business case draft 1 - GOV.UK

Making Earth Observation Work for UK Biodiversity Conservation - Phase 3

Environment Systems Ltd. ii

Glossary

APGB: Aerial Photography for Great Britain contract Ancillary data: Data from sources other than remote sensing that is used to assist in analysis and classification. Annex I habitats: Habitats listed in Annex I of the Habitats Directive; habitats with a high conservation value Aspect slopes: Referring to the compass direction a slope is facing BAP Priority Habitats: Habitats listed as priority habitats in the Biodiversity Action Plan Biogeographic region: An area of floral or faunal distribution with similar or shared characteristics Catchment: The entire area of a landscape that drains into the same river Crick framework: A ‘multi-scale’ approach to habitat mapping, utilising differing scales of imagery and supporting data and concurrently producing a range of harmonised outputs that are relevant from a landscape scale to a particular site eCognition: Software for developing segmentations and rule-based classifications Ecosystem approach: An approach to land management that integrates management of land, water and living resources EO: Earth Observation Geoinformatics: Combining and modelling spatial datasets Invasive species: Species that are non-native to an environment in affect said environment negatively; after habitat loss considered the second greatest threat to biodiversity Landsat: Land Satellite – 30m Satellite imagery available from the USGS (United States Geological Survey) LiDAR: Light Detection and Ranging – Air borne sensor which gives vertical height data Monte Carlo simulation: A problem solving technique used to approximate the probability of certain outcomes by running multiple trial runs, called simulations, using random variables. Multispectral imagery: Imagery captured by dividing the spectrum into many bands; main type of images acquired by remote sensing radiometers NIR: Near infrared a spectral band in aerial and satellite imagery NVC: National Vegetation Classification (Rodwell, 1991a, 1991b, 1992, 1995 and 2000) OBIA: Object based image analysis – process of classification where objects are considered rather than individual pixels Phenology: Referring to cyclic and seasonal natural phenomena Pre-processing: Ortho, atmospheric, topographic and other corrections to prepare imagery for classification. Productivity: The rate at which new biomass is produced by an individual, population or community PSMA Public Sector Mapping Agreement, RGB: Red green blue, also known as true colour often used for referring to aerial photography RPAS: Remotely Piloted Aircraft Systems Rule-base: A series of structured statistical rules (e.g. NDVI < 0.9) applied to satellite imagery, airborne imagery and/or thematic data layers to produce a user defined map. SAC: Special Area of Conservation Segmentation: Grouping of pixels based on similar values – a type of automated vectorisation (digitising). Semi-natural habitats : These are habitats where human induced changes can be detected or that is human managed, but which is primarily formed from a natural habitat in terms of native Spectral signature : Information that is recorded for each image pixel or image object which is related to the species or habitats present in the pixel.

SPOT: Satellite Pour l'Observation de la Terre, French satellite supporting the HRG sensor SSSI: Site of Specific Scientific Interest UAV data: Data obtained with and Unmanned Aerial Vehicle (i.e. an aerial platform that can carry a sensor)

Version: Draft v2. 12/01/15

Parker, J. A., Medcalf, K. A., Turton, N., Hussein, M. H, Smith, G., Breyer, J. Making Earth

Observation Work for UK Biodiversity Conservation – Phase 3: Material for an Outline

Business Case for a national EO service for habitat mapping.


Environment Systems Ltd. iii

Contents Glossary.................................................................................................................................ii Material for an Outline Business Case for a national EO service for habitat mapping............ 1

1 Introduction ................................................................................................................ 1

2 The Crick approach as part of a habitat toolkit ............................................................ 1

2.1 Vision for the Crick approach ............................................................................... 2 2.2 The Crick approach ............................................................................................. 2 2.3 Objectives of the Crick approach ......................................................................... 4 2.4 What are the main benefits? ................................................................................ 5 2.5 Adaptability, transferability and practicality .......................................................... 6 2.6 Key uses and users ............................................................................................. 6

3 Assessing change in the landscape .......................................................................... 12

4 Alignment with strategic initiatives and surveys ........................................................ 14

5 Why scale up and roll-out the approach across the UK? .......................................... 17

5.1 Why scale up? ................................................................................................... 17 5.2 Why begin roll-out of the Crick approach now?.................................................. 18 5.3 When would benefits be realised? ..................................................................... 18

A NATIONAL EO SERVICE FOR HABITAT MAPPING ...................................................... 20 6 What will a national habitat assessment service comprise and how might it be

delivered? ........................................................................................................................ 20

6.1 Data components .............................................................................................. 21 6.2 System components .......................................................................................... 23 6.3 Mapping Process ............................................................................................... 24 6.4 People (delivery partners) ................................................................................. 26

7 Delivery model .......................................................................................................... 29

7.1 External Dependencies ..................................................................................... 30 7.2 Risks and Issues ............................................................................................... 30

8 Business Plan........................................................................................................... 31

9 Assumptions and presentation of costs .................................................................... 31

9.1 Scaling up of costs ............................................................................................ 33 9.2 Further analysis of factors influencing the cost of rollout and their effects.......... 38

10 Demonstrating effectiveness of the techniques for Habitats Directive reporting .... 40

10.2 A rolling programme of delivery ......................................................................... 42 11 Value for Money .................................................................................................... 46

11.1 The value of information to improve ecosystem management for biodiversity (Value for Ecosystem Services) ................................................................................... 46 11.2 Comparisons with the costs of current initiatives (Cost Comparison) ................. 47 11.3 Value for scientific purposes (Science Value) .................................................... 48 11.4 Avoidance of fines for failure to meet obligations to protect the highest value designated land in the UK (Avoidance of Fines) ........................................................... 49

References ......................................................................................................................... 50 Appendix 1: Fuller description of potential project stakeholders & other interested parties .. 52 Appendix 2: Risk register .................................................................................................... 55


Environment Systems Ltd. iv

Table of Figures

Figure 1: Crick Framework .................................................................................................... 5

Figure 2: Flow chart of the process of mapping using the Crick approach, orange lines show

the alternative pathway for areas with existing good mapping requiring update .................. 24

Figure 3: Local Nature Partnerships .................................................................................... 28

Figure 4: Histogram of distribution of predicted per unit cost ............................................... 34

Figure 5: Tornado plot showing sources and extent of effect of the mapping stages on the

predicted aggregate costs for Norfolk .................................................................................. 35

Figure 6: Illustrative costed rolling programme of initial mapping to support Habitats Directive

Reporting ............................................................................................................................ 44

Figure 7: Illustrative costed rolling programme of updates to support Habitats Directive

Reporting ............................................................................................................................ 44

Figure 8: Discounted total cost of a rolling programme of mapping to support Habitats

Directive Reporting ............................................................................................................. 44

Table 1: Data outputs ............................................................................................................ 3

Table 2: Range of stakeholders identified ............................................................................. 7

Table 3: Examples of adoption and further use of products derived from the Crick approach 9

Table 4: Meeting local and national needs for information about habitats............................ 11

Table 5: Summary of change detection techniques ............................................................. 13

Table 6: How a National EO service for habitat mapping can help address known habitat

surveillance challenges ....................................................................................................... 14

Table 7: Sequence in which Crick approach outputs / benefits will be made available ........ 19

Table 8: Imagery and data requirements ............................................................................. 21

Table 9: Required stages of the Crick approach .................................................................. 25

Table 10: Potential involvement of key delivery stakeholders* ............................................ 26

Table 11: SWOT analysis of the delivery models ................................................................ 29

Table 12: Average cost of mapping a unit area (km) ........................................................... 32

Table 13: Scaling factors for consideration when costing mapping of different zones of the

UK ....................................................................................................................................... 37

Table 14: Projected cost per country for initial mapping ...................................................... 37

Table 15: Projected cost per country for updating the mapping ........................................... 38

Table 16: Initial mapping costs for Norfolk ........................................................................... 39

Table 17: Indicative cost of map update for Norfolk ............................................................. 40

Table 18: Proposed ranges of costs of different techniques for a hypothetical SPOT scene 48


Environment Systems Ltd. 1

Material for an Outline Business Case for a national EO service for habitat mapping

1 Introduction

This work forms part of Phase 3 of the Defra and JNCC funded project: Making Earth

Observation (EO) Work for UK Biodiversity Conservation.

This document forms the full report on the outline business case and rationale for a national

EO service for habitat mapping. It is supported by two short summary documents that

present the main findings of this report.

A business case is a document that should accompany a project from initial concept to final

implementation. This document provides a preliminary assessment, in the form of material

suitable for inclusion in an Outline Business Case1 (see: Treasury Green Book), which can

be refined and developed to produce a full business case. One of its main purposes is to

help potential sponsoring organisations make decisions about their support for the rollout of

a national EO service for habitat mapping of the Crick approach.

It presents the rationale for adoption and rollout of EO habitat mapping across the UK using

the Crick approach to demonstrate that the spending proposal provides a strategic fit with

policy needs at both national and local levels and should deliver synergy for organisations

operating at both these scales. An initial high level assessment describes how rollout of the

Crick approach can be achieved, presenting options for delivery, with a preferred option

identified. Initial cost estimates are provided for the creation of a national EO Habitat

Mapping Service at the country level. The cost-effectiveness of the approach for Habitats

Directive (Article 17) reporting of Annex I habitats at the national scale is examined and a

value for money assessment provided.

It draws on the experience of three related Defra research projects. These compared the

relative merits of the Crick approach with other habitat mapping methods and presented the

results of testing the habitat mapping methods in four differing environments in the UK. It

also draws on the experience of producing the Habitat Inventory for Wales2 and habitat

mapping in the UK Overseas Territories3 that employ very similar techniques.

2 The Crick approach as part of a habitat toolkit

Habitat surveillance and monitoring is crucial to provide a broad picture of the status (extent,

distribution and condition) of semi-natural habitats across the UK, so that we know enough

about specific species and habitats of conservation concern to provide protection, meet our

legislative obligations and ensure that sites are managed appropriately. Good knowledge of

the status of semi-natural habitats meets a range of other policy requirements, land manager

1https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/190609/Green_Book_guidance_short_plain_English_guide_to_assessing_business_cases.pdf

2 http://www.gwylio.org.uk/ and Lucas et al. 2011

3 http://www.cieem.net/data/files/Resource_Library/Conferences/2013_OT_SIG/OT2013-KatieMedcalf.pdf

https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/190609/Green_Book_guidance_short_plain_English_guide_to_assessing_business_cases.pdf

https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/190609/Green_Book_guidance_short_plain_English_guide_to_assessing_business_cases.pdf

http://www.gwylio.org.uk/

http://www.cieem.net/data/files/Resource_Library/Conferences/2013_OT_SIG/OT2013-KatieMedcalf.pdf



and planning considerations and increases our understanding and appreciation of our

natural resources, as evident in the ecosystem services approach.

Habitat practitioners already use field survey and remote sensing techniques. The

techniques developed in the Crick approach are designed to enhance this “toolkit”, providing

further choice with a tested and proven approach.

The approach and framework are named after Mark Crick of the JNCC, who worked hard to

develop and promote the use of remote sensing in habitat mapping.

2.1 Vision for the Crick approach

To strengthen the evidence base within existing resource constraints so that there is

consistent and robust information about habitats in the UK to support policy and land

management decisions, and help monitor the impact of these decisions.

2.2 The Crick approach

The Crick approach is a ‘multi-scale’ approach to habitat mapping, utilising differing types of

imagery covering a range of spatial scales and spectral information, with supporting data to

produce a range of harmonised outputs that are relevant at both landscape scale to a

particular site.

In the Crick approach, state-of-the-art Earth Observation analysis mapping techniques,

imagery and supporting data are used to:

provide knowledge about where particular habitats are located, or potentially located,

within distinct regional landscapes (with a current focus on Annex I and Priority

Habitats);

provide a range of other outputs as part of the mapping process, such as more

detailed localised habitat maps, and measures to support the assessment of habitat

condition;

provide a ‘living map’ that can be improved upon by local bodies and updated

periodically when new imagery becomes available;

map to real world objects, such as field parcels, so that outputs are recognisable and

practical for use at the local level.

The Crick approach is “a multi-purpose habitat and ecosystem measurement and mapping

system” that can produce different outputs from a stock of data sources. The MEOW outputs

can be more easily reworked, revisited, improved upon and merged with other information

than traditionally generated habitat maps.

The approach utilises ecological desk survey and familiarisation fieldwork to understand the

habitats and biogeographic context of habitats within the area. Geoinformatic techniques are

then used to split the environment into components for analysis, with multi-temporal, multi-

spectral and multi-spatial data. Objects are created at a suitable for the habitats under

consideration within each component of the landscape, and rules are developed to

distinguish them from one another from spectral and other supporting data such as slope

and heterogeneity. The rule base is then subject to field evaluation and can be re-iterated

where habitats are not being distinguished at sufficient levels of reliability. As well as the final



classification, a range of data products which fit within this spatial framework can be output

to describe the ecology of the area including known or potential conservation interest.

Integral to the mapping process from the outset, is the engagement and input of local

bodies. Additional follow-up fieldwork thus building on the habitat information provided.

If there is an existing habitat mapping (e.g. a local phase 1 map) this can be used as a start-

point for production of ‘the complete habitat map’, a rule based is produced to identify

changes that can occur in the ecological envelope of possible changes. This includes habitat

boundary and vegetation classification change, poor original spatial representation of

features can be updated to better represent the features on the ground.

If this follow-up work is done as part of a structured process, the Crick approach can provide

a framework for the collation of information suitable for national reporting of some aspects of

the status of habitats (primarily range and extent). The map evolves and is strengthened

through local fieldwork and can also be revisited and improved upon periodically as new

aerial or satellite imagery becomes available. As such, it is ‘living map’ that forms part of a

habitat mapping toolkit.

A range of measures of the condition of habitats are being developed, these do not equate

to the condition measures used in Annex I reporting but rather, are designed to support land

managers with site management (e.g. mapping invasive species, changes in site wetness

providing measures of vegetation productivity or site nutrient status within and surrounding

sites). Such factors are difficult to assess on the ground, and again, the Crick approach

seeks to enhance the habitat toolkit.

The outputs (listed in Table 1) produced using the Crick approach can be revisited,

improved upon and merged with other information to provide further outputs to support local

needs, for example for mapping ecosystem services, connectivity analysis or modelling

species distributions.

Table 1: Data outputs

Outputs of the Crick approach

Description

Complete coverage habitat map

Mapping of the complete range of features, including agricultural land

and semi‑natural vegetation across regional areas with more detailed

mapping in areas of semi-natural habitats.

Information and maps to support recording of specific habitats (e.g. Annex I and Priority Habitats)

It is possible to extract the likely range, and in some cases the specific extent of internationally significant habitats from the ’complete coverage’ map. This can be used for targeting fieldwork, thus providing a structured process to support local collation of information for national reporting.

Site scale habitat map

In more localised areas, such as nature reserves, known to contain semi-natural habitat, additional information can be used such as very detailed imagery to look at specific features and condition of habitats.

Additional system outputs

The complete coverage habitat map can be re-worked to provide outputs that are suitable for use as map layers to support the analysis of ecosystem goods and services because habitat maps describe the structural function of the vegetation, condition of habitats and surrounding land cover.

Mapping involves a multi-stage process:



Stage 1: Data acquisition

Stage 2: Image processing

Stage 3: Preparation of supporting data

Stage 4: Familiarisation fieldwork

Stage 5: Rule base development and implementation

Stage 6: Testing fieldwork

Stage 7: Product generation

Stage 8: Gap filling through fieldwork to get to Annex I habitats

2.3 Objectives of the Crick approach

Implementation of the Crick approach has the following objectives:

support national, regional and local activity to improve knowledge of the status of

habitats to ensure their continued protection and appropriate management;

strengthen the evidence base for statutory reporting for the Habitats Directive (Article

17 reporting for Annex I habitats) within existing resources; and,

provide a platform and improve the capacity and capability, particularly at the local

level, to use EO and geoinformatic techniques to respond to the growing need for

information to support the conservation of natural resources.

Working with local partners, ecological knowledge is incorporated into image classification.

With local knowledge of landscape and ecological context applied to image classification, it

is possible to delineate areas in which particular habitats are restricted. This allows the user

to separately map vegetation assemblages which are spectrally similar but that occur in very

different settings (thus avoiding sources of confusion in the image analysis). A ‘rule-base’

describing the spectral and ecological and landscape characteristics of different habitat

types can then be developed and applied, allowing maps to be constructed using an expert

automated system.

The approach is supported by The Crick Framework which describes the capacity of EO to

support the identification of Annex I and Priority Habitats and the requirements for habitat

mapping. A wide range of interacting factors has been considered along with ecological

knowledge, to develop a generic classification system that proposes categories (tiers) of

habitat groups (Figure 1).

For instance, small scale or narrow habitats such as field margins can only be mapped with

spatially detailed image data, and are therefore placed in tier 2b or 3b. Similarly, certain

habitats are only associated with particular geological substrate conditions so are likely to be

in the tier 2c or 3c.



Figure 1: Crick Framework

2.4 What are the main benefits?

The Crick approach allows us to understand which habitats are present, where they occur in

the landscape and the condition these habitats are in. The approach also:

complements existing methods: supporting and augmenting existing techniques in a

practical, cost-effective way (see section 9), being particularly effective in targeting

fieldwork;

provides a range of useful outputs from one process, all of which are relevant to

current policy and which have the capacity to meet local, regional and national

needs;

is adaptable: the existing rule base and data can be improved upon and merged with

other information to provide further outputs to support and address future needs;

captures existing knowledge (previous habitat maps) as part of the mapping

process;

is transferable: pilot projects have demonstrated that the approach can be used in

the differing environments of the UK, subject to suitable imagery being available;

is fast, efficient and practical: a very much larger area can be classified at one time

than is possible under existing fieldwork-based methods.

enhances local, regional and national capability and knowledge of the use of EO and

geoinformatic techniques.

To ensure continuity, existing habitat data (where it is of adequate quality) should form the

starting point for mapping using the Crick approach, the practicalities of this are further

discussed in section 1.



2.5 Adaptability, transferability and practicality

It is possible to produce similar outputs to those of the Crick approach using existing

methods, such as visual and automated interpretation of aerial photography, fieldwork,

automated mapping techniques or interpretation of existing habitat maps and data. In this

respect, the outputs produced by the Crick approach are not unique but the process itself

has particular advantages.

The techniques developed are considered to be:

fast and efficient: the Crick approach requires experienced ecologists and remote

sensing specialists but a very much larger area can be classified at one time than is

possible under existing methods (e.g. production of a Phase 1 map required a 10-

year field survey campaign in Wales, whereas the Wales Habitat Map, which was

developed using similar EO based techniques took 6 years, about half of which

involved development of the techniques).

adaptable: the developed rule bases and processed imagery and data can be re-

analysed, built upon or adapted as necessary to produce additional outputs or

products tailored to other policy needs that utilise habitat mapping (e.g. to map areas

using a different habitat classification system or to provide a more detailed

assessments of a site or area). It can draw in information from existing habitat maps

and data. This keeps the cost of follow-on work very low in comparison with repeat

survey, manual re-interpretation of imagery or other traditional field survey

techniques and allows good use to be made of existing knowledge. It also means

that multiple outputs relevant at the local level can be produced during one mapping

process with some tailoring of outputs to meet specific local needs.

transferable: to all UK environments (evidenced by work in Norfolk, Wales, upland

environments in England and Scotland, ongoing in this project). From a technical

perspective successful transfer of the techniques depends on understanding the

ecological drivers of the area (best sourced from local experts) and obtaining imagery

at suitable spatial and temporal scales for the habitats present in the area of interest.

practical: providing a technique for directly mapping some habitats of high

conservation value as well as providing information showing the location of grassland

habitats of potentially high conservation value for further field validation at an

individual species level. In common with other maps derived from EO, the techniques

are particularly efficient for mapping areas with an extensive coverage of semi-

natural vegetation (e.g. upland areas) and difficult to reach areas, such as coastal

areas (where they can improve upon the accuracy of habitat mapping of individual

sites).

2.6 Key uses and users

Through stakeholder consultation and known early adopted uses of the approach (expanded

on in section 2.6.1), a range of potential uses and users were identified. The techniques

outlined in the Crick approach are considered to be particularly valuable for:

assisting with filling gaps in knowledge of the presence and extent of habitats in the

wider countryside;

improving the spatial definition of habitats within designated sites;



generating habitat maps and data to meet a wide range of landscape scale

approaches to biodiversity delivery;

generating map layers to support the analysis of ecosystem goods and services,

ecological networks, ecological restoration, pollution modelling, species distribution

modelling and threat maps for non-native species and plant and animal health

issues;

producing evidence for management plans for larger sites or discrete areas and to

inform local planning;

identifying threats to habitats and ways of mitigating against and monitoring these (as

evidenced through pilot projects, for example mapping the presence and extent of

alien species).

They are best suited to addressing evidence needs arising from the following policy drivers:

United Nations Convention on Biological Diversity (Aichi targets);

EC Habitats Directive;

Biodiversity 2020 and country strategies;

National Ecosystems Assessment;

Water Framework Directive;

Environment Protection Act (Common Standards Monitoring).

Defra, the Devolved Governments and some other government departments (such as the

Forestry Commissions) are responsible for policies and regulations for the natural

environment and together with the JNCC, the advisory body on nature conservation issues

for the UK, and delivery partners (statutory conservation agencies and Environment

Agencies) are key users.

There are a very large number of stakeholders with an interest in funding, delivering or

making direct use of the outputs the Crick approach. In addition to stakeholders who are

users of the outputs, there are suppliers and organisations with a more strategic interest in

the Crick approach (Table 2 and Appendix 1). For many stakeholder organisations

(especially those with a national, or UK wide remit) there are multiple potential uses for the

outputs of the Crick approach (Parker, J., et al. 2014).

Table 2: Range of stakeholders identified

Stakeholder Description

Defra UK Government department

JNCC Public Body that advises the UK Government and Devolved Administrations on UK-wide and international nature conservation.

Devolved Administrations Scottish Government – Environment and Rural Affairs Dept. Welsh Government – Directorate of Sustainable Futures Government of Northern Ireland – Department of the Environment

Other Government Departments

Forestry Commission England Forestry Commission Wales DECC, MOD, DCLG, DFID, DoT

Country Conservation Agencies

Natural England, Scottish Natural Heritage, Natural Resources Wales & Northern Ireland Environment Agency

Environment Agencies Environment Agency Scottish Environment Protection Agency Northern Ireland Environment Agency



Stakeholder Description

Local Authorities Local Authorities (426 in UK)

Organisations responsible for designated landscapes

National Park Authorities (England, Scotland, Wales) National Parks England, National Parks Wales AONBs, National Scenic Areas

Partnerships (at local scale)

Local Nature Partnerships (LNPs – 50 areas in England, some overlap), Local Sites Partnerships (LSPs – 53 in England), ‘Living Landscapes’ (Wildlife Trusts – ~100 in England & Wales) Other partnerships (e.g. IDBAs)

Organisations involved with the pilot project

Northern Upland Chain LNP, Yorkshire Dales NPA, Northumberland NPA, NEYEDC (data centre), CEH, Norfolk Biodiversity Information Centre, Nidderdale AONB, North Pennines AONB, Norfolk Wildlife Trust and RSPB.

Organisations responsible for transport corridors

Canal & River Trust (England, Wales) Network Rail, British Waterways / Scottish Canals Highways, Inland Waterways (NI),

Organisations with a strategic interest

UK Space Agency - Space for Smarter Government Programme, UK Earth Observation Framework

National Ecosystem Assessment

The UK National Ecosystem Assessment, The UK National Ecosystem Assessment Follow on

Other ecosystems services initiatives

BESS biodiversity & ecosystem service sustainability, Ecoserv-GIS, Site Informer, Polyscape, Invest, Aries, SENCE,SCCAN, JNCC Spatial Framework

Scientific research organisations

CEH (UK soil observatory, Countryside Survey), FERA (Agricultural Change and Environment Observatory Programme), Forest Research, James Hutton Institute

Key data providers

Copernicus (satellite imagery), Ordnance Survey (MasterMap), Rural Payments Agency (RLR), Environment Agency (LiDAR) Country Conservation Agencies (aerial photography (APGB)/ wide range of other mapped data). Licencing and implications of this is discussed in section 11.1.

Specialist groups with specific remits

UK Biodiversity Indicator Steering Group UK Biodiversity Indicator Forum GB Non-native Species Secretariat / Invasive Species Ireland Strategic Regional Coastal Monitoring Programmes Terrestrial Biodiversity Group, Natural England National Trust

As the Crick approach captures data on the habitats of the whole landscape at a relatively

fine scale, including features such as hedgerows, field margins, scrub and other patches of

habitat, it is also useful for local mapping and modelling purposes such as ecosystem

service assessment, offsetting analysis, water framework mapping and modelling and

biodiversity modelling. At the local scale, Local Authorities and local partnerships are

important users of the approach, particularly with regard to the delivery of sustainable

development through planning.

The Crick approach can be used to consider aspects of ‘condition’ and change in condition

for specific sites where important habitats, such as heathland, occur. Research is underway

to develop approaches to support condition monitoring as a range of potentially useful

outputs have been identified. The Crick approach could be used to monitor factors such as



productivity of the vegetation within a site and on the surrounding land with the outputs

feeding into a risk based approach to monitoring of designated sites.

2.6.1 Evidence of demand

Recent consultation with key national stakeholders carried out as part of a review of CS2007

(Haines-Young et al, 2012) concluded that:

“for habitats and species of conservation importance (e.g. those listed in the Habitats

Directive) detailed information on their distribution and condition will be needed for reporting

purposes in relation to 2020 targets, and for assessing the impact of policy interventions

designed to sustain ecological function and the integrity of our natural capital. The success

of measures to create coherent and resilient ecological networks will also be a focus for

future monitoring in some areas”.

At least five percent of England’s land area is Local Wildlife Site. A recent survey (Wildlife

Trusts, 2014) of 48 of the 53 Local Wildlife Site (LWS) partnerships across England found

that due to resource constraints only about 6% of LWSs are being monitored each year; in

some areas none were being monitored due to a lack of resources. Of 6,590 LWSs

monitored between 2009 and 2013 more than 11% of sites were found to be lost or

damaged. Five partnerships confirmed that no monitoring was undertaken in the last five

years (these partnerships collectively have 11% of England’s total number of sites). Of the

47 partnerships that responded, 43 stated that additional resources were needed for survey

and monitoring and 23 partnerships required resources for site selection. With limited

resources, support from the Crick approach to assist with targeting areas likely to be sites of

local importance (Table 3). It can also be used to highlight where there is a risk of negative

change in condition of a site and this is likely to be beneficial to many partnerships, for

example the expansion of stands of species like European gorse, or increasing vegetation

productivity, suggesting an input of nitrate.

The production of the National Habitat Map of Wales and pilot work in Norfolk carried out as

part of Phase 2 of this research (Medcalf et al. 2013) have generated demand for further

mapping and the production of additional outputs using the maps and rule-bases developed

(Table 3). In Norfolk, where the approach was piloted in 2013, local adoption has already

resulted in several organisations using the map for such purposes; further uses are planned

and volunteers are validating information already provided.

Table 3: Examples of adoption and further use of products derived from the Crick approach

Users involved Description

Projects in Norfolk (utilising the mapping produced for Norfolk during Phase 2 of this research project)

Norfolk Biodiversity Information Service, which is County Council (CC) funded

Following pilot mapping in in 2013, the County Council adopted the Crick approach, and funded the mapping of a ‘complete coverage habitat map’ for the rest of the county. Structured methods are being developing for validating the map in detail using volunteers

The complete coverage habitat map is being used to create an ‘Ecological Network’ map, which will be used to analyse habitat connectivity

A map of ‘threats’ is being produced to support activities that concern non-native species issues; this includes the mapping of Himalayan Balsam on the Norfolk




broads from ultra-high resolution data.

Old Phase 1 habitat maps from the 1980’s are being digitised so they can be used with the new map

Planned uses (not yet started): Green Infrastructure mapping, opportunity mapping, condition monitoring at a site level. Uses being considered: Ecosystem Service analysis/modelling support for planning decisions.

Broads Authority

Updating existing pollution and sediment models.

Norfolk Wildlife Trust

Using the ‘complete coverage habitat map’ as baseline for Living Landscape areas to target restoration.

Direct site management: a detailed site based map in East Wretham Nature Reserve is being used to inform plans to monitor rabbit numbers and nutrient spread.

BTO Used in predictive models, with bat records to predict future bat species distributions in Norfolk.

Projects in Wales (utilising the Phase 1 mapping of the whole of Wales using a similar OBIA based approach)

Glamorgan CC & Carmarthen CC

Re-analysis of satellite data used in the national phase 1 map to produce a bespoke ‘traffic light map’ of potentially species rich grassland to target fieldwork to identify additional Local Sites (also known as SINCs, Local Wildlife Sites).

NRW

National ecosystem services mapping across Wales to support the National Environment Framework http://www.werh.org/natural-environment-framework.php.en feeding into the UKNEA

Bridgend CC Ecosystem services mapping in Bridgend County Borough Council.

WAG Mapping of pollination ecosystem services in the Dyfi Biosphere reserve.

NRW

National ecosystem services mapping across Wales to support the National Environment Framework http://www.werh.org/natural-environment-framework.php.en

Case study of the application of ES mapping in Cambrian Mountains National Park (for CCW)

Targeting agri-environment option placement in the Glastir scheme

Caerphilly CC Identification of historic features in the landscape (e.g. historic field patterns, hill forts)

Other work

Anguilla (UKOT’s)

Habitat & ES mapping in terrestrial and marine environments. Mapping and analysis of shoreline change.

Dorset CC Habitat mapping, connectivity mapping and some ES mapping.

ADAS UK Ltd

Mapping of areas of bog of potentially high nature conservation value in NI across a mountain range to target field effort to find the best potential intact blanket bog sites. Rapid response project with a short timescale to complete the project over a very wide area.

Isle of Man update of

The Isle of Man has an existing Phase 1 habitat survey which is mapped to field parcels. This was updated using remote sensing to attribute each polygon with a

http://www.werh.org/natural-environment-framework.php.en

http://www.werh.org/natural-environment-framework.php.en




vegetation likelihood of change; this layer can then be used to target field work to update the map in a very cost effective way.

Dartmoor pilot

The mapping tasks for assessment included aspects of blanket bog condition assessment (mapping of peat hags, heather distribution and drains), archaeological prospecting and inventory, historical mining analysis, moorland and heathland extent, and hedgerow/field boundary mapping. LIDAR and CASI (hyperspectral) data were used for the tasks in this pilot study, with Geomatics Group surveying approximately 18% of the National Park (180 km

2), capturing 50cm resolution LIDAR

data and 1m resolution CASI data.

There are common national and local needs and some differences in needs, mainly

concerning the particular habitat classification of outputs of the Crick approach (Table 4).

These can be accommodated by the approach as the outputs can be related to differing

habitat classification systems.

Table 4: Meeting local and national needs for information about habitats

User group Summary of user need Recording systems into which

outputs need to integrate

Notes

Both local and national

Condition monitoring of SAC’s / SSSI sites and their surrounding areas to identify risks to sites

Risk features:

single stand potentially invasive species;

increased nutrient / productivity within site;

increased nutrient / productivity

on land surrounding the site;

change in site features such as wetness

Such risk features could form a useful ‘early warning’ or risk based monitoring approach.

National needs

Description of Annex I and BAP Priority Habitats in each area

Extent and location:

Annex I habitats

BAP Priority habitats

Local needs Description of habitats across the landscape in accordance with a recognised classification.

Inclusion of features such as hedges and areas of scrub

Phase 1 habitat survey

IHS

EUNIS (level 3)

Locally and nationally significant habitats



3 Assessing change in the landscape

Two main uses of habitat maps are to map the existing stock of habitats in an area and to

consider change in stock. Both factors are equally important and remote sensing can be

useful in both situations. Where there is an existing good quality habitat map then remote

sensing can be used to identify change and update the mapping. Where there is not

suitable mapping then remote sensing can be used to create such a map.

Where there are existing habitat maps which are widely accepted as having been an

accurate description of an area, then the Crick approach should work to update and validate

the change in these areas, rather than giving a completely new view of the area. This will in

some cases involve breaking mosaics into individual polygons, but by working within the

existing maps and policy or figures derived from the existing maps can be built upon and

used for change analysis rather than establishing a completely new baseline. This reuse of

existing field data is only possible where a good degree of cartographic accuracy has been

possible related to aerial imagery. This change detection map can be further enhanced by

fieldwork, which can be fed back into the map, the final product becoming a ‘living map’.

One of the key uses for habitat data is to map existing stocks of high quality and

internationally significant habitats and to periodically record any change in the habitat.

Habitat change has traditionally been assessed on a site by site basis either in the field or by

visual comparison of aerial photography. This normally entails doing a completely new

survey and comparing the results to the baseline survey. Therefore each survey tends to

cost a similar amount. Additionally change in condition of habitats is driven by serval key

factors such as the presence of monotypic invasive species such as gorse or a change in

water regime or productivity. Measuring and mapping the extent and nature of change in

condition is quite complex to do in the field, remote sensing can assist with some aspects of

assessing change in habitat condition, such as the change in vegetation extent, increase in

bare ground or peat and changes in productivity.

The ability to make comparisons in habitat status over time is vital as it provides the

evidence for informed action in the delivery nature conservation and sustainable

management. Field surveyors, policy makers and other users are keen to achieve

comparability between existing Phase 1 habitat maps, other spatial ecological data derived

from field surveys and maps created from remotely sensed images. As the Crick approach

uses a rule-base, Phase 1 is just one of many possible habitat outputs, combining several

sub-classes which can be combined in other ways for different habitat classification systems

such as Eunis.

Three change detection techniques are available for EO approaches (Table 5). A study

carried out in Wales following the Wales habitat mapping project found that ‘map to image’

comparisons provided the most cost effective and repeatable technique for change detection

and map update (Breyer, 2014).



Table 5: Summary of change detection techniques

Change detection

type

Description

Image-to-image

Direct comparison of two images based on spectral values. Requires year-round data with comparable phenology, otherwise seasonality has too great an influence on differences between images, indicating much more change than has genuinely occurred.

Post-classification

Comparison of map classes after classification of land cover from remotely sensed data. Limited by the confidence in either map or a large range of variables (input data, ecological interpretation) during the mapping phase.

Map-to-image

By knowing the likely habitat from the existing map the expected spectral values for each object can be predicted at the time of year of any new image. Objects which do not match this expectation can be identified either for further analysis or field visits to determine the new habitat. Where habitat parcels are expected to be quite homogenous, increased variability in the spectral response would be a further indicator of change in at least part of the object.

Reliable base-line data is essential for change detection. Once the Crick approach has been

used to map habitats and other features in the landscape (e.g. fields and distinct areas of

semi-natural vegetation, measures of condition) the process required to update this

information becomes less time consuming. This is because knowledge has been captured

using a ‘rule-base’ that defines each habitat type in terms of spectral and ancillary data. The

recorded thresholds and features used to classify the habitats (e.g. field boundaries) can be

used as a start-point to compare with more recently acquired imagery. Data held in the ‘rule-

base’ can be re-used to set criteria such as thresholds designed to highlight areas of

potential change within existing objects. For instance, grasslands that have been improved,

changes in the extent of heather or trees, and changes in bare or built environments, could

be automatically identified for further investigation.

The greater the conformity between existing data and updated maps created from new

technologies, the greater the likelihood of successful change detection. In some instances

Phase 1 surveys already exist for an area. Where these surveys have captured appropriate

objects, such as field boundaries, vegetation parcels and OS MasterMap features, it is

possible to detect change from these initial objects. An ongoing study with the Isle of Man

Government is using this approach to detect the change in the landscape since a previous

Phase 1 habitat survey.

To ensure continuity, existing habitat data (where it is of adequate quality) should form the

starting point for mapping using the Crick approach. As part of Phase 3 of the MEOW

project, this approach is being trialled. In the North York Moors some of the larger upland

polygons mapped using Phase 1 survey are being split into their component mosaic habitats

based on spectral signatures, this will allow a finer grain of habitat classification and change

detection.

The Welsh study following the Wales habitat mapping showed that before undertaking

change detection, an assessment of the likelihood, level and direction of change of any class

should be undertaken. Change detection needs to be developed into a rolling monitoring

system.



There is potential to develop a similar monitoring approach for measures that indicate the

condition of habitats such as a ‘wetness index’ but this requires further investigation to

develop indices to ensure the measures produced are directly comparable using differing

image sources in differing bio-geographical environments. The metrics investigated are

currently able to determine relative values. In order to turn these metrics into quantifiable

condition monitoring techniques, more research is needed into their response to the daily,

seasonal and climatic perturbations of the natural environment to identify real change to a

site.

Frequency of surveillance of different parameters (e.g. range/extent, condition) should be

appropriate to their sensitivity to change (for example, range is often less sensitive to change

than condition). The required frequency of condition assessment is ideally determined by

local threat and management needs.

4 Alignment with strategic initiatives and surveys

JNCC Surveillance Strategy: The JNCC have identified several significant strategic

challenges that affect surveillance and monitoring functions at country level (JNCC, 2013),

these are summarised in Table 6.

Table 6: How a National EO service for habitat mapping can help address known habitat surveillance challenges

Challenges and

issues Result

How can a National EO service for habitat

mapping assist?

There are European requirements for UK scale reporting

4,

however there are few specific legal requirements for UK scale surveillance or monitoring.

Overall funding for biodiversity is low compared to many areas of government policy. Surveillance spending is piecemeal, meeting multiple needs. The outcome is that there is relatively poor knowledge of the status of habitats outwith the statutory series. Available information cannot be readily compared or integrated

5.

Can bring consistency of output and accuracy/ certainty to support ongoing UK scale reporting requirements and improved knowledge of habitat status, esp. outside the statutory series. This will reduce the risk of failure inherent in supplying inaccurate or challengeable figures used to address reporting needs. Comparable mapping across the UK would provide a high level dataset to support species and habitat sampling and modelling.

Budget constraints and changes to evidence delivery functions

No new money will be available to fund additional mapping

Rollout of a national EO service for habitat mapping will need to be funded from current biodiversity budgets.

An increased focus on ecosystem services and natural capital

The JNCC foresee an increased demand for monitoring to provide effective evidence at a local scale

6, and more autonomy

in local delivery and evidence is likely.

Rollout of a national EO service for habitat mapping supports delivery at the local level with landscape scale maps and additional system outputs that can meet this demand.

Increased scrutiny and challenge on evidence quality.

Need to be able to provide transparent evidence to substantiate policy reporting

7.

The outputs, rule-base and measures of certainty are held in the map. Consistency of approach and its application can be demonstrated.

4 http://jncc.defra.gov.uk/PDF/comm13%20D07.pdf

5 http://jncc.defra.gov.uk/PDF/uktbss_Surveillancerationale.pdf

6 & 7

http://jncc.defra.gov.uk/PDF/comm13%20D07.pdf


http://jncc.defra.gov.uk/PDF/uktbss_Surveillancerationale.pdf




The value of national habitat mapping has the potential to be beneficial beyond the needs of

Defra and JNCC for habitat surveillance (e.g. in providing a framework for helping

harmonise a range of monitoring programmes) and there is good strategic alignment with the

following programmes and initiatives:

European Space Agency (ESA): ESA plans to launch several new ‘Sentinel’ satellites as

part of its Copernicus programme. Launched in April 2014, Sentinel-1 is already operational,

capturing radar data of the land and sea, the data produced is too coarse for use in the Crick

approach but can be used for wider scale condition trends. Sentinel-2 is a polar-orbiting,

multispectral high-resolution imaging mission for land monitoring that is due to launch in

2015/16. This satellite will have a very high pass rate over the UK and will provide suitable

imagery for use in the Crick approach. When data from this satellite becomes available there

should be a significant reduction in the cost of imagery to projects like this licencing and data

access, are discussed in section 11.1.

Other EO initiatives: There are many other EO initiatives providing satellite imagery at ever

increasing spatial, spectral resolution and pass rates, such as JAXA and NASA and small

start-ups such as Planet Labs and Skybox imaging which rely on many small cheap

satellites in a large constellation. These should also benefit a National EO service for habitat

mapping by increasing the likelihood of obtaining suitable data at the time it is needed.

UKEOF ‘Efficiencies in Observation’: UKEOF are working to improve environmental

observation activities through encouraging collaboration across UK organisations. Working

together can help avoid duplication, develop more robust networks and improves knowledge

and data accessibility. Natural England have mapped geographic coincidences in

environmental monitoring sites, and it is apparent that there were high levels of geographical

coincidences in monitoring sites in England and Wales. Having consistent regional and

national level maps could help UKEOF partners to better understand the nature of these

sites and assist with efforts to harmonise existing monitoring programmes and other survey

work. A consistent national map would also serve as an important component of a national

framework for sampling.

Space for Smarter Government Programme (SSGP): is a UK Space Agency-led

programme. The ambition is to help the public sector create sustainable operational services

from satellite data and products to enable smarter, more efficient operations, reduce risk and

enhance policy making. A key goal is to make use of existing investment in space, such as

the Copernicus programme and the Sentinel satellites, and allow UK government to become

a first intelligent customer for satellite products and services that could generate economic

growth through export. The outputs, objectives and plan for the delivery of the Crick

approach align well with the objectives of the SSGP because:

they propose cost-effective operational services that use satellite data generated

through existing development (e.g. Copernicus programme) and which improve the

evidence base (providing a more resilient and traceable means of reporting for the

Habitats Directive). The outputs also provide a platform for the cost-effective

development of new evidence for many areas of policy concerning natural resources;

the mapping process involves customers and delivers EO based outputs with the

intention that they be further developed. Involvement is at a variety of levels within



Government – from locally based organisations (such as local nature partnerships),

to government departments and their specialist supporting agencies and advisory

bodies;

the operational solution proposed seeks to put in place a sustainable and more

efficient operational service than the current means employed for Habitats Directive

reporting;

all EU Member States report on Article 17 of the Habitats Directive and face

environmental challenges that would benefit from robust habitat based evidence in

support of policy. The UK is ahead of the field in terms of developing a toolkit of

solutions for comprehensive repeatable habitat mapping. This presents scope for

export of approaches and knowledge.

SSGP are preparing best practice guidance for business case development for space-

enabled operational services8.

Countryside Survey (CS): It has collected a vast resource of survey information and trends

identified by Countryside Survey have long been used to try and answer some of the

challenges, but it was never designed for this purpose. A workshop, held in September 2014

explored the potential benefits of earth observation methods to CS, including those

developed through the Crick approach. A number of potential benefits to CS were identified:

improving understanding of the findings by gathering information on the wider

landscape context of each CS square, to address questions such as “Is this square

typical of its surroundings?”;

looking at cause of high impact change picked up by field survey, using EO and other

data such as agri-environment scheme uptake, to address questions such as “What

pressures and drivers of change are acting on the natural capital of the square?”;

supporting assessment on the ground of physical aspects of habitats, such as %

cover of dominant invasive species, such as gorse, or precise location of habitat

boundaries;

routine EO data capture through programmes such as Copernicus, offers the

possibility of detecting change outside of existing survey windows, if a change is

noticed or an area is deemed to be at high risk then it is a straightforward job to find

the imagery from each pass of the satellite and analysis it to see when the change

took place etc;

provide measures of habitat condition to complement field survey findings – e.g.

measures of degree of wetness and phenology;

using previous CS results along with historical EO data to understand context and

speed of change.

The upland pilot study area included two countryside survey squares for which ultra-high

resolution data from RPAS was acquired and analysed. It was possible to classify the

different habitats from this data including useful features for monitoring such as areas of

bare earth. It was also possible to use RS indices to pull out data on wetness and

productivity on both the square and the surrounding area. Data from the RPAS was

combined with worldview satellite data to give a much better separability for certain habitats

such as different woodland types and wetlands providing initial evidence that some of the

8 http://www.spaceforsmartergovernment.uk/ssgp-focus-2014-15/



potential benefits identified at the workshop can be realised for CS. This has been reported

on in the Phase 3 extension report: WP2 Using EO to map CS squares.

Professional fieldwork: Professional field survey at phase 1, extended phase 1, NVC and

condition monitoring is currently used for the main evidence for a wide range of projects from

site specific to local/regional habitat surveillance and monitoring. There is a wealth of

experience in professional field survey, and it will always form an important part of answering

some of these questions, adding vital information to any habitat mapping. Professional field

surveyors use very well understood methodologies, although there is often intra-surveyor

variation and it is very time consuming to cover large areas with resurvey takes as long as

initial survey.

Citizen science: There are many initiatives to increase awareness, understanding and

enjoyment of biodiversity, and engage more people in conservation. Citizen science is a

growing source of environmental data, and there is a large body of work establishing how far

the data collected can be used in projects. With the potential to provide large amounts of

habitat data, citizen science can serve as part of a validation process, potentially acting to

keep any mapping current. There is also a role for the Crick approach outputs in assisting

with communication and wider accessibility with local people. Looking at habitat

assessments and recording additional data could form an important component for schemes

that seek to link people to nature. Norfolk Biodiversity Centre and partner organisations are

using the habitat map produced using the Crick approach to look for different habitats and

species. This approach is helping to validate the map and add rich additional data to it, using

a citizen science approach.

5 Why scale up and roll-out the approach across the UK?

5.1 Why scale up?

Scaling up will lead to consistent data being available at nested levels. This is not

currently the case - different habitat mapping approaches exist which cause many

integration issues; drawing together and making comparisons between existing data

can be very difficult. Without action this situation will persist. In any scale up

operation consistency needs to be addressed through guidance, mapping to common

habitat classifications and definitions which necessitates a delivery framework to

achieve consistency in outputs and allow data auditing.

With consistent habitat knowledge better decisions can be made to manage

biodiversity, and better monitoring can be carried out to understand and respond to

trends in species and habitats.

Data can be compatible across administrative boundaries and so can be used for

the management unit required. Data can also therefore be used for Local Nature

Partnerships, Nature Improvement Areas, catchment planning, landscape

assessments and for commitments to EU reporting.

A suitable framework will be available for use by local biodiversity groups, giving

information about habitats across an area, that is more up to date than much county

recording, and which can act as a basis for feeding in extra local data.

At a county or regional level, systematic data provides a useful framework for

strategic planning, at a scale suitable for feeding into ecosystem approaches for



resource management, and provides quantifiable data on the extent of habitat for

local biodiversity targets and plans.

Uncertainties revolve around provision of suitable imagery, consistency and integration of

data, management of the time periods to allow the data to feed into national reporting, and

management of data update. If suitable imagery is not available within the window of

mapping, areas will be left out of the mapping cycle or out of sync, with little or old

information. Rigorous management of the projects and maintenance of data standards is

required to maintain data quality otherwise there would be too much variation.

If a programme of roll-out is not followed up, there is still likely to be a series of different

habitat mapping approaches going ahead. This will essentially mean that the status quo

persists where habitat data is highly variable in terms of scale, coverage and habitat system.

Whenever data are required at a national level, the data needs re-working and modelling into

an approximation of the UK habitat situation, having good information in some areas and no

information available in others.

5.2 Why begin roll-out of the Crick approach now?

A ‘preparation phase’ which would set up documentation, data production standards and a

system to address data licencing and allow access to consistent data is required as a first

step before systematic roll-out can occur. Additionally further testing of the update process is

required as scaling up changes the nature of delivery, in areas where good (but out of date)

data exists. That withstanding:

demand already exists for the data produced by the techniques (e.g. for Habitat

Directive reporting);

outputs and data from the Crick approach are being adopted locally (Table 3) as the

data provided can be used to generate new outputs that support local policy needs;

the techniques are already proven to support identification of these habitats;

use of the technique in Norfolk, Wales and upland environments in the UK

demonstrates that the techniques are capable of consistent implementation;

we are missing opportunities to protect and enhance the natural environment by

our lack of systematic detailed mapping;

to address recognised weaknesses and gaps in knowledge about the status of

Annex I and Priority Habitats;

to maximise the return by additional reuse on substantial Government investment in

geographic data, including national aerial photographic cover, LiDAR, LPIS, Digital

Surface Models, Satellite data through EU contributions to Copernicus;

to benefit from low cost imagery, suitable for this type of habitat mapping, soon to

be provided by the Copernicus programme.

5.3 When would benefits be realised?

Table 7 demonstrates the likely sequence in which outputs and benefits will be realisable to

users. For many users, the value of the data then increases year on year if the data is

actively managed, kept updated and used to produce further data on the environment. Full

potential of the data for monitoring requirements may take longer to be realised (such as for

condition monitoring). Having a baseline of the extent of habitats allows changes to be

recorded. Some benefits will start to accrue from an early stage as evidenced by uptake of



the products (Table 3). There would also be a range of other advantages realised at an early

stage, including improving site knowledge (locally) and public understanding of science.

Table 7: Sequence in which Crick approach outputs / benefits will be made available

Benefits realised by local

stakeholders

When benefits

would begin to

be realised?

Benefits realised by national stakeholders

In selected areas throughout the UK:

systematically mapped products available;

additional fieldwork and checking (locally) of content maintains and enhances accuracy and completeness of the map;

Local Records Centres facilitate reuse of the Crick products locally and produce data and maps that address local policy needs

An updated habitat map is made available for validation and use by local stakeholders in selected areas (Wales and other areas with good starting data).

Immediate – 3 years (with

value and effectiveness

increasing over time)

National data hub in place providing:

consistent data standards (processing of input imagery and potentially for high-level supporting data products);

opportunity for licence negotiation Priority areas for initial mapping identified.

Good UK wide understanding of the range of start-points for the mapping required at the local level.

Potential new measures of condition are developed and tested. Methods disseminated to local partners.

Additional areas continue to be systematically mapped bringing the benefits described above until initial mapping is completed (in 7 years).

Condition measures available for use locally.

3 – 7 years (with value and effectiveness

increasing over time)

National Records Centres able to make data from selected areas available to national stakeholders groups such as conservation NGOs.

Condition measures and other analyses can be carried out on nationally validated data.

4 years

Annex I reporting in 2019 is informed by combined validated baseline data available from selected local partnerships and up to date data from Wales.

9 years

Annex I reporting in 2025 is informed by a full set of validated initial maps. Selected areas will have been updated.

Partnerships in all parts of the UK have access to up to date habitat map for use in local decision making.

13 years Annex I reporting in 2031 is informed by current habitat maps (within 3 years of reporting year)



A NATIONAL EO SERVICE FOR HABITAT MAPPING

This part of the Business Case describes how a National EO Service for habitat mapping

based around the Crick approach could be delivered and discusses the practicalities

involved. It is based on a high level assessment of available information and includes an

initial assessment of risks and dependencies. Initial costs and the assumptions upon which

they are based are presented. This is followed by evaluation of the cost-effectiveness and

value for money of the Crick approach as the means for delivering of habitat data for

reporting on Annex I habitats.

The outputs of the Crick approach must add value and/or deliver efficiencies to enhance the

“toolkit” and be a practical proposition for organisations to complement existing techniques.

The approach described is considered fit for purpose for generating data for Habitats

Directive reporting, amongst other uses, and assumes mapping is carried out for the whole

landscape with more detailed mapping carried out in areas with more complex, more

intricate habitats.

6 What will a national habitat assessment service

comprise and how might it be delivered?

A national service for habitat mapping needs data, systems for analysing the data and

people to run the analysis, have custody of the data and use it.

Data

availability of a range of datasets to support the assessment of habitats;

EO data at a variety of scales to give suitable spatial and temporal resolution for

looking at habitats;

other supporting datasets to provide context and a consistent spatial framework;

data hub to provide consistent management of the national datasets, image archives,

high quality processing of the imagery and easy access to the data.

Systems

an Object Based Image Analysis (OBIA) approach with suitable software and

hardware for processing imagery and running image analyses;

a mechanism for compiling local ecological knowledge;

EO / geoinformatic systems for data manipulation, cleaning and update.

People

national guidance for standard setting and data auditing;

technical remote sensing and ecological skills;

data management skills;

local ecological knowledge;

local partnerships to manage and distribute the data locally.



6.1 Data components

For the best possible outputs, a range of suitable data is required to support identification of

habitats including imagery, digital terrain models (DTMs), soil/geology and agricultural

system data (i.e. a Land Parcel Information System). A summary of these and their role in

the mapping process is provided in Table 8.

Table 8: Imagery and data requirements

Feature to be identified Imagery / data need Current data access situation

Rectification of imagery, splitting the landscape into discrete areas and the identification of some features (e.g. valley bottoms, steep slopes and north facing aspect slopes)

High resolution DTM (5m) available consistently across the whole study area

Available for GB within APGB contract

Identifying rivers and urban / built areas

OS MasterMap Available under PSMA and One Scotland agreements, but does have implications for dissemination of products

Identification of individual land parcels – agricultural fields.

Land Parcel Information System (LPIS) is the ideal solution. MasterMap can be used but requires additional analysis to ‘close’ polygons open because of gates. MasterMap is better in some areas than others for this.

Classification of relatively uniform areas of habitat which are spectrally distinct above a minimum mapping unit of 5 image pixels.

SPOT or similar resolution imagery (10m multispectral imagery). Likely to be able to use Sentinel-2 data when it becomes available.

SPOT available commercially. Sentinel-2 data will be free at the point of use once operational.

Identification of small and intricate or narrow vegetation features e.g. hedgerows, small patches of scrub classification of (e.g. creeks)

VHR satellite data, (~ <5m) e.g. RapidEye or Worldview data or CIR aerial photography.

RapidEye and WorldView available commercially. CIR photography is available under APGB contract

Extra information to identify features determined by cyclical changes to vegetation (e.g. to distinguish permanent grassland from arable land)

Broad scale high temporal frequency data e.g., Landsat 7/8 Imagery

Landsat data is freely available.

QA and checking of rule base development during the mapping process

RGB aerial photography CIR photography is available under APGB contract

Identification of features characterised by their specific structural characteristics (e.g. reed beds)

Leaf on LiDAR LiDAR is collected by EA and SEPA, but coverage is patchy.

Soil and geology data for identifying features with a strong relationship to their soil type (e.g. calcareous grasslands)

1:25,000 or 1:10,000 soil & geological data

Geological data at 1:250,000 available under Defra license. Soils data only available under commercial license from Cranfield University

Local habitat data Pre-existing local habitat data consistently collected.

Variable coverage



This data is held by various central government and private organisations. Licensing of

commercial data such as the OS MasterMap, soils and commercial imagery needs to be

established at an early stage in planning, as licensing issues are known to have impacted on

regional and national mapping projects in the past. Licensing issues encountered by

LCM2007 resulted in substantial additional costs to the delivery of the project and onward

exploitation (Haines-Young et al. 2012).

The currency of any mapping is reliant on the date of the imagery available. Therefore, as

part of a rolling programme, any national mapping data products will contain a range of

dates, with each region being completed with imagery at a specific date.

RPAS data is very useful at the site level but it is not at a suitable scale for being used for

the national scale habitat assessment, therefore has not been considered for the National

Scale EO Service roll out. A network of service providers could be used to deliver site

assessments nationally.

6.1.1 Plans for the supply of imagery in cloud affected areas

All areas of the UK are prone to cloud cover. In areas which are regularly affected by cloud

cover, much less usable satellite imagery will be available from standard image capture

programmes. Tasking of imagery can be requested where the satellites are ‘turned-on’ for or

pointed toward target specific areas, giving greater chance of capturing clear imagery.

Upland and western areas in particular are the most likely to require image tasking. This has

been taken account of in the costing exercise (Section 9). There may be merit in establishing

the potential of radar data such as Sentinel 1, for areas affected by cloud cover. This

imagery has a coarser spatial resolution than the optical data used in the Crick approach

and the implications of this would need to be assessed.

6.1.2 Aerial photography

Aerial photographic (AP) data is essential for the Crick technique for finer scaled habitat and

for ecological interpretation. Both RGB and NIR data are needed and it is particularly

important for the segmentation process that identifies woodland and hedgerows. There are,

however, certain features of large scale aerial data capture that add to the complexity and

cost of mapping. Most of these considerations stem from the fact that large scale AP capture

is subject to significant variation arising from:

variation in solar illumination conditions and changes in sensor present limitations.

variables presented by the growing season / timing of seasonal changes in the

growth of vegetation

Terrain, which presents a limitation across the study sites because mountainous and

textured terrain produces significant shadow in winter imagery due to the low sun

angle. It is therefore best practice not to use imagery captured from mid-October to

early March.

leaf-on / leaf-off imagery poses the greatest cost implications, it is necessary to have

consistent leaf on imagery to get good data for woodland and hedgerow

segmentations.

occasionally imagery varies beyond the capacity of colour balancing or smoothing

and therefore prevents the application of a consistent rule-base, this again increases

costs.



The current Defra funded APGB (Aerial Photography for Great Britain) contract runs for the

next three years and sets a new standard for the capture of geographic data in Great Britain,

with a three year update programme rather than the previous five year cycle. It covers the

collection of high resolution aerial photography, detailed 3D height models and Colour Infra-

Red imagery for the whole of England, Wales and Scotland.

6.1.3 Data provision hub

For consistency of data inputs and reducing the time taken to access data from multiple

sources, there is a need for a data repository or system to enable access to the nationally

held imagery and ancillary data needed for a rollout of the Crick approach. The Satellite

Applications Catapult, an independent technology and innovation company, will play a key

role delivering the Space for Smarter Government Programme (SSGP) and is establishing

potential storage of sentinel data. A SSGP project is currently identifying and assessing

options and logistics of a hub and spoke type model for satellite imagery data management.

6.2 System components

Scale: The Crick approach works at a landscape level, typically the scale of a county,

operational catchment or several Landscape Character Areas. In areas larger than this, it is

not possible to capture local knowledge effectively. An analysis to identify suitable areas is

underway for all countries based on known active nature partnerships or areas where these

might be facilitated. This is further discussed in section 11.1. By way of example, there are

88 counties in England and initial analysis suggests that in England, around 33 regional

projects (dependent upon image swath) would be needed to deliver national mapping. Each

of these ‘regions’ is likely to comprise several Crick ‘projects’ determined by imagery

footprints together with the bio-geography of that region. At present, these projects are

restricted to 60km x 60km SPOT footprints and the RapidEye 77km swath width currently

used as the main sources of imagery, but the larger swaths from the Copernicus satellites

(290km swath width) will enable larger areas to be covered, allowing biogeographic areas to

be the main drivers of projects; at present the image scene has a major influence on choice

of project area.

Software: There are a number of software packages for image processing and for running

OBIA analyses. Currently the most powerful for the OBIA methodology uses eCognition

software. There are some open source software alternatives which can be used together

with other proprietary software, which could support the process particularly for updating and

monitoring of existing ‘objects’ which may provide alternatives and/or cost savings. These

should be kept under evaluation but are not currently suitable and considered medium term

options.

The eCognition software is particularly advantageous as the interface is clear, enabling

ecologists with some remote sensing background to use it and to set rules. It also allows

multiple scales and types of segmentation meaning distinct objects in the image can be

delineated that accurately reflect the different size, appearance and manifestation of the

environmental features on the ground (from long hedgerows to large areas of blanket bog).

This software needs high capacity computer systems to run the ‘engines’ on the complex

calculations and analysis undertaken. It can take 10-24 hours of elapsed computer time to

calculate the objects for a Crick project on a very fast and highly specified work station.



6.3 Mapping Process

The flow of the mapping process is illustrated in Figure 2 and the distinct stages which have been been used for costing are shown in



Table 9.

Figure 2: Flow chart of the process of mapping using the Crick approach, orange lines show the alternative pathway for areas with existing good mapping requiring update



Table 9: Required stages of the Crick approach

Stage Requirements

Stage 1: Data acquisition

Image searching, ordering, and downloading. Search criteria include how many different scenes will be needed in terms both of area and seasonality of semi-natural habitats and main agricultural systems. (In the ideal situation scenes within the same year should be chosen, if scenes across years are necessary an analysis must be made to ensure that change will not be a major issue). For example in wales a 3 year window was tolerated, Not all imagery needs to be of the highest spatial resolution. If Landsat scenes are useful for fully describing land cover within the arable cycle, which is highly variable. A SPOT/RE/equivalent sensor (Sentinel 2), with both a spring and summer image for the whole area is an essential requirement.

Sourcing and ordering ancillary data sets including DTMs, existing survey information, field boundary datasets, roads and rivers etc.

Stage 2: Image processing

Image correction to ready-to-use status

Each scene needs processed using a high quality DTM. This must be accurate enough for ortho-registration of imagery from different sensors and different times of year to exactly overlay.

Atmospheric, topographic and radiometric correction needs completing and finally mosaicking, colour balancing, re-sampling, sub-setting by date

Stage 3: Ancillary data preparation

The ancillary data needs tiling, cleaning, sub setting and any data extraction.

Stage 4: Familiarisation fieldwork

Remote sensing and ecological specialist need to take the imagery into the field to check the manifestation of the habitats in the actual images used.

Stage 5: Rule base development and implementation

The approach relies on iterations of the rule base: Iteration 1 includes incorporation of any data used in the segmentation and initial landscape splitting. There then follows a stage with internal validation and reiterating until there is a first draft or a first iteration of the mapping.

Stage 6: Testing fieldwork

To establish fitness-for-purpose local partners should take the initial map out into the field and verify the habitats, any that are not being described well should be described by target notes so that the rule base can be modified to better select them.

Stage 7: Product generation

Use of Geoinformatics to lump/split RS classes into required classification system.

Stage 8: Gap filling through fieldwork to get to Annex I habitats

Fieldwork to get beyond what EO can do to bridge the gap to national reporting needs for the presence and extent of habitats.

Stage 9: Management

Project management activities (throughout delivery)

Stage 10: Software and hardware requirements

Most software license costs would rest with whoever undertook the mapping and be influenced by how many licences they currently hold. If a delivery organisation sought to use an alternative to eCognition (which is technically feasible) it should be recognised that this would be considerably more difficult and more time consuming.

Stage 11: Documentation

Supporting documentation to provide the framework and guidance and data quality assurance across the UK (mainly delivered early in the process).



6.4 People (delivery partners)

Ecological and remote sensing skills within the delivery team are a prerequisite. The stages

at which the main stakeholder organisations are considered likely to be involved in delivery

are summarised in Table 10, based upon the experience of the pilot projects and mapping in

Wales.

Table 10: Potential involvement of key delivery stakeholders*

Stage

Delivery stakeholder group

Central (Defra/JNCC)

Country Delivery

Agencies

Consultancy Local (e.g. LNPs,

AONB, WTs, LRCs)

1: Data acquisition

National base data collation

Satellite imagery

Local data

2: Image processing **

Satellite imagery processing

3: Ancillary data preparation

Centrally held data

Locally held data

Creation of masks

4: Familiarisation fieldwork

5: Rule base development & implementation

Rule base development

Field checking

6: Testing fieldwork

7: Product generation

8: Gap filling through fieldwork to get to Annex I habitats

9: Management

10: Software and hardware requirements

11: Documentation

Data currency

Data standards *purple denotes involvement of delivery partners in the main stages of delivery; blue denotes involvement in sub-stages.

**Image processing could also be undertaken by a data hub

In the scenario in Table 10 it is assumed that DEFRA and JNCC and the Country Delivery

agencies will be involved as key stakeholders and data providers, they will have an overall

role in supporting the development as it progresses and an interest in the results and setting

the standards. The local partner will have a key role in supplying local data, expertise and

knowledge and in providing field work to support to the process.

Local Nature Partnerships in England (Figure 3) which (for comparison with other countries)

are typically at a county scale, are considered likely to be the ideal size for delivering Crick

projects in a national rollout.

The Local Nature Partnerships (or similar bodies, with a similar remit in other parts of the

UK) could have a large role in the National EO service, providing field expertise, ecological

knowledge and field checking together with involvement in validation of the mapped outputs.

They will continue to add habitat information once the map is produced through fieldwork by



local volunteers or through other activities. There is a wealth of experience in professional

field survey. Field survey is critical to accurately identify habitats defined by small scale

infrequently occurring indicator species and it is currently the only way to determine specific

subsurface features e.g. peat depth.

It is envisaged that Local Records Centres / National Record Initiatives e.g. NBN could be

responsible for copies of the habitat map and will disseminate information about its

availability and potential for use to support policy at the local level. Responsibility for co-

ordination at a local level would come from the Local Nature Partnership or leading partner

group. This could vary depending on local setups.

National agencies such as GI departments of government agencies have the expertise to

contribute to setting standards, overall co-ordination of metadata, presentation of colours in

maps, ensuring consistency with minimum mapping units etc. as well as providing external

QA. They are also required to ensure consistent data standards, both for the processing of

input imagery and a potential common high-level data product would be a role for the

national co-ordination team. Potentially they could also be responsible for co-ordinating the

purchase of imagery (to ensure VFM is sought). It is also necessary to track the progress of

mapping, making available the necessary data in sufficient time for national reporting

deadlines.



Figure 3: Local Nature Partnerships



7 Delivery model

Four possible delivery models have been identified:

1. Top down: Public sector led mapping of the whole of the UK in structured and regular

manner, i.e., one area after the other until everywhere is covered with central

coordination of production.

2. Bottom-up: Driven and produced by demand from local/regional bodies.

3. Data led: The programme of mapping would be driven by image or data availability,

starting with those areas for which data is available at best cost.

4. Combined approach: In the combined approach the project would be likely to include

local areas keen to be involved where data was readily available with the Public

Sector providing overall project control including setting and ensuring standards

These have been subject to a SWOT analysis to help determine a preferred model of

delivery (Table 11).

Table 11: SWOT analysis of the delivery models

Strengths Weaknesses Opportunities Threats

Top down:

Consistent approach Limited local input, therefore not picking up scale and variability

Can prioritise mapping of areas to meet known policy needs (e.g. wider countryside)

Delays if staff are diverted onto other more urgent tasks.

Little local adoption

Bottom-up:

Allows early adopters to lead the way.

As data is available work can begin.

Difficult to direct to areas of need if no interest or capacity to respond locally.

Early adopters participate in Knowledge Exchange and demonstrate best practice and local uses

Could result in patchy coverage that is difficult to stitch together

Data led:

Target delivery areas based on image and other data availability.

Local area might not have a suitable partnership to be ready to be involved.

Efficient for co-ordinating image tasking in areas left behind after the areas with existing good imagery are complete.

Could be economical if it fitted well with plans for improved LPIS datasets

Local areas that are keen to be involved feel overlooked

Combined approach:

Able to strike balance between data availability, user needs and opportunities

Requires more planning and preparation at outset than other approaches

Offer a flexibility to build in opportunities to a longer term plan as they arise

Poor management could lead to missed opportunities or an overly complex and time consuming programme

A combined approach that is primarily data-led is considered the most viable way to deliver

an ongoing programme of national rollout because it can best take advantage of available

and timely imagery, or areas of the country where LPIS data is available, can respond to

local demand, target areas where government need to address gaps in knowledge whilst

ensuring consistent standards.



The consistency of the approach to mapping and extra field data recording for map update

could be driven centrally. The data will therefore be relevant for both local and central

analysis and modelling.

There can be flexibility in the approach to ensure it is practical. For example, local

engagement enables the best product possible to be produced, but also brings added value

as organisations working at this level will be key users of the products. However, if demand

is low in some areas, or there simply is not the capacity locally to participate, then

responsibility for in-fill could be taken up by an overseeing body (e.g. public sector supplies

specialist skills (e.g. GI or ecology support skills) through commercial contracts or by

supplying staff from the relevant government agencies. This was how the map in Wales was

produced.

7.1 External Dependencies

Provision of skills: There need to be sufficient staff available with remote sensing and

ecological skills within the delivery teams to meet any set delivery schedule. The availability

of local ecological support is highly desirable but it would be possible to map some areas

without this provided there is sufficient capacity in the national delivery teams.

Continuity of supply of suitable imagery at anticipated costs: Analysis of the planned

lifecycles of existing and proposed satellites, through to 2025. Defra have recently let a new

contact for acquisition on a 3 year cycle for the UK for RGB and CIR aerial photography.

Contractual arrangements: Some forms of agreement are needed to secure engagement of

those involved with delivery (with the exception of volunteers). This will be an ongoing

dependency of the project as it is intended that there be a rolling programme of delivery.

One option that could be considered for securing and formalising engagement at the local

level might be to periodically invite participation through ‘calls of interest’ such as those used

to help establish Local Nature Partnerships.

The mapping involves discrete multi-stage tasks and there will be a range of internal

dependencies related to these that will need to be identified as the business case is further

developed.

7.2 Risks and Issues

Risks have been reviewed and mitigation measures proposed and an initial risk register is

provided at Appendix 2. As business planning proceeds many of these risks can be actively

managed and reduced in terms of their likelihood of occurring and their severity. The main

risks are:

Lack of commitment by funders, national or local delivery partners;

Unforeseen costs for licenses or imagery (e.g. if Copernicus data is not available);

Getting the central standards and system in place – costs could rise if the central hub for delivery of Copernicus data is not in place when scheduled; and,

Loss of key skilled staff to the ‘habitat hub’ – cause delays – training / capacity building.



8 Business Plan

A business plan for roll-out of a national EO service for habitat mapping, particularly one

employing a delivery approach that seeks to combine elements of a data-driven, centrally

and locally led models will require a significant element of forward planning in a preparation

phase. This should provide the framework for taking the national EO service for habitat

mapping forward in a sound, transparent and considered way through a partnership

approach and with an appreciation of the objectives and long-term nature of commitments.

This will provide partners with the information they need to progress discussions and help

them reach agreements towards implementing the service. It would set out plans for all

aspects of implementation and consider how the service could be further developed in the

longer term. It would provide sufficient information to proceed to a technical specification for

implementation and takes account of immediate and longer-term needs. There needs to be

assessment and forward planning, for example, covering:

the capacity of interested parties to deliver against a preferred schedule for rollout.

Where are the skills currently? Which organisations have capacity to deliver? What

input do they need to provide? How will the products be integrated into

organisations?

planning for reporting / planning for consistency of outputs, within individual mapping

projects, between project areas and for the development of the ‘living map’

integrating within a hub and spoke model / assignment of responsibilities

post mapping data management

planning for local engagement

9 Assumptions and presentation of costs

The estimated unit costs (£/km2) of implementation of the Crick approach at a national scale

were derived from analysis of costs incurred in a pilot project in Norfolk during Phase 2 of

the project, accounting for scale and landscape differences. The Norfolk pilot project

covered:

production of a complete coverage habitat map for the county (referred to as : during the mapping process areas known to contain substantial areas of semi-natural vegetation were mapped at a finer scale (this applied to about 7% of Norfolk);

time spent by local NBIS staff supporting the mapping process and field checking areas identified as of potentially high nature conservation status after the map was produced to assign actual habitat status.

The cost of acquiring and processing data from RPAS in Norfolk was excluded as data

acquisition from RPAS is not proposed as part of national rollout. Arable areas were not

mapped to crop type.

The self-funded contributions from central bodies, such as Defra, JNCC and from individual

members of the Steering Group are not included. Local organisational overheads are also

not included as these are highly variable, depending on the organisations involved (these

would increase the value of voluntary contributions to the project). Staff costs are based on

day rates associated with the contractual arrangements.

Costs of the pilot project were categorised using an eleven stage process of work activities (see



Table 9) each with sub-stages against which distinct items of cost / elements of delivery

were identified. These costs were assigned to:

capital costs (for the purposes of this exercise this includes software, hardware and

data costs)

staff costs (project team time for delivering the mapping)

in kind contributions (time provided to the project by NBIS staff and volunteers)

The cost of updating the map is an estimate derived through expert review of each element

of the known costs of initial mapping, taking into account existing experience of updating

maps using the Crick approach gained in recent projects.

To assist with estimating the costs of national implementation and future map updates and

identifying where efficiencies from scaling up are expected, the costs were further

categorised as:

ongoing or one-off costs; and,

linear or non-linear depending on whether or not they increased proportionally when scaling up the process

The average unit costs of initial mapping and estimates of the cost of updating using the

Crick approach in Norfolk are provided in Table 12.

Table 12: Average cost of mapping a unit area (km)

Cost category

Cost per unit area (Norfolk pilot area)

Initial mapping (actual)

(£ / km2)

Update of the initial map (estimate) (£ / km

2)

Capital costs 4.40 2.30

Staff costs 23.70 8.70

In-kind costs 8.50 3.00

Total £36.60 £14.00

Staff costs of the core (central) delivery team comprise about two-thirds (65%) of the total

cost during initial mapping and map update.

In Norfolk, costs of time NBIS staff and volunteer time at the local level comprised a

substantial contribution making up nearly a quarter (23% and 21% for initial mapping and

update respectively) of the cost of delivery.

It is estimated that updating the initial map will cost approximately 40% of the cost of

production of the initial map.



9.1 Scaling up of costs

The Crick approach is relatively flexible; the mapping process and resource capabilities can

be adapted to take account of differing circumstances locally and variability in levels of

support available nationally (e.g. via Copernicus). This flexibility of delivery is set within the

context of mapping habitats across very variable environments within the UK. Thus rollout

presents a complex process encompassing variation in product specification, with technical,

environmental and practical challenges. Equally, rollout also presents opportunities to deliver

economies of scale and make best use of local and national resources.

To understand and accommodate this variation, a modelling approach has been developed

to incorporate experience from previous implementations of the approach and existing

national land cover mapping exercises to describe and document the uncertainties,

economies of scale, assumptions and impacts on cost of a national roll-out.

The modelling approach was broken down into three elements:

the mapping process;

landscape type; and

regional issues.

9.1.1 The mapping process

The mapping process costs are those related to the technical specification and implementation of the implementation of the approach across the 11 stages described in



Table 9. An estimate of the minimum and maximum cost for each stage was arrived at.

Each item of known cost (from the Norfolk pilot and other costs that would be required for

scale up) was subject to expert review within the project team. Known factors, uncertainty

and assumptions considered likely to cause variation in the cost of delivery due to the actual

processes involved in implementing the Crick approach when scaled up for a national scale

mapping service rollout were documented. During this process every effort was made to

exclude the effects of working in differing environments although it is recognised it will have

some bearing on the variation of some costs (for example, in some areas it will be necessary

to produce landscape masks to separate the calcareous rock from the acidic rocks, this

might be a time consuming process if no detailed geology map is available).

The minimum, average and maximum cost has been estimated for each sub-stage of

mapping based on the team’s experience; these reflect the staff time and capital cost

associated with the most efficient, most likely and most challenging situations that might be

feasibly be encountered. For the imagery costs, the minimum cost assumed all data could

be obtained from existing data agreements, including the Public Sector Mapping Agreement

(PSMA), Aerial Photography for Great Britain (APGB) contract and Copernicus programme,

and there would be no additional commercial data required. The maximum cost assumes

that all satellite data would need to be purchased from commercial providers. The costs

associated with each sub-stages are mostly independent of one another. As a result a

mapping exercise may incur a mixture of minimum, average and maximum costs depending

upon the particular circumstances that arise.

The inter-relationships of factors and wide range of assumptions applied confers a high level

of uncertainty in making an assessment of how the variation in input costs will affect the

overall cost of mapping at the national scale. Simulations are a preferred method of data

analysis in cases where the variability in parameters is an important factor in determining the

magnitude of predicted cost. A Monte Carlo simulation was applied to provide statistical

assessment of the likely variation in unit costs of implementing the Crick approach nationally

based on current knowledge.

The model outputs provide an average estimate and range in the cost of rollout with 95%

certainty. This per unit cost translates to an aggregated range of costs for Norfolk of

£170,000 to £214,000, with mean value of circa £192,000. This translates to a cost per

kilometre of approximately £36 with a range of between £32 and £40 (with 95% certainty) as

shown in Figure 4. Given the uncertainties at this stage about the input costs that might be

incurred, we have made assumptions about the distribution of costs - that is to say the

likelihood of the averages for different sub-stage falling somewhere between the respective

maximum and minimum cost. This has involved calculating a mode for each pair of

minimum-maximum costs and choosing a triangular or PERT distribution which assigns less

weight to maximum and minimum values compared to the mode.



Figure 4: Histogram of distribution of predicted per unit cost

The modelling process highlights the main stages of the mapping process that are

contributing to uncertainty, and therefore variability, in the predicted the costs (Figure 5).

This sensitivity analysis suggests that stages 10, 5 and 1 are highly correlated with the

output, meaning that these have the greatest effect on the magnitude of predicted cost.

Changes in software and hardware requirements affect the predicted cost the most, whereas

the effect of gap filling fieldwork on the predicted cost is marginal.

Figure 5: Tornado plot showing sources and extent of effect of the mapping stages on the predicted aggregate costs for Norfolk

This is very useful as it points to the areas where further work can be targeted to better

understand the likely costs of implementation when scaling up nationally, which should lead

to more certainty and will help strengthen the business case going forward, de-risking the

process.



The stages in the Crick approach that account for the main variability in costs of the delivery

process in order of influence are:

Stage 10: Software and hardware requirements – the number of EO analysis and

processing software licences (eCognition, ENVI etc.) required is the main cause of

variation. If all the software needed had to be purchased at full commercial rates it

could significantly influence the cost. Normally this cost would be shouldered by the

delivery partner for different analysis projects, but has been included as a cost

consideration as it is potentially a factor that needs to be considered further.

Stage 5: Rule base development and implementation – the variability in this cost

is influenced by time required to set up the rule base and the level of experience and

staff knowledge of the area to be analysed.

Stage 1: Data acquisition – the availability or otherwise of data from the Copernicus

programme is the main source of variability in data acquisition costs, the existence of

a national hub for data sourcing and the ability of local records centres or country

conservation agencies to supply supporting data could also impact on the overall

uncertainty.

There was insufficient time available to conduct a Monte Carlo simulation for the cost of map

updating, but such an analysis is recommended for the future for the reasons already

provided.

9.1.2 Landscape type and regional differences

Cost variation arising from differing landscape type is related to the numbers of habitats and

land cover types and their arrangement; these factors have been considered in a separate

exercise from the analysis of variation in the cost of the mapping process, and used as

scaling factors for arriving at national costs.

In general, lowland intensively managed and urban landscapes are more straightforward to

map as the land cover types are usually arranged in regular homogenous blocks with

relatively distinct boundaries between units. There also tends to be more ancillary data

available in these areas. Conversely, the unenclosed uplands may contain many subtly

different and highly heterogeneous habitats which blend into each other across ecotones

rather than hard boundaries.

The project team has drawn on their experience of the UK landscape and the ease with

which EO-based products can be derived in differing environments to produce a simplified

model to allow initial assessment of the likely variation in costs that might arise from working

in the differing bio-geographical areas of the UK. This has produced the scaling factors in

Table 13.

Firstly, three categories were adopted based on the Environmental Zones used in

Countryside Survey which were aggregated to give the following landscape types:

Arable-dominated lowlands and urban

Marginal pasture-dominated

Uplands

From experience of UK-wide and European mapping activities it is known that the acquisition

of imagery data can be challenging in the northern and western areas of Europe.



The regional factors are cloud cover, solar illumination, length of season, image availability,

user engagement, ancillary data availability etc. Within the UK the regional gradient is

generally in a southeast to northwest direction and can be divided into three general

categories:

South eastern part of the British Isles (England south of line from Severn to Wash)

Central part of the British Isles (Wales plus England north of line from Severn to

Wash)

North western part of the British Isles (Northern Ireland and Scotland)

Alpine areas (parts of Scotland)

By scoring each of the landscape and regional zones relative to a base situation, which

represents the ‘lowest cost’ of delivery (in this case the arable-dominated lowlands and

urban of the south eastern part of the British Isles) a table of scaling factors can be

produced. The scaling factors attempt to encompass the difficulty of working in the uplands

as opposed to lowlands and the change from arable dominated to upland dominated

communities (Table 13). So, it is thought that working in the uplands of the North western

areas of the British Isles will cost approximately 1.4 times more to successfully map, than the

arable lowlands.

Table 13: Scaling factors for consideration when costing mapping of different zones of the UK

Scaling factors

Regions

South eastern

Central North western

Alpine

1.0 1.1 1.2 1.3

Landscape types

Arable-dominated lowlands 1.0 1.0 1.1 1.2 n/a

Marginal pasture-dominated 1.1 1.1 1.2 1.3 n/a

Uplands 1.2 1.2 1.3 1.4 1.6

Clearly this has implications for the relative costs of mapping the four countries, with mapping of Scotland likely to incur higher costs per unit area mapped than the other countries of the UK.

9.1.3 Model implementation

The cost ranges and scaling factors were then used to estimate values for the UK and the

four separate countries (Table 14 and Table 15) for both initial base mapping and updating.

This was carried out by calculating the proportion of each landscape type within each of the

regional zones within each country. The landscape types were estimated from the 1 km grid

data for the Environmental Zones held within the Countryside Information System. These

proportions were then multiplied by the appropriate scaling factors, the total land area and

the unit cost being considered.



For initial mapping the lower and upper cost limits (95% certainty limits) were informed by

the Monte Carlo simulation. As no Monte Carlo simulation was done for map updating, the

absolute minimum and maximum costs were used, resulting in greater overall level of

variation in the costs presented.

Table 14: Projected cost per country for initial mapping

Cost (£ million)

Lower* Average Upper*

Unit cost (to which

scaling was applied) £32.00 / km

2 £36.60 / km

2 £40.40 / km

2

England 4.7 5.4 6.0

Wales 0.9 1.0 1.1

Scotland 3.7 4.2 4.6

Northern Ireland 0.7 0.8 0.9

United Kingdom 10.0 11.4 12.6

*cost is that calculated by the Monte Carlo simulation, with the 95% certainty lower and upper limits applied

Wales has already been mapped using an OBIA approach and an update of this map is

likely. The cost of initial mapping of Wales is included in Table 14 for illustrative purposes as

there is no intention or need to reproduce the initial mapping.

Table 15: Projected cost per country for updating the mapping

Cost (£ million)

Lower Average Upper

Unit cost (to which

scaling was applied) 7.50 / km

2 14.00 / km

2 18.40 / km

2

England 1.1 2.1 2.7

Wales 0.2 0.4 0.5

Scotland 0.9 1.6 2.1

Northern Ireland 0.2 0.3 0.4

United Kingdom 2.4 4.4 5.7

The work and modelling approach represent a first iteration through the business case

costing for a national roll-out of the Crick approach and provide a viable and evidence-based

approach on which to base a more in-depth analysis. If Monte Carlo simulations are

produced for updating the costs associated with updating the map the 95% certainty values

can be used to provide projected costs for updating the map; this will reduce the overall

variation in costs.

It is appreciated that there are likely to be quite large uncertainties in some of the estimates

of upper and lower costs derived from the mapping in Norfolk and in the scaling factors. It is

suggested that further consideration is given to the categories used for the landscape and

regional types and their consequent individual and combined scaling factors. As a first step,

the costs of implementation of the Crick approach in the full range of areas in the UK in

which it has been applied could be used to calibrate / validate the scaling factors.



9.2 Further analysis of factors influencing the cost of rollout and

their effects

Implementing rollout of a National EO service for habitat mapping will bring cost savings

through economies of scale but there will also be additional costs that arise because as the

process scales up, structures and procedures need to be put in place to support and

facilitate delivery and ensure consistency of the outputs.

An analysis of factors likely to have significant impacts on mapping costs as a result of the

process of rollout across the stages of delivery identifies the savings and additional costs

likely to accrue. During this analysis the impact of the planned Copernicus satellite delivering

imagery free of cost to the project has been taken into account. However, the cost of setting

up structures such as a national and / or habitat data hub, have not, as the practicalities of

this are being addressed in other projects (see section 6.1.3). These will be significant costs,

albeit not necessarily incurred directly by the project itself. They will need to be taken into

consideration as the business case is further developed.

The cost factors were separated into those identified as being able to deliver economies of

scale (due to a declining rate of increase as the scale of rollout increases) and those unlikely

to deliver economies of scale (due to linear increase with the scale) but which may be

affected by a range of other factors that lead to increased or decreased costs. The factors

with the greatest overall impact on predicted costs, where more savings are identified than

potential additional costs are:

Economies of scale

Image and data acquisition – There could be significant savings depending on the

Copernicus imagery availability and storage systems. It takes considerable time to

searching for and source suitable imagery. If free imagery and data are available

from Copernicus which is held in a searchable well attributed data system (similar to

that used for Landsat 8 for example) this could reduce the cost.

Rule base development: There will be saving of scale and rule transferability,

working systematically around the UK. We recommend areas, for partnership groups,

that will cover approximately ‘6 Crick Projects’; the LNPs are the ideal size as in

some areas the local groups will have already formed into working partnerships.

Image processing and ancillary data preparation – savings from automated batch

processing and preparation of national datasets ready for use locally are likely to be

larger than increased costs from scaling up associated with more data processing

and storage, licenses and data maintenance.

Conversely, there are factors that cause anticipated increases to costs:

Increased costs

Hardware and software – costs increase significantly as further organisations

become involved in delivery without prior knowledge and experience of delivering the

approach and need to put in place the addition hardware and software needed to

deliver the project. Time required for training on the software is also a significant



consideration. This has the single biggest impact on additional capital costs and

outweighs capital savings.

9.2.1 Variation in types of cost in the Norfolk pilot

Capital costs have the largest variation when scaled up (Table 16and Table 17). The figures

presented are based on absolute minimum and maximum values calculated from the Norfolk

pilot as a breakdown of the type of costs was not readily available from the Monte Carlo

simulation. Rule base development is an area that accounts for large variation in estimated

staff cost, dependent upon whether there is existing habitat mapping to MasterMap available

locally, or not.

Table 16: Initial mapping costs for Norfolk

Type of cost

Initial mapping costs for Norfolk (£/km2)

Minimum Average Maximum

Capital cost 1.50 4.40 10.30

Staff cost 15.90 23.70 30.60

In kind contribution 6.80 8.50 10.20

Total 24.20 36.60 51.10

Table 17: Indicative cost of map update for Norfolk

Type of cost

Cost of map update for Norfolk (£/km2)

Minimum Average Maximum

Capital cost 0 2.30 2.80

Staff cost 5.10 8.70 11.90

In kind contribution 2.40 3.00 3.70

Total 7.50 14.00 18.40

10 Demonstrating effectiveness of the techniques for

Habitats Directive reporting

Cost effectiveness is illustrated using the example of Habitats Directive reporting. The UK

has a statutory obligation to report on the conservation status of 77 semi-natural habitats

listed in Annex I of the Habitats Directive, commonly referred to as Annex I habitats. These

are habitats considered to be internationally important or scarce across Europe. Article 11 of

the Directive requires Member States to implement surveillance of the conservation status of

Annex I habitats and Article 17 requires reporting of their status every six years.

For the 2013 report there was a huge variety of data utilised, which came from a range of

sources including Phase 1 maps, NVC surveys, common standards monitoring and SAC

surveys (2014, pers comm9). Only the latter deals directly with Annex 1 habitats, the others

9 Paul Robinson, JNCC



use other classification systems and so need some element of translation and analysis to be

used. The surveys will also have taken place over a period of several years. In many cases

the data used in 2013 will also have been used in 2007 but where it was deemed safe to

assume that it had not changed, the data was carried forward.

The approach was also different between the two reporting rounds. In 2013, each of the four

countries of the UK produced a national picture based on the data and expert opinion

available to them. These were then collated into a UK picture, with an element of calibration

applied. In 2007, the report for each habitat was coordinated at a UK level by the leads of

the appropriate Lead Coordination Network, so there were probably fairly different

approaches across different habitat types.

A transition to the Crick approach will substantially change the way data is derived for

Article 17 reporting and this will need to be justified to the EU which will require a

demonstration that in implementing change there are analyses in place to ensure there is

understanding (e.g. via modelling) of the extent to which the resulting data on habitat status

are a consequence of real change or a change in procedural approach.

The results of accuracy assessment (Medcalf et al., 2013), feedback from local habitat

practitioners and the assessment by the research team of how well the techniques worked,

all suggest that the EO techniques developed delivered through the Crick approach are fit for

purpose for supporting the mapping and surveillance of high priority habitats. The techniques

have been shown to be capable of consistent implementation and pilot projects demonstrate

that they are transferable. This does not mean that the techniques can be used to identify,

map and assess the condition of all Annex I habitats in isolation and it is not suggested that

they entirely replace field survey, but rather that the Crick approach directs it in a practical

and targeted way by supporting and augmenting current surveillance techniques, improving

the currency and coverage of the surveillance effort. Utilising the Crick approach in this type

of ‘supporting role’ delivers a range of improvements that will strengthen Annex I reporting

(Box 1):

Box 1: Benefits that the Crick Approach will bring to Article 17 reporting

improved consistency in data. Data is derived from a common methodological

approach which has controlled processes and QC

an improved evidence base, that is more readily traceable and updateable and that

makes good use of in-kind knowledge and contributions to cost;

some improvements to accuracy and completeness - through better definition of

boundary extent (a known weakness of field-based techniques and to a lesser extent

API) and through effective targeting of field resources to deal with known gaps in

knowledge.

potential for improved timeliness and responsiveness, subject to staff capacity and

resources being available. A rolling programme that would deliver a good level of

national coverage would maintain data at 6-7 years old and once initial mapping is

complete the timing of updates could be better tailored to reporting cycles. (Figure 6).



better more targeted use of staff providing fieldwork support to assess the status of

Annex I habitats

reduced costs of future updates, which are forecast to cost 60% less than the cost

of initial mapping.

Strengthen the ability locally and centrally to respond to the demand for

assessment of the impact of a range of other policy interventions designed to sustain

ecological function and the integrity of our natural capital.

potential for additional new measures of condition for monitoring the status of

Annex I habitats



Box 2: Demonstrating efficiencies through reuse of the data

The EO techniques developed have particular advantages that help deliver efficiencies in accordance with the principle of “gather once use many times”.

Mapping habitats will become more cost-effective when the inputs (imagery, data, rule-base) or outputs (maps, measures) are reworked, further developed or reused. The Crick approach lends itself well to this because the rule base and image segmentation work can be built upon or adapted. This keeps the cost of any subsequent analysis using automated techniques very low in comparison with repeat survey, manual re-interpretation of imagery or other traditional survey techniques.

The business case illustrates that in Norfolk, Wales and elsewhere maps produced initially to support the identification of habitats are being used for a wide range of other uses.

10.1.1 Improving governance

Adopting the Crick approach for Article 17 would allow JNCC to demonstrate to the EC that

many characteristics of good governance are in operation:

accountability – providing an improved evidence base enables JNCC, Defra and the

devolved administrations to explain and be answerable for their reporting of habitat

status to the EC.

transparent decisions – the basis of decisions are recorded and accessible through

the rule-base;

responsive – the delivery stakeholders can respond to demands for information

locally and centrally (with ability to target areas for gap-filling, meeting local demand

from early adopters and producing data that can be reused and reclassified for other

purposes);

equitable and inclusive - the involvement of local bodies and the subsequent local

adoption and use of the outputs are a central theme of the Crick approach;

effective and efficient – follows the principle of gather once use many times,

participatory – local engagement is actively sought and citizen science is built into

delivery of the Crick approach.

10.2 A rolling programme of delivery

An illustrative UK wide costed rolling programme has been modelled based upon a seven

year period for rollout in the UK to demonstrate the rate at which progress could reasonably

be made to support Article 17 reporting (Figure 6 and Figure 7). In summary, in this

illustrative rolling programme, at the end of 2031/32, the following will have been completed:

Initial mapping and updates 1, 2 and 3 for Wales

Initial mapping, a single update of the rest of the UK and 50% of a second update for

the rest of the UK

Post 2025/26, a complete update of all the UK will be completed within each Habitats

Directive Reporting period. The timing of map updates can then be more closely attuned to

reporting cycles to optimise the currency of the map for reporting purposes.



Much of the map updating is provisionally scheduled after initial mapping is completed, but

an update of the mapping of Wales could be rolled out earlier as it was mapped

predominantly using imagery from 2006 and therefore could already benefit from update.

Initial mapping starts at £0.5 million in 2015/16, when it is envisaged some mapping would

be completed in tandem with further preparatory work and rises to £1.65 million annually

from 2016/17 throughout 2021/22.

To calculate the present value of the total cost of a rolling programme of mapping the annual

costs were discounted at 3.5% using 2014/15 as both base and price year. Discounting is a

technique used to compare costs and benefits that occur in different time periods. It is a

separate concept from inflation, and is based on the principle that, generally, people and

society in general prefer to receive goods and services now rather than later (see: HM

Treasury Green Book). This is known as ‘time preference’. The estimated present value of

the total cost of the rolling programme of 17 years between 2016/17 and 2031/32 is

approximately £14 million. Figure 8 summarises the annual cost and present value.

The estimated present value of the cost of each habitat reporting cycle is approximately:

£6.71 million (2015/16 to 2019/20 inclusive);

£4.43 million (2020/21 to 2025/26 inclusive);

£2.76 million (2026/27 to 2031/32 inclusive).



Figure 6: Illustrative costed rolling programme of initial mapping to support Habitats Directive Reporting

Financial Year 2015/2016

2016/ 2017

2017/ 2018

2018/ 2019

2019/ 2020

2020/ 2021

2021/ 2022

2022/ 2023

2023/ 2024

2024/ 2025

2025/ 2026

2026/ 2027

2027/ 2028

2028/ 2029

2029/ 2030

2030/ 2031

2031/ 2032

Habitat reporting year (grey)

Total cost /year £millions* 0.5 1.65 1.65 1.65 1.65 1.65 1.65

Progress (Wales) Completed (Habitat map of Wales)

Progress (rest UK) 4.8% 21% 37% 52% 68% 84% 100% Initial mapping completed * These figures are based on the average projected costs, taken from Table 14. It excludes costs for initial mapping of Wales, which has already been completed.

Figure 7: Illustrative costed rolling programme of updates to support Habitats Directive Reporting

Financial Year 2015/ 2016

2016/ 2017

2017/ 2018

2018/ 2019

2019/ 2020

2020/ 2021

2021/ 2022

2022/ 2023

2023/ 2024

2024/ 2025

2025/ 2026

2026/ 2027

2027/ 2028

2028/ 2029

2029/ 2030

2030/ 2031

2031/2032

Habitat reporting year (grey)

Total cost /year £millions** 0.2 0.2 1.2 1.2 1 1 1.2 1.2

Progress (Wales) 50% 100% Update 1 completed 50% 100% Update 2 completed 50% 100%

Update 3 completed

Progress (rest UK) 25% 50% 75% 100% Update 1 completed

25% 50% Update 2 in progress

** These figures are based on the average projected update costs, taken from Table 15.

Figure 8: Discounted total cost of a rolling programme of mapping to support Habitats Directive Reporting

Financial Year 2015/ 2016

2016/ 2017

2017/ 2018

2018/ 2019

2019/ 2020

2020/ 2021

2021/ 2022

2022/ 2023

2023/ 2024

2024/ 2025

2025/ 2026

2026/ 2027

2027/ 2028

2028/ 2029

2029/ 2030

Present value of total cost (£millions)

Total cost of mapping (UK) £millions

0.5 1.85 1.85 1.65 1.65 1.65 1.65 1.2 1.2 0 0 1 1 1.2 1.2

Discounted total cost of mapping (at 3.5%, constant prices and base year 2014/15)

£0.48m £1.73m £1.67m £1.44m £1.39m £1.34m £1.30m £0.91m £0.88m £0 £0 £0.66m £0.64m £0.74m £0.72m £13.9m



10.2.1 Alternative approaches

Aside from continuing with the current methods used for Article 17 reporting there are no

identified alternatives that will deliver national data for Annex I reporting. Countryside Survey

which is sample based cannot deliver the data required because of the dispersed and

fragmented nature of many of the habitats. Land Cover Map provides a consistent approach

to mapping the whole of the UK, however this is at an insufficient spatial or thematic scale

for many regional uses on the ground.

If a programme of roll-out is not followed up, there is still likely to be a series of different

habitat mapping approaches going ahead, which will cause many integration issues e.g.

relating to scale and approach. This will essentially mean that the situation remains similar to

the current situation where habitat data is highly variable in terms of scale, coverage and

habitat system. Any time data is required at a national level, the data needs re-working and

modelled into an approximation of the UK habitat situation, with good information in some

areas and no information available in others.

10.2.2 In kind contributions

The Crick approach enables Article 17 reporting to make very good use of in-kind

contributions from local specialists and volunteers. The setup is mutually beneficial as there

is knowledge and skills transfer and the map is made available locally for further use.

With many monitoring programmes that support policy, costs tend to increase over time,

primarily because of the increasing costs of staff time. Fieldwork can be particularly difficult

to deliver and as such the risks of overspend on projects with fieldwork are high. In CS2007

there was a 22% overspend on fieldwork (Haynes-Young, et al., 2011). In this proposed

approach fieldwork is mainly to be carried out by staff from local bodies supported by

volunteers who are best placed to do this work efficiently as they know their locality. Thus

the risk of overspend on fieldwork is reduced in comparison with programmes such as

Countryside Survey and the risks are shared with partners. However it does have risks

associated with it such as the quality of the data provided should be subject to rigorous

quality check to avoid variability and local staff are under pressure from a growing work load

and cuts. Some investment at local levels will be required to make this work.

10.2.3 Cost of re-use of the data

Once imagery and other data has been collected, and processed to a high standard ready

for use, it can easily be re-analysed and reused, meaning the outputs can be improved upon

and merged with other information to provide further outputs to support local needs. Specific

questions can be addressed very quickly using the existing data, such as attributing areas

with additional information from basic indices such as wetness or productivity. Many of these

uses are listed in Table 3. Such additional analyses can be used to identify the condition of

habitats, assess the risk they are at from invasive species or target areas for fieldwork.



11 Value for Money

The budgets available to organisations for monitoring and surveillance are unlikely to

increase, so increasing the range of techniques available to habitat practitioners is only of

value if a newly introduced method allows the toolkit as a whole to deliver better value for

money. Four approaches to the assessment of value for money were considered in the

Phase 1 report (Medcalf et al, 2011); these are presented in order of importance. It is

recommended that these be more fully updated as the Business Case develops:

1. Value of information to improve the management of ecosystems and biodiversity (Value for Management)

2. Comparisons with the costs of other mapping techniques (Cost Comparison).

3. Value for scientific purposes (Science Value)

4. Comparisons with likely infraction fines for failure to meet obligations to protect the highest value designated land in the UK (Avoidance of Fines)

Each of these are now considered in the following sections.

11.1 The value of information to improve ecosystem management

for biodiversity (Value for Ecosystem Services)

Information that improves our knowledge of the location, status and trends of higher priority

habitats is fundamental to ensuring the protection and beneficial management of such

habitats and the species that depend upon them, both in a local and strategic context. In the

UK, there is an acknowledged gap in the provision of such information for many of the higher

priority habitats; the information currently available provides only part of the picture. This in

turn leads to uncertainty when evaluating the environmental and societal benefits arising

from existing and planned future investments in natural resource protection.

There are a range of policy initiatives with elements that specifically target the management

and protection of rare and vulnerable habitats of recognised importance. Most of these

initiatives have a broader remit, contributing to the delivery of wider ecological and

environmental benefits, such as the WFD, with additional ecosystem goods and services

such as water storage, provision and flood mitigation. As a consequence it is often not

possible to isolate the actual expenditure on Annex I or BAP Priority Habitats per se.

It is clear that there is substantial expenditure on policy initiatives that target resources

towards these habitats and the species that depend upon them. Funding for these is

provided both by the UK Exchequer and from European funding mechanisms. The annual

costs of meeting biodiversity objectives of the UK BAP in England were estimated at

upwards of half a billion pounds (Lawton, et al, 2010). Defra has estimated that some

£395 million was spent managing SSSIs between April 2000 and March 2008 (NAO, 2008).

Considerable effort is now being put into valuing natural capital by Defra both in the UK and

internationally. A study by Cao et al (2009) estimated the combined cost of meeting

environmental objectives for water management (flood risk and water quantity measures),

climate change mitigation (protecting the major carbon stores in peat and woodland), soil

quality, resource protection (including improving water quality) and public access at £517.3

million. These estimates do not include any costs associated with regulatory compliance



and demonstrate the additional costs that need to be taken into account when habitats are

considered and managed for the wider goods and services they provide beyond their

inherent ecological benefits.

There is an increasingly recognised need to view ecological management of habitats in a

wider context, and an aspiration to plan for a coherent and resilient managed ecological

network with habitat management delivered at the landscape scale. Taking a forward look,

the Lawton report (2010) estimates that the annual cost of establishing a coherent and

resilient ecological network in England, a framework that will be necessary for the long-term

sustainability of higher priority habitats, is likely to lie somewhere between £600 million to

£1.1 billion. The report recognises the need to plan for the medium and longer term stating

that “the sooner we act, and the better we are at focussing our actions to enhance the

Network, the lower the eventual cost will be”; an important acknowledgement that no or

delayed action will result in higher costs overall due to a loss of environmental goods and

services.

Earth Observation already makes a substantial contribution to identifying, mapping and

understanding the importance of managing land for a range of ecosystem services;

LCM2007 for example provided key information on the more commonly found Broad

Habitats for the UK National Ecosystem Assessment. Further use of EO techniques will

enable improved identification and monitoring of higher priority habitats for a range of

purposes. Improved mapped inventories, in particular, will improve the efficacy and make

more cost-effective a wide range of current policy measures directed at natural resource

protection. They will also facilitate work to value natural capital and inform the early stages

of policy development in relation to initiatives to improve ecological networks. The costs of

initial mapping using the Crick approach will be of the order of £14 million this would

represent a very small percentage (< 0.02%) of upwards of £500 million in policy costs of

existing funded initiatives. In this context investment in EO techniques appears to represent

very good value for money.

11.2 Comparisons with the costs of current initiatives (Cost

Comparison)

In Wales, Phase 1 mapping was completed, over the period 1979 to 1997 at an estimated

cost of around £10 million. Many such mapping initiatives have taken place, often at a

county scale, across the UK often using differing mapping techniques. If a programme of

roll-out is not followed up, there is still likely to be a series of different habitat mapping

approaches going ahead anyway which will cause many integration issues due to

differences in scale, and approach.

Phase 1 of this research project sought data for cost comparison of habitat mapping

techniques but concluded that there were no published data comparing different techniques

and that costs available were derived in differing ways. In order to achieve a framework of

comparative costs information was compiled from various sources including ad-hoc

discussions, published and unpublished literature and varied anecdotal experiences over

running a broad range of projects. Typical comparative costs were presented for field-based

methods, semi-automated EO-based methods and automated EO-based methods based on

set assumptions for an area the size of a 60km x 60km SPOT scene (Medcalf et al, 2011) in

a hypothetical example supposing an area of half uplands and half agricultural

lowlands/forestry to provide a standard for comparison (Table 18). Applying the same



criteria to the relevant stages of the costs for the initial and map updating costs for Norfolk

confirm these estimates for automated EO-based methods.

Table 18: Proposed ranges of costs of different techniques for a hypothetical SPOT scene

Initial mapping Re-survey [1] Total

Field-based

methods £340K-£570K £340K-£570K £680k - £1,140k

Semi-automated

EO-based methods £90K-£140K £90K-£140K £180k - £280k

Automated EO-

based methods £65 - £110K. £45 - £80K £110k - £190k

The Crick approach is slightly cheaper than semi-automated methods and much more cost

effective than traditional field survey at the initial mapping stage. The costs of updating the

map reduce dramatically in contrast to other habitat mapping techniques. In comparison to

other habitat mapping techniques the Crick approach appears to offer very good value for

money.

11.3 Value for scientific purposes (Science Value)

Investments in taking forward techniques developed for habitat surveillance and monitoring

of higher priority terrestrial and freshwater habitats are likely to provide scientific spin-off into

new areas which have cost and value for money implications. Of particular importance to

the Habitats Directive will be the development of measures of the condition of habitats.

It is becoming increasingly evident that certain invasive species are causing damage to

ecosystems, native habitats and species in the UK. In 2008, a strategy was developed to

meet the challenge posed by invasive non-native species in Great Britain. Of particular

concern are plant species such as Japanese Knotweed, Himalayan Balsam and Giant

Hogweed, and more recently the fungal pathogen phytophthora, which is causing extensive

damage to Larch plantations and other trees and shrubs. The cost of controlling these

invasive species can be enormous. The National Trust has estimated that the cost of

eradicating Japanese Knotweed alone from the UK would be in excess of £1.5bn (National

Trust), or £2.3bn per year to control all non-native species (Biodiversity is life 2010). EO has

the potential to contribute to the identification of the distribution of these invasive species

and targeting appropriate control treatments. An EO approach has been piloted to

identifying the likely distribution of Japanese Knotweed in Caerphilly CBC; the approach is

now being rolled out across all key areas of Wales. The Crick approach was used for the

successful mapping of Himalayan Balsam in the Norfolk Broads as part of Phase 2 of this

project. Such approaches could be delivered as one-off activities locally, or at larger scales

on the basis of need and subject to funding.

[1] Assumes a full resurvey for all three methods, so that information is available on all land use ecosystems services

applications as well as BAP and Annex 1 mapping. Note that other options should also be considered for monitoring e.g. risk

based sampling.

https://secure.fera.defra.gov.uk/nonnativespecies/index.cfm?sectionid=55



There are likely to be other opportunities for scientific spin-off into new areas. Broad based

stakeholder groups such as UK Forum for Earth Observation Applications provide an

opportunity for knowledge transfer to identify these. For example, techniques and

approaches developed for habitat surveillance and monitoring of higher priority terrestrial

and freshwater habitats may provide techniques that are applicable to the environmental

monitoring of marine environments, which can be difficult and potentially dangerous to

survey from the ground, and are subject to increasing EU regulations. In addition,

opportunities could exist to better understand how Priority Habitat extent and condition

relates to measures of biodiversity and ecosystem services in the wider countryside. Data on

management for these sites alongside weather information and data for other possible

drivers of change also provides a great deal of potential for understanding changes in

condition and extent.

11.4 Avoidance of fines for failure to meet obligations to protect

the highest value designated land in the UK

Infraction proceedings can be taken by the European Commission against Member States

for failure to comply with an obligation under an EU treaty. The failure to properly transpose

and enforce an EU obligation can eventually lead to a fine. The penalty represents a

material loss to the UK Exchequer and falls outside of Parliament’s intentions in relation to

the proper administration of European funding.

The total value of disallowance penalties accrued or provided for since the introduction of the

CAP 2005 in 2008- 09, is currently £580 million[2]. This illustrates the serious costs involved if

non-compliance is shown. This case is of relevance from an EO perspective because EO

has potential for use as a rapid and cost-effective solution.

To date no fines have been imposed on the UK government for non-compliance with aspects

of Directives relating to the protection of higher priority habitats. In other EU member states,

Ireland was referred to the Court of Justice for a continuing failure to protect sensitive

peatlands from peat extraction; similarly Spain has been given a written warning (a precursor

to pursuing an infraction fine) for failing to protect Natura 2000 sites by allowing opencast

mining developments.

Concerns have been expressed by UK wildlife organisations that current spending cuts on

environmental protection could lead to damage to habitats and destruction of protected

species, and a situation arising where non-compliance (and hence infraction fines) for failing

to protect habitats from damage could be more likely (Belfast Telegraph 2011).

To summarise the situation with regard to infraction fines, there is a risk, and even when not

subject to fine, the resolution of infraction proceedings are both stressful and resource

intensive. These risks would certainly be reduced by demonstrating that improved evidence-

based surveillance and monitoring systems have been developed and put in place to report

with more confidence on the status and trends of these habitats.

[2] http://www.nao.org.uk/wp-content/uploads/2014/07/Department-for-Environment-Food-and-Rural-Affairs-and-

Rural-Payments-Agency-Accounts-2013-14.pdf



References

Breyer, J., 2014, Recommendations for updating the Habitat Inventory of Wales, A Report

prepared by Environment Systems for NRW.

Cao, Y., Elliott, J., McCracken, D., Rowe, K., Whitehead, J. & Wilson, L. (2009). Estimating

the scale of future environmental land management requirements in the UK, Report to the

Land Use Policy Group. ADAS UK Ltd and the Scottish Agriculture College.

Haynes-Young, R., Potschin, M., Deane, R., and Porter, J., (2012), Policy Impact and Future

Options for Countryside Survey. Final Report, Defra Project Code WC1030, 152pp.

JNCC and Defra (on behalf of the Four Countries’ Biodiversity Group). (2012). UK Post-2010

Biodiversity Framework. July 2012. Available from: http://jncc.defra.gov.uk/page-6189.

Lawton, J., Brown, V., Elphick, C., Fitter, A., Forshaw, J., Haddow, R., Hilborne, S., Leafe, R., Mace., Southgate, M., Sutherland, W., Tew, T., Varley, J., Wynne, G. (2010) Making Space for Nature: A review of England’s wildlife sites and ecological network.

Lucas, R., Medcalf, K., Brown, A., Bunting, P., Breyer, J., Clewley, D., Keyworth, S. and Blackmore, P., 2011, Updating the Phase 1 habitat map of Wales, UK, using satellite sensor data, ISPRS Journal of Photogrammetry and Remote Sensing, 66 (1), pp81-102.

Medcalf K. A., Parker J.A., Turton, N., and Finch C. (2011) Making Earth Observation Work

for UK Biodiversity Conservation – Phase 1. JNCC Report No. 495 Phase 1, JNCC

Peterborough 2014

Medcalf K. A., Parker J.A., Turton, N., and Bell, G. (2013) Making Earth Observation Work

for UK Biodiversity Conservation – Phase 2. JNCC Report 495 Phase 2, JNCC

Peterborough 2014.

National Audit Office (2008) Natural England’s Role in Improving Sites of Special Scientific

Interest. Report by the Comptroller and Auditor General. The Stationery Office.

National Trust [www] Japanese Knotweed Clearance Reaches Milestone http://www.nationaltrust.org.uk/main/w-global/w-news/w-latest_news/w-news-japanese_knotweed.htm, (Accessed 24th January 2011). Parker J. A., Medcalf K. A., Turton, N., (2014, draft) Making Earth Observation Work for UK

Biodiversity Conservation – Phase 3: Cost effective methods to measure extent and

condition of habitats: Identifying common requirements for Crick outputs and approaches.

Rodwell, J.S. (ed.) 1991a. British Plant Communities. Volume 1. Woodlands and scrub.

Cambridge University Press.

Rodwell, J.S. (ed.) 1991b. British Plant Communities. Volume 2. Mires and heath.

Cambridge University Press.

Rodwell, J. S. (ed.) 1992. British Plant Communities. Volume 3. Grassland and montane

communities. Cambridge University Press.

Rodwell, J.S. (ed.) 1995. British Plant Communities. Volume 4. Aquatic communities,

swamps and tall herb fens. Cambridge University Press.

http://www.nationaltrust.org.uk/main/w-global/w-news/w-latest_news/w-news-japanese_knotweed.htm

http://www.nationaltrust.org.uk/main/w-global/w-news/w-latest_news/w-news-japanese_knotweed.htm



Rodwell, J.S. (ed.) 2000. British plant communities. Volume 5. Maritime communities and

vegetation of open habitats. Cambridge University Press.

Wildlife Trusts (2014) The status of England’s Local Wildlife Sites 2014.



Appendix 1: Fuller description of potential project stakeholders & other interested parties

Stakeholder Description Notes

Defra UK Government department Responsible for policy and regulations for the natural environment (incl. biodiversity and ecosystems services) and for a wide range of other policy areas that concern or impact upon the natural environment (incl. sustainable development, environmental protection and pollution control, climate change adaptation and mitigation). Defra only works directly in England, it works closely with the Devolved Administrations generally leads on negotiations in the EU and internationally. Evidence requirements at all stages of the policy cycle. Consistency of evidence important for UK and national scale reporting

JNCC Public Body that advises the UK Government and Devolved Administrations on UK-wide and international nature conservation.

Devise strategies for collecting and using data. Provide objective advice on the priorities for creating a cost-effective evidence-base for use across the UK and the investment required. Provide support to ensure that European / international requirements are understood and implemented through national policies and actions

Devolved Administrations

Scottish Government – Environment and Rural Affairs Dept. Welsh Government – Directorate of Sustainable Futures Government of Northern Ireland – Department of the Environment

Responsible for all issues relating to the natural environment that are not explicitly reserved to the United Kingdom Parliament at Westminster.

Other Government Departments

Forestry Commission England Forestry Commission Wales DECC, MOD, DCLG, DFID, DoT

The Forestry Commissions are responsible for woodland creation and management. Natural Resources Wales has taken over the functions of Forestry Commission Wales, as well as some functions of Welsh Government. Biodiversity Duty - all public bodies to have regard to biodiversity conservation when carrying out their functions.

Country Conservation Agencies

Natural England, Scottish Natural Heritage, Natural Resources Wales & Environment and Heritage Service (Northern Ireland)

Statutory advisory agencies to country governments, with wide range of regulatory, delivery and reporting responsibilities for policy and legislation concerning the natural environment Natural Resources Wales has taken over the functions of the Countryside Council for Wales

Environment Agency, SEPA

Environment Agency Scottish Environment Protection Agency Northern Ireland Environment Agency

Lead authority for implementation and monitoring of the Water Framework Directive (incl. River Basin Management Planning) Advisory, delivery and regulatory functions with role to protect and improve the environment Responsibilities for flood and coastal risk management, water quality and resources, conservation and ecology Natural Resources Wales has taken over the functions Environment Agency Wales

http://en.wikipedia.org/wiki/Reserved_matters

http://en.wikipedia.org/wiki/United_Kingdom_Parliament

http://en.wikipedia.org/wiki/United_Kingdom_Parliament

http://en.wikipedia.org/wiki/Westminster




Local Authorities Local Authorities (426 in UK) The NPPF is clear that pursuing sustainable development includes moving from a net loss of biodiversity to achieving net gains for nature, and that a core principle for planning is that it should contribute to conserving and enhancing the natural environment and reducing pollution. Local planning authorities (LPAs) have legal obligations to take nature conservation into account when determining planning applications. LPAs and neighbourhood planning bodies should seek opportunities to work collaboratively (e.g. with LNPs to develop and deliver a strategic approach to protecting and improving the natural environment based on local priorities and evidence

Organisations responsible for designated landscapes

National Park Authorities (England, Scotland, Wales) National Parks England, National Parks Wales AONBs, National Scenic Areas

National Parks are Planning Authorities. These organisations produce area management plans such as National Park Management Plans (England, Wales) / National Park Plan (Scotland).

Partnerships (at local scale)

Local Nature Partnerships (LNPs – 43 areas in England), Local Sites Partnerships (LSPs – 53 in England), ‘Living Landscapes’ (Wildlife Trusts – ~100 in England & Wales) Other partnerships (e.g. running IDBAs)

Identifying and facilitating management of natural resources Delivering landscape scale initiatives Often supported by local or regional scale Environmental Data Centres

Organisations involved with the pilot project

Northern Upland Chain LNP, Yorkshire Dales NPA, Northumberland NPA,NEYEDC (data centre), CEH, Norfolk Biodiversity Information Centre, Nidderdale AONB, North Pennines AONB, Norfolk Wildlife Trust and RSPB.

As above for National Parks, LNPs and Environmental Data Centres CEH run Countryside Survey (for CEH, also see “scientific research organisations”)

Organisations responsible for transport corridors

Canal & River Trust (England, Wales) Network Rail, British Waterways / Scottish Canals Highways, Inland Waterways (NI),

National influence over management of substantial wildlife corridors Biodiversity Duty - all public bodies to have regard to biodiversity conservation when carrying out their functions.

Organisations with a strategic interest

UK Space Agency - Space for Smarter Government Programme, UK Earth Observation Framework

SSGP are producing guidance on the production of business cases concerning the adoption of EO techniques UKEOF co-ordination role – provides opportunities to disseminate info to many organisations

National Ecosystem Assessment

The UK National Ecosystem Assessment, The UK National Ecosystem Assessment Follow on

Other ecosystems services initiatives

BESS biodiversity & ecosystem service sustainability, Ecoserv-GIS, Site Informer, Polyscape, Invest, Aries, SENCE,SCCAN, JNCC Spatial Framework

https://www.gov.uk/government/policies/protecting-biodiversity-and-ecosystems-at-home-and-abroad/supporting-pages/local-nature-partnerships




Scientific research organisations

CEH (UK soil observatory, Countryside Survey), FERA (Agricultural Change and Environment Observatory Programme), Forest Research, James Hutton Institute

Often run environmental national or UK scale observatories on behalf of Government that are part of ‘big data’ initiatives Run national scale monitoring activities – such as Countryside Survey

Key Data providers Copernicus (satellite imagery), Ordnance Survey (Mastermap), Rural Payments Agency (RLR), Environment Agency (LiDAR) Country Conservation Agencies (aerial photography / wide range of other mapped data)

For creation of real world objects that are fundamental to the OBIA approach Detailed height data improves mapping and allows some habitats to be distinguished Aerial photography

Specialist Groups with specific remits

UK Biodiversity Indicator Steering Group UK Biodiversity Indicator Forum GB Non-native Species Secretariat / Invasive Species Ireland Strategic Regional Coastal Monitoring Programmes Terrestrial Biodiversity Group, Natural England National Trust

Development of indicators and approaches for assessing progress with achieving them.



Appendix 2: Risk register

Issue 1 – for review prior to inclusion in Business Case – covering risks related to rollout of the Crick approach

Role of Steering Group: Notify PM of any external risk to the project.

Make decisions on PM’s recommended counter-measures

Strike balance between level of risk and potential benefits that project may achieve.

ID Description of risk Impact Probability Counter-measures Owner Date Identified

Last update

Status

Risks of a technical nature

1. Cloud cover results in failure to acquire necessary imagery– increases costs associated with image tasking and could result in delays

High Med Define areas at risk. Plan and allow funding necessary for image tasking and fieldwork. Recognise that Crick approach will not provide complete coverage.

Open

2. Unforeseen costs for licenses – nationally e.g. MM and locally from LRCs

Med Med Begin discussion at an early stage with suppliers. Check feasibility of using LPIS data where available. Check status of local licensing at early stage with LRCs.

Open

3. Poor quality aerial imagery leads to large increase in cost of producing rule bases

High Med Improve data collection protocols and specify acceptance criteria for acceptable photography from contractors via central procurement process. Review quality of available AP’s when planning rollout sequence.

Open

4. Stringent requirements for image processing aren’t adhered to – affects quality of outputs

High Med Specify requirements through protocols and include as a requirement of delivery in procurement process

Open

5. Poor quality aerial imagery leads to large increase in cost of much increased costs

High Med Improve data collection protocols and specify acceptance criteria for acceptable photography from contractors via central procurement process. Review quality of available AP’s when planning rollout sequence.

Open

Risks associated with skills and capacity to deliver

6. No central support system in place (e.g. via a National and / or habitat mapping hub) to support delivery – delays, inconsistent standards

High High SG to track progress with funding and setup of an EO hub and provide information on needs arising from Crick approach. SG to progress obtaining funding and drawing up plans for a Habitat hub

Open

7. Habitat hub is more difficult to setup than anticipated – delays, costs.

Med Low Early planning of requirements and costs Open

8. No central support system in place (e.g. via a National and / or habitat mapping hub) to support delivery – delays, inconsistent standards

High ? SG to track progress with funding and setup of an EO hub and provide information on needs arising from Crick approach. SG to progress obtaining funding and drawing up plans for a Habitat hub

Open




Last update

Status

9. Poor implementation of ecological rule base development

Med Med Ensure business plan deals with training needs and specifies appropriate skills. Follow through via procurement requirements.

Open

10. Insufficient ‘central’ flexibility to adapt to local needs – lose local buy-in, and produce information that isn’t sufficiently locally relevant

Med Low Ensure that Crick approach implemented as planned Open

11. Loss of key skilled staff in the ‘habitat hub’ – cause delays.

Med Med Business Plan needs to identify training needs and level of additional capacity building required.

Open

12. Loss of ecological skills at the local and national (delivery agencies) level – delays and overall quality of outputs for national reporting.

High Med Review status of skills (via past studies and if necessary consultation) as part of business plan.

Open

13. Lack of capacity to respond to unforeseen needs – staff and skills shortage leaves project vulnerable to unforeseen problems

High Med Review overall skills needs and produce statement of need via Business Planning

Open

Risks related to commitment to approach:

14. Lack of confidence from users at the national level – habitat products underutilised, potential for impact on national reporting needs.

High High Communication and knowledge exchange strategy. Active engagement with key individuals to ensure benefits to organisations are understood by users. Assess likely impact on overall aims arising from non-use by key sectors at the national level if risk is considered high.

Open

15. Lack of interest locally - increases demand on central capabilities (habitat hub) increases costs and slows delivery. Potential target areas for gap-filling remaining poorly understood.

High Med The wider interest in engagement at a local level is unknown. Communication is planned to gauge this as part of current project.

Open

Risks related to funding:

16. Lack of commitment by funders – Rollout will stall without finance. Slow uptake by funders could result in delays to programme of rollout.

High Med Active engagement by Defra / JNCC with funders. Produce costed Business Case and develop this as project progresses. Follow up with business Plan

Open




Last update

Status

17. Costs estimates are inaccurate being either:

1) overestimated – delivery unnecessarily expensive – seen as poor VFM by potential sponsors - doesn’t attract the necessary funding for further investment.

2) underestimated - funding shortfall will arise with insufficient funds available to cover agreed programme of implementation & development.

3) Incorrectly profiled – leading to funding shortfalls or resources being available ahead of the ability to implement sites or some types of monitoring.

High Low Continue to refine and make available costs in the Business Case, especially for areas of cost identified as having uncertainty. Continue to check projected costs against costs incurred in delivery of the pilots and other projects. Review and further develop a rolling programme of delivery in a Business Plan.

Open

18. Cost estimates do not show sufficient detail to enable investment decisions by potential sponsors – causing loss of confidence and lack of support from potential sponsors & stakeholders. Funding and other opportunities missed.

High Low Open

19. Skills and staff availability at local level overestimated – increased costs and delays.

High Med The capacity for engagement at a local level is unknown (outwith the pilot areas). Communication is planned to gauge this as part of current project.

Open

20. Unknown costs for setup and running of central co-ordination hub – leads to lack of commitment by potential funders.

High Med Agree delivery model, secure organisational commitments (staff) and cost to include in Business Case

Open

Other risks:

21. Copernicus data is not available or not FOC –increase costs and / or reduces image data available to project

High Ensure BC identifies the impact of this on costs and treat as contingency cost until situation is known.

22. Disease outbreak or other restrictions on access to land locally will impact on speed and ability to deliver programme of rollout.

Med Low Monitor risks and ensure some flexibility to adjust the sequence of areas mapped during the rolling programme. Develop contingency plans if risk increases.

23. Data processing – hardware / software failure. Causes significant delays.

High Low Identify key equipment requirements for project. Ensure that alternatives and/or spares are available if required.

24. All or parts of the UK leave the EU – statutory policy driver is potentially removed

Low Low Ensure funding is not contingent on a single policy driver.

business case draft 1 - GOV.UK

Documents

Transcript of business case draft 1 - GOV.UK