From data to information, from information to

knowledge

Computational Sustainability and land use Prof.ssa Maria Grazia Albanesi

Università di Pavia - Facoltà di Ingegneria

Outline

Research/Development field: Computational Sustainability

ACI Project

The problem: Land use

Solution: Anthropentropy indicator

First approach: GIS data

Second approach: open data Google Earth & social network

Third approach: open data & open software

DANTHE Project

Goals

Methodology

Case studies

A cosa può servire il progetto? Proposte e discussione

Presentation

Maria Grazia Albanesi is the chief of

Informatica per la Sostenibilità

Computational Sustainability Unit

csu.unipv.it

Site: DCAlab

Multidisciplinary approach:

Dipartimento di Scienze Economiche e Aziendali

Dipartimento di Scienze della Terra e dell’Ambiente

http://csu.unipv.it/

International framework

Computational Sustainability

First conference: 2009 USA

Definition (1987) rapporto della World Commission on Environment and

Development (WCED), commission Brundtland, Our Common Future:

«Sustainable development is development that meets the needs of the present without compromising the ability of future generations to meet

their own needs”

It contains within it two key concepts:

The concept of 'needs', in particular the essential needs of the world's poor, to which overriding priority should be given; and

The idea of limitations imposed by the state of technology and social organization on the environment's ability to meet present and future

needs.

Sustainability

Implicit in teh definition:

Temporal spread of years

We nedd to measure the effects of

defintion

Opposing needs present-future

I progetti

Progetti precedenti/attuali:

Progetto ACI: un nuovo indicatore (Fattore di Antropentropia, FA) di consumo di suolo per

misurare il livello di degrado del consumo di suolo. Febbraio 2013

Progetto DANTHE: un modello predittivo e il relativo sistema di supporto alle decisioni

basato sull’indicatore FA per predire il futuro consumo di suolo di nuovi insediamenti

urbani. Novembre 2013.

Progetti/attività correnti:

Completamento del progetto ACI

Studiare le correlazioni tra il nuovo indicatore e altri bioindicatori, eventualmente

proponendone di nuovi (per es: lo stato delle foreste). Con Scienze della Terra e

dell’Ambiente.

Studiare la correlazione tra il FA e indicatori economici (popolazione, livello di ricchezza,

ecc…)

Fare trasferimento tecnologico verso le aziende.

Land use: definition

The percentage of land occupied by:

houses

Buildings (firms, schoiols, hospitals ecc.)

roads

railways

Intensive agriculture sites (i.e., plant nursery)

Land use: negative effects

Soil sealing

Fragmentation and reduction of biodiversity

Pollution

111 788 ha/year in EU*

Land take in Italy: + 6.3%** (1956-2006)

Sources: * European Environment Agency, **ISPRA

8

Fragmentation

9

Effects

Effects of fragmentation Landscape Fragmentation in Europe, EEA –FOEN T.R n. 2/2011

Effect of fragmentation Landscape Fragmentation in Europe, EEA –FOEN T.R n. 2/2011

Un uso scorretto della frammentazione

Fragmentation

Uncorrect indicator

Decentration

Belgio 97 Ecuador 66 Gran Bretagna 90 Irlanda 62 Danimarca 87 Grecia 61 USA 82 Siria 55 Canada 80 Cina 44 Francia 78 Egitto 43 Germania 74 Zimbabwe 38 Svizzera 74 India 30 Russia 73 Lesotho 26 Italia 68 Afghanistan 24 Nepal 18

0 20 40 60 80 100

Belgio

Gran Bretagna

Danimarca

USA

Canada

Francia

Germania

Svizzera

Russia

Italia

Ecuador

Irlanda

Grecia

Siria

Cina

Egitto

Zimbabwe

India

Lesotho

Afghanistan

Nepal

Percentage of habitants who live in urban areas

New indicators are needed

Classical approach

Related to population

Area

Bio-indicators

15

Requested features

Fragmentation and continuity

Easy to be computed and updated

Automatic or semi-automatic

Framework

Models DPSIR

Type of indicators (A, B, C, D)

Sources:

E. Smeets and R. Weterings, “Environmental indicators: typology and overview,” European Environment Agency Technical Report n. 25/1999 European Environment Agency, “Urban sprawl in Europe - The ignored challenge,” EEA Report n. 10/2006 European Environment Agency, “Towards a green economy in Europe - EU environmental policy targets and objectives 2010–2050,” EEA Report n. 8/2013

16

A new indicator: Anthropentropy

Anthropos (Άνθρωπος) = man & Entropy (disorder)

The anthropentropy of a territory

AF = DA / (S – NA)

Where

S: Area of the territory

DA: Death Area

NA: Neutral area

Neutral Area

Neutral area is the part of the territory containing:

inland water (lakes, lagoons etc.) > 2kmq

lands located > 3,000 m above sea level.

Death Area

The anthropic places are enlarged of a buffer of 50 m in each direction (bidimentional metric).

Death area and fragmentation

20

The metric

AF

0 < AF <= 0.2 OK optimal environment preservation for land use, ideal condition 0.2 < AF <= 0.4 First worrying level 0.4 < AF <= 0.6 Serious level of degradation 0.6 < AF <= 0.8 Very serious level of degradation 0.8 < AF <= 1 Irreversible situation A possible application: to help local government in planning new urban expansions.

Data sources

Two main problems:

Availability of a description of land use in terms of anthropic places

Accuracy of scale comparable with the size of the dilation (50 m) accomplished by the algorithm.

Data sources:

Corine Land Cover Data and GIS software

Social network (UGC content and crowdsourcing)

IV – Legend of CLC

Workshop - L’informatica e i social network al servizio dell’ecologia: il progetto ACI Pavia, 18 aprile 2013

CLC types 2000

The goal of the project

AF

Each municipality is labelled with a AF value

ACI: Antropentropia Comuni Italiani (i.e., Italian Municipality Anthropentropy)

Liguria

ICCGI13: Nice, we are here!

Example: Lombardia

26

Pavia

27

VI – Sintesi dei risultati: Province & Comuni FA

The total Anthropentropy map for Italy

FA

Only a little part of Italy is covered with Corine Land Cover data What about white areas?? Here the social network came into the game!!! Users of social network Facebook linked to www.albanesi.it

UGC data and crowdsourcing

A “bottom-up”, collaborative procedure generates, from Google Earth maps, a map with anthropic places and neutral zones. Characteristics of the procedure: “Bottom” side: It requires a low level of computer skills Based on open software and open data (Google Earth maps and Gimp). “Up” side: Fully automated algorithm based on computer vision operators (software Matlab)

UGC Maps: examples

Procedure

Collect and check the UGC maps

Calibration (determination of scale)

AF computation by completely automatized image

processing procedure.

Calibration

Image processing

Maps are processed by image

processing primitives

Morphological operator: dilation Programming environment: Matlab Image toolbox

Our approach: critical issues

Drawbacks:

Precision and knowledge of the territory vary form map to

map Collaboration is fundamental

Advantages:

It increase awareness of the territory and environment

degradation by the citizen The calculation procedure of AF can be used in a scenario

"What if?“ It reacts more quickly to changes in the area (bottom - up

knowledge)

Some computed AF by UGC maps

45 maps received and processed in the first 3 months (rejected percentage <1%)

Municipality S (Area in kmq) Neutral Area AF

Scarperia 79.3700 0.00 0.1637

Erice 47.3000 0.00 0.214

Verbania 37.6200 0.37 0.674

Pomigliano d'Arco 11.4400 0.00 0.98040

Praia a Mare 22.9100 0.00 0.4370

An example

37

Cosa si riceve dagli utenti?

Procedura

Acquisizione e check della mappa UGC

Calibrazione

Lettura dati input

Calcolo del FA con elaborazione visuale/interattiva

Possibilità di loop a seconda della dimensione del comune.

Calibration

Computer vision algorithm

Mathematical morphology operator of dilation MATLAB

(Image toolbox)

Some values

Comune Superficie (kmq) Area neutra FA

Scarperia 79.3700 0.00 0.1637

Erice 47.3000 0.00 0.214

Verbania 37.6200 0.37 0.674

Pomigliano d'Arco 11.4400 0.00 0.98040

Praia a Mare 22.9100 0.00 0.4370

Adding temporal perpspective

DANTHE project How can you estimate the future land use? Hypothesis on new anthropic places (ap)

DANTHE = Dynamic ANTHropentropy Expansion

A new ap is, in turn, capable of attracting new sites (anthropogenicity)

Antrhopic places are divided in typologies

New expansions Uniform

Gravitational

>Road

43

DANTHE

The expansions

44

I luoghi antropentropogeni

45

Predictive modelDANTHE

New sites?

worst case approach

Predicted expansions in terms of growth/year

Term: 2050

𝐹𝑖 = 𝑆𝑓 − 𝑆𝑖

𝑆𝑖 ∙ 𝑦𝑒𝑎𝑟 , 𝑖 = 𝑎𝑛𝑡ℎ𝑟𝑜𝑝𝑜𝑔𝑒𝑛𝑖𝑐 𝑠𝑖𝑡𝑒 , 𝑖 = 1, 2, … 10

46

𝐹𝑖𝑚𝑎𝑥 > 𝐹𝑖 ∀𝐹𝑖

Example of computation of Fmax

47

𝐹𝑚𝑎𝑥 = 0,0503

48

𝐹𝑚𝑎𝑥 = 0,1053

Example of computation of Fmax

Uniform expansion

49

Uniform expansion

(with constraint)

50

Gravitational expansion

51

Road expansion (1)

52

Prediction(1)

53

Prediction (2)

54

Prediction (3)

55

Valutazione di sostenibilita’

Constraint of Land Use Sustainability (CLUS)

56

𝐴𝐹 2050 – 𝐴𝐹 𝑡

𝐴𝐹 𝑡< 0.05 𝐴𝑁𝐷

𝐴𝐹 𝑡 – 𝐴𝐹 2014

𝐴𝐹 2014 < 0.20

Future Constraint

Past constraint

Steps: The predictive model computes all teh expansions until 2050 The system computes AF(2014) e AF(t) Test of CLUS and conclude if the new site is sustainable or not.

Valutazione di sostenibilita’

57

Implementation

PHP

HTML

CSS

Javascript

About 3500 lines, with ad hoc libraries

58

59

DANTHE & DPSIR

DPSIR

Tipi di indicatori (A, B, C, D)

Sources:

E. Smeets and R. Weterings, “Environmental indicators: typology and overview,” European Environment Agency Technical Report n. 25/1999 European Environment Agency, “Urban sprawl in Europe - The ignored challenge,” EEA Report n. 10/2006 European Environment Agency, “Towards a green economy in Europe - EU environmental policy targets and objectives 2010–2050,” EEA Report n. 8/2013

60

Thank you!

More info: dcalab.unipv.it/ies mariagrazia.albanesi@unipv.it

To our children To let them know that we tried to change (G. Porro)