Guidelines for Developing and Compiling Vulnerability Maps...

138
Improved Methods for Aquifer Vulnerability Assessments and Protocols (AVAP) for Producing Vulnerability Maps, Taking Into Account Information on Soils WRC Project K5/1432 Deliverable 2 of GIS Component Guidelines for Developing and Compiling Groundwater Vulnerability Maps Using GIS 10 March 2006 Compiled by: Abraham Thomas Department of Earth Sciences, University of the Western Cape, P. Bag X17, Bellville 7535 & Julian Conrad and Zahn Munch GEOSS (Pty) Ltd Unit 19, 9 Quantum Street, TechnoStell, Techno Park, Stellenbosch 7600

Transcript of Guidelines for Developing and Compiling Vulnerability Maps...

Guidelines for Developing and Compiling Vulnerability Maps Using UGIf Model

Improved Methods for Aquifer Vulnerability Assessments and Protocols (AVAP) for Producing Vulnerability Maps, Taking Into Account Information on Soils

WRC Project K5/1432

Deliverable 2 of GIS Component

Guidelines for Developing and Compiling Groundwater Vulnerability Maps Using GIS

10 March 2006

Compiled by:

Abraham Thomas

Department of Earth Sciences,

University of the Western Cape,

P. Bag X17, Bellville 7535

&

Julian Conrad and Zahn Munch

GEOSS (Pty) Ltd

Unit 19, 9 Quantum Street, TechnoStell,

Techno Park, Stellenbosch 7600

TABLE OF CONTENTS

1SECTION 1

11Introduction

11.1Introduction

11.2Assessing Ground Water Vulnerability

31.3Process of Ground Water Vulnerability Assessments

51.4Key elements of vulnerability assessment

51.5Aquifer Sensitivity and Groundwater Vulnerability Methods

61.5.1Overlay and Index Methods

61.5.2Process-based simulation methods

61.5.3Empirical Statistical Methods

71.6Uncertainties associated with vulnerability assessments

71.7GIS-based algorithms and approaches to vulnerability assessment

71.7.1Main characteristics of GIS-based algorithms

91.7.2Selected GIS-based approaches to vulnerability assessment

101.8Purpose of the Guideline Document

101.9Organisation of the Guidelines and Procedures

101.10References

12SECTION 2

122GUIDELINES FOR DEVELOPING AND COMPILING GROUNDWATER VULNERABILITY MAPS USING UGIf MODEL

122.1OVERVIEW OF UGIf MODEL

122.1.1Model structure

142.1.2Modified UGIf model

142.1.3Various Menus available in UGIf

142.1.4Capabilities and Applicability

152.2Data Requirements and Outputs

162.2.1Data Requirement for Vulnerability Assessments using Modified UGIf

162.3Limitations, Prediction Accuracy and Reliability

162.3.1Limitations

172.3.2Prediction Accuracy and Reliability of Assessment

172.4Vulnerability Assessment Using UGIf Model

172.4.1Steps involved in vulnerability assessment for BTEX compounds

192.4.2Preliminary results of the revised UGIf approach

222.5References

23SECTION 3

233GUIDELINES FOR DEVELOPING AND COMPILING VULNERABILITY MAPS USING THE ReSIS METHOD

233.1Introduction

243.2The ReSIS layer method

243.3Elements of ReSIS layer model

263.4Required layers

263.5Assigning a vulnerability rating using the Zone Vulnerability Matrix

263.6Deriving a vulnerability weighting

263.7Intrinsic vulnerability

263.8Quantitative assessment of uncertainty

273.9Conclusion

273.10References

28SECTION 4 - APPENDICES

28Appendix 1

284Opening of ArcView Project for Vulnerability Assessment

284.1Preliminary steps

284.2Opening of ArcView Project file of UGIf model

33Appendix 2

335Procedure for Running the Direct Recharge Model

345.1Preparation of land use map in grid format acceptable to the model

365.2Land Use Grid Map Preparation

415.3Preparation of hydrologic soil group map

435.4Hydrologic Soil Group Grid Preparation

475.5Combine Grids and Assign NRCS Curve Numbers

525.6Preparation of meteorological data table

535.7Potential Recharge Estimation

63Appendix 3

636Non Point Source Pollution Assessment Using UGIf

636.1Model Inputs and Outputs

646.2Modelling Steps / Procedure for Estimation of BTEX Pollution in Recharge

646.3Assigning EMC Values

746.4Calculation of NPS runoff pollution load and initial recharge pollutant fluxes

806.5Combining Input Grid Maps for NPS Pollution Assessment

86Appendix 4

867Groundwater Vulnerability Assessment Using UGIf

867.1Calculation of Vadose Zone Travel Time of BTEX Compounds

897.2Running of Programs under Menu ‘Groundwater Vulnerability Assessment’

907.2.1Groundwater Vulnerability of Conservative Pollutants

947.2.2Assessment of Groundwater Vulnerability of Organic Pollutants (BTEX)

100Appendix 5

1008Groundwater Vulnerability Assessment using the ReSIS method: Coastal Park Area

1008.1Input data sets

1018.2Processing natural recharge

1028.3Deriving the Soil Zone rating

1058.4Rating the Intermediate zone

1098.5Finding the rating for the saturated zone

1108.6Combining the layers to produce a preliminary vulnerability

LIST OF FIGURES

4Figure 1.1The vulnerability assessment process (Source: NRC, 1993).

13Figure 2.1Structure of UGIf model (drift = superficial deposits).

13Figure 2.2Simplified flow chart for estimating pollutant mass fluxes at the water table.

15Figure 2.3Interface of the UGIf model for assessing NPS BTEX pollutant fluxes.

20Figure 2.4Attenuation Factor Values for Benzene in Recharge Waters of Coastal Park.

20Figure 2.5Leaching index potential values for Benzene in Recharge Waters of Coastal Park.

21Figure 2.6Ranking Index for Benzene in Coastal Park.

21Figure 2.7Distribution of travel time in years for a conservative contaminant.

25Figure 3.1Flowchart of ReSIS layer model

88Figure 4.1Attribute table showing travel time of BTEX comounds.

89Figure 4.2Distribution of Benzene travel time in days.

90Figure 4.3Interface of added programs for groundwater vulnerability assessment in UGIf.

93Figure 4.4Attribute table showing groundwater vulnerability of conservative pollutant.

94Figure 4.5Distribution of chloride travel time in years.

98Figure 4.6Attribute table showing groundwater vulnerability from screening level models.

99Figure 4.7.Attenuation Factor Values for Benzene in Recharge Waters of Coastal Park.

101Figure 5.1Functionality on the ReSIS menu item

102Figure 5.2User defined rates for recharge reclassification

103Figure 5.3Land types within the Coastal Park study area

104Figure 5.4Land type Ha7 terrain unit profile from the Land type memoirs

104Figure 5.5Predominant Land Types and DEM derived terrain units in Coastal Park

105Figure 5.6Soil vulnerability in the Coastal Park study area

106Figure 5.7Selecting the Import data option

106Figure 5.8Fields from the database that can be selected for interpolation

107Figure 5.9Specifying a new geographic projection if different from input data

107Figure 5.10Interpolate from menu item using the IDW method

108Figure 5.11Selecting the extent of the output grid

108Figure 5.12Clipping the newly created grid to the extent of the study area

109Figure 5.13The interpolated grid can now be reclassified

109Figure 5.14Reclassifying the depth to water level grid using EPA ratings

110Figure 5.15Selecting a weight for each layer

110Figure 5.16The final ReSIS rated vulnerability grid

SECTION 11 Introduction

1.1 Introduction

Groundwater is a very important natural resource widely used for different purposes like drinking, irrigation, industrial use etc. Two-thirds of South Africa, including more than 280 towns and settlements, are largely dependent on groundwater for their drinking water supply and development. Groundwater in urban environments all over the world is recently being becoming a natural resource of strategic importance owing to its limited resources, quality deterioration, increasing demand, and limited replenishment in urban set up (Thomas and Tellam, 2005). An adequate characterization of ground water resources in a region or province is an essential first step in developing effective, long-term ground water management programs. The information on availability of groundwater resources and its quality forms one of the keys to economic development in rapidly expanding urban, industrial, and agricultural regions. The sustainability of groundwater resources on all scales (whether urban, suburban, mixed or rural) concerns its quantity and quality. In order to meet the requirement of increasing urban population on a sustainable basis it should be available in good amount and of good quality (Thomas, 2001).

Groundwater resources are often highly vulnerable to contamination from human activity. The vulnerability potential of an aquifer to ground water contamination is in large part a function of the susceptibility of its recharge area to infiltration (US EPA, 1998). The importance of ground water has long been recognized, but the potential for ground water to become contaminated as a result of human activities at or near the land surface has only been recognized in recent years (NRC, 1993). Groundwater contamination causes degradation of water quality. Once contaminated, ground water is very expensive to clean up; in many cases, cleanup may not be possible within a reasonable time (Mackay and Cherry 1989, Haley et al. 1991). In addition, ground water is the only source of drinking water for many rural areas. As a result, appropriate protection measures must be put in place, a point currently recognized by water resource managers, decision makers, developers and planners all over the world. Protection measures are taken based on assessment of the likelihood of contaminants to reach groundwater resources, which in other words is called vulnerability assessment.

1.2 Assessing Ground Water Vulnerability

In general, ground water vulnerability assessments are aimed at determining the tendency or likelihood for contaminants to reach a specified position in the ground water system after introduction at some location above the uppermost aquifer (NRC 1993). Vulnerability assessments help evaluation of susceptibility to potential threats of pollution and identify corrective actions that can reduce or mitigate the risk of serious consequences from human activities on land. Vulnerability assessments combine the physical and chemical components of ground water (i.e., hydrogeologic setting) with indicators of the nature and extent of potential contaminant sources to determine the potential impact of these anthropogenic influences on the ground water quality (Hamerlinck and Arneson, 2003). Understanding the natural hydrogeologic and geochemical processes as well as the associated anthropogenic effects on a ground-water resource is required for complete scientific understanding of ground-water vulnerability (Focazio et al, 2002). Ground water vulnerability is an amorphous concept, not a measurable property. It is a probability (i.e., "the tendency or likelihood") of contamination occurring in the future, and thus must be inferred from surrogate information that is measurable. In this sense, a ground water vulnerability assessment is a predictive statement much like a weather forecast, but for processes that take place underground and over much longer time scales (NRC 1993).

The potential for contaminants to leach to ground water depends on many factors, including the composition of soils and geologic materials in the unsaturated zone, the depth to the water table, the groundwater recharge rate, and environmental factors influencing the potential for biodegradation. The composition of the unsaturated zone can greatly influence transformations and reactions. For example, high organic matter or clay content increases sorption and thus lessens the potential for contamination. The depth to the water table can be an important factor because short flow paths decrease the opportunity for sorption and biodegradation, thus increasing the potential for many contaminants to reach the ground water. Conversely, longer flow paths from land surface to the water table can lessen the potential for contamination for chemicals that adsorb or degrade along the flow path. Groundwater recharge rates affect the extent and rate of transport of contaminants through the saturated zone. Finally, environmental factors, such as temperature and water content, can significantly influence the degradation of contaminants by microbial transformations (NRC 1993).

There are certain general geologic and hydrologic factors that influence an aquifer's vulnerability to contamination as shown in Table 1, along with examples of features that lead to low or high vulnerability. Although these factors may look quite simple at first inspection, many of them interact in the environment to create more complex and subtle distinctions in vulnerability than the extreme situations in Table 1. In addition, many of these factors affecting vulnerability are highly variable and difficult to characterize over any given area.

Useful information can be gained by going through the process of assessing ground water vulnerability. Ground water vulnerability assessment is a potentially useful management concept for guiding decisions about ground water protection and thus requires the cooperative efforts of regulatory policy makers, natural resource managers, educators, and technical experts. Vulnerability assessments require the cooperative efforts of regulatory policy makers, natural resource managers, and technical experts. In performing vulnerability assessments these three groups are united by a common goal: the protection of ground water by the development and implementation of different management practices or policies, based on vulnerability to contamination, that minimize or prevent contamination of ground water resources.

Table 1. Principal geologic and hydrologic features that influence an aquifer's vulnerability to contamination (after Johnston, 1988).

Feature Determining Aquifer Vulnerability to Contamination

Low Vulnerability

High Vulnerability

A. Hydrogeologic Framework

Unsaturated Zone

Thick unsaturated zone, with high levels of clay and organic materials.

Thin unsaturated zone, with high levels of sand, gravel, limestone, or basalt of high permeability.

Confining Unit

Thick confining unit of clay or shale above aquifer.

No confining unit.

Aquifer Properties

Silty sandstone or shaley limestone of low permeability.

Cavernous limestone, sand and gravel, gravel, or basalt of high permeability.

B. Ground Water Flow System

Recharge Rate

Negligible recharge rate, as in arid regions.

Large recharge rate, as in humid regions.

Location within flow system (proximity to recharge or discharge area)

Located in the deep, sluggish part of a regional flow system.

Located within a recharge area or within the cone of depression of a pumped well.

(Source: NRC,1993)

1.3 Process of Ground Water Vulnerability Assessments

Assessing vulnerability is dynamic and iterative process requiring determination of the purpose of the assessment, followed by selection of a method, identification of the type, availability, and quality of data needed, performance of the actual assessment, and, finally, use of the information gained from the assessment process to make decisions on ground water resource management (NRC, 1993). The major components of vulnerability assessment (Figure 1.1) include: determining the purpose of the assessment; selecting an assessment method, dealing with issues of uncertainty and evaluation; identifying the needs, availability, and quality of data; and eventually using the completed assessment in managing ground water resources.

Figure 1.1The vulnerability assessment process (Source: NRC, 1993).

The first step in the process of vulnerability assessment is to identify the purpose of the assessment. As shown in the flowchart, an assessment's purpose is influenced by a variety of factors including the organization's ground water policy goal, technical considerations such as the form of the output and the cost of the assessment, and institutional issues such as the time frame for the assessment and resource availability. Purposes of vulnerability assessments are many range from improving information and education through analyzing the impact of alternative ground water policies, providing a tool for allocating resources, and guiding the decisions of land users or land use managers.

The next stage in the process of vulnerability assessment is to select a suitable approach for conducting the assessment. Various methods are available for vulnerability assessment. This stage of the assessment process includes choosing a model or technique for the assessment, identifying the uncertainties inherent in the model and the data needed for the assessment, and testing the model and its assumptions.

Considerations regarding the availability and quality of the data required are highly related to the performance of an assessment. These questions influence both the choice of technique for the assessment and the confidence of policy makers and regulators in making decisions based on the results.

Once an assessment is complete, various management actions are taken to protect ground water quality or minimize contamination. Management actions range from altering land use practices, targeting resource allocations, or disseminating vulnerability information through an educational program to collecting additional data on factors relating to vulnerability or ground water quality. Findings and recommendations on the use and improvement of vulnerability assessments and related research should also appear in the end of the process.

As the flowchart shows, the approach used to assess ground water vulnerability is central to the process, but is also directly affected by inputs or considerations entailed by the purpose, data availability, and management use of the assessment. The selection and development of a method for vulnerability assessment is not simply a question of appropriate science, but also reflects concerns over the need for the assessment, the availability of suitable data, the level of uncertainty in the model or the data, and the impact of this uncertainty on the management actions resulting from the assessment (NRC, 1993).

1.4 Key elements of vulnerability assessment

Key elements to consider in a vulnerability assessment for a particular application include the reference location, the degree of contaminant specificity, the contaminant pathways considered, and the time and spatial scales of the vulnerability assessment. The reference location is the position in the ground water system specified to be of interest. The ground water table is the reference location used in most existing techniques. However, managers may determine that another reference location is more useful for their purposes. Vulnerability assessments may or may not account for the different behavior of different contaminants in the environment. Thus, there are two general types of vulnerability assessments. The first addresses specific vulnerability, and is referenced to a specific contaminant, contaminant class, or human activity. The second addresses intrinsic vulnerability and is for vulnerability assessments that do not consider the attributes and behavior of specific contaminants. In practice, a clear distinction between intrinsic and specific vulnerability cannot always be made. Contaminants can enter aquifers by a variety of pathways. Most existing assessment techniques address only transport that occurs by simple percolation and ignore preferential flow paths such as biochannels, cracks, joints, and solution channels in the vadose zone. The omission of preferential flow paths is likely a significant limitation of vulnerability assessments in many environments. Some overlay and index methods have attempted to address contamination that might occur by wells and boreholes by mapping those features in combination with the results derived from other assessment methods. The overall utility of a vulnerability assessment is highly dependent on the scale at which it is conducted, the scale at which data are available, the scale used to display results, and the spatial resolution of mapping. The combination of these elements makes up a vulnerability assessment method.

1.5 Aquifer Sensitivity and Groundwater Vulnerability Methods

Numerous approaches or methods have been developed and used for assessing aquifer sensitivity and ground water vulnerability. They range from sophisticated models of the physical, chemical, and biological processes occurring in the vadose zone and ground water regime, to models that weight critical factors affecting vulnerability through either statistical methods or expert judgment. The National Research Council (1993) has classified these methods into three major classes: (1) overlay and index methods that combine specific physical characteristics that affect vulnerability, often giving a numerical score, (2) process-based methods consisting of mathematical models that approximate the behaviour of substances in the subsurface environment, and (3) statistical methods that draw associations with areas where contamination is known to have occurred (NRC, 1993). With the advent of Geographic information systems (GIS) implementation of some of these methods become much easier as it can act as computing environment for executing some types of assessments and for displaying the results of virtually all types of assessments.

1.5.1 Overlay and Index Methods

Overlay and index methods, are based on combining maps of various physiographic attributes (e.g., geology, soils, depth to water table) of the region by assigning a numerical index or score to each attribute. These relatively simple applications assign a numerical index or rating to mapped physiographic and anthropogenic attributes of a region. The ratings are then combined to generate a composite sensitivity/vulnerability rating. The ratings can be considered equally or weighted according to the relative magnitude of their influence in the overall assessment determination.

In the simplest of these methods, all attributes are assigned equal weights, with no judgment being made on their relative importance. Thus, in situations where simple confluence of the specified attributes occurs in a given area (e.g., sandy soils and shallow ground water) are deemed vulnerable. Such methods were the earliest to be used and are still favored by many state and local regulatory and planning agencies. There are overlay and index methods that attempt to be more quantitative by assigning different numerical scores and weights to the attributes in developing a range of vulnerability classes, which are then displayed on a map. Popularization of GIS technology has made it increasingly easy to adopt map overlay and index methods.

1.5.2 Process-based simulation methods

Process-based simulation methods (methods employing process-based simulation models), require analytical or numerical solutions to mathematical equations that represent coupled processes governing contaminant transport. Methods in this category are many and range from indices based on simple transport models to analytical solutions for one-dimensional transport of contaminants through the unsaturated zone to coupled, unsaturated-saturated, multiple phase, two- or three-dimensional models.

Process-based simulation model methods predict how long a contaminant will take to reach a given depth and/or the amount of contaminant by mathematically modeling the processes influencing contaminant fate and transport. The complexity of the models can range from simple transport model indices to multi-phase, multi-dimensional modeling of contaminant movement through saturated and unsaturated zones.

1.5.3 Empirical Statistical Methods

Statistical methods having a contaminant concentration or a probability of contamination as the dependent variable form the basis for the third category vulnerability assessment methods. These methods incorporate data on known areal contaminant distributions and calculate the probability of contamination by characterizing contamination potential for the specific geographic area using data from known contamination distribution in the area. Statistical methods are used by regulatory agencies that have the regional databases on ground water contamination needed to develop models. This method, too, is quite complex – especially if applied over a large area.

1.6 Uncertainties associated with vulnerability assessments

Vulnerability assessments will always be subject to uncertainties due to:

· Lack of data;

· Errors in data used;

· Incomplete understanding of the environmental processes;

· Errors in aggregating information;

· Errors inherent to statistical measures of association;

· The inclusion or exclusion of variables in most approaches is often arbitrary and based on expert opinion as to the weighting between factors.

· The indices are typically not based on observations or measurements of groundwater contamination and even when physically based models are used they are often prone to errors in model assumptions or in selecting the input parameters.

· The approaches or models are rarely validated or tested against observational data (Worrall et al., 2002).

Some of the uncertainties or errors can be measured while other cannot be measured.

1.7 GIS-based algorithms and approaches to vulnerability assessment

Geographical Information Systems (GIS) has been widely used in identifying the potential sources of groundwater contamination and the possible sites of pollution. It allows modelling of groundwater pollution and vulnerability and hence algorithms have been developed all over the world. A GIS system is well suited to carrying out analysis of spatially referenced layers of information. The ability to convert data layers into cell-based data sets facilitates the rapid achievement of a final result of a groundwater vulnerability assessment. It is relatively straightforward to compile vulnerability maps combined with other infrastructural information so as to be easily referenced and understood by planners and decision makers. Process based approaches and index based approaches of vulnerability assessment can be implemented in a GIS. For example many researchers all over the world are using the famous DRASTIC model (Aller et al., 1985) implemented in GIS software like ArcInfo (using the concept of overlay of raster layers and map calculation).

1.7.1 Main characteristics of GIS-based algorithms

Most groundwater originates as excess rainfall locally infiltrating the land surface, and thus activities at the land surface threaten groundwater quality. GIS-based algorithms (subjective or deterministic) have been used to model the spatial variability in land surface and sub-surface at various conflicting points, even in close proximity (Conrad, 2004). Instead of applying universal controls over land or soil use and effluent discharge to the ground, modeling the natural contaminant attenuation capacity of the strata overlying the saturated aquifer in GIS can be used in determining aquifer pollution vulnerability (Foster, 1998). Index (subjective) methods tend to be the most simple to apply and can be used over a range of spatial scales, but include subjective categorizations with uncertainties that cannot be quantified.

Properly designed hybrid methods that combine statistical and deterministic or process-based components and exclude subjective categorizations can provide insights on important processes controlling vulnerability over a range of spatial scales while maintaining objectivity and hence scientific defensibility. On the other hand, a subjective hybrid method that combines results of an objective model with a subjective categorization scheme to produce indexes of vulnerability would lose objectivity and ultimately may not be scientifically defensible

The soil zone can play a significant role in attenuating chemical concentrations, particularly through processes such as filtration; solution and precipitation; biochemical transformations and volatilisation. The soil zone thus needs to be included in groundwater vulnerability assessments due to likely presence of clay, organic contents and microbial populations, however the soil characteristics also influence the scale of nutrient and pesticide leaching from a given agricultural practice. However it must be noted that in many point sources of contamination the sub-surface contaminant load is applied below the soil zone at the base of excavations, such as pits, trenches, leaking underground tanks and quarries, and the attenuation capacity of the soil zone does not contribute to reducing the overall vulnerability. Detailed studies on the characteristics of the soil zone are being carried out by the University of Stellenbosch and a revised soil classification is being produced according the attenuation capacity of South African soils.

The intermediate zone or vadose zone is very important in assessing groundwater vulnerability. Not only is it strategically placed, between the ground surface and the saturated zone, it is most effective in pollution attenuation and even elimination as vadose zone water movement is typically slow and restricted to the smaller pores with larger specific surface.

In the intermediate zone there is significant potential for:

· Interception, sorption and elimination of pathogenic bacteria and viruses;

· Attenuation of heavy metals and other inorganic chemicals, through precipitation (as carbonates or hydroxides), sorption or ion exchange;

· Sorption and biodegradation of many hydrocarbon and synthetic organic compounds.

However, water movement in the intermediate zone is complex and its ability to attenuate pollution difficult to predict. Marked changes in the behaviour of some pollutants can occur if the polluting activity has sufficient organic or acidic loading to bring about significant change in the Eh or pH of the zone. In the case of persistent mobile pollutants the vadose zone merely introduces a large time delay before arrival at the saturated zone, without any significant attenuation occurring.

There are opposing schools of thought concerning the saturated zone and its role in vulnerability assessments: some consider the natural mobility and persistence of pollutants in the saturated zone should be considered. Others state that vulnerability assessments should only consider the zones between the ground surface and the saturated zone as important, the objective being to provide planners with the information that will ideally try and prevent pollution from even reaching the saturated zone.

In most situations the degree of contaminant attenuation will be largely dependent on the intermediate zone pollutant pathways and residence times. While natural flow rates for most porous formations do not exceed 0.2 m/d, when averaged over long time periods, in the presence of preferential pathways and fractured formations, flow rates may be more than an order of magnitude higher. Thus accelerated flow rates due to preferential flow paths need to be taken into account, particularly where microbial, biodegradable and readily retarded contaminants are concerned.

Scientists can provide water-resource decision makers scientifically defensible information for the assessment of groundwater vulnerability. To the extent that uncertainties in the assessment can be elucidated either quantitatively or qualitatively, the scientific defensibility and ultimate usefulness of the product will increase. Science objectives should be clearly distinguished from water-resource management objectives. Ultimately, successful groundwater vulnerability assessments blend scientifically defensible analyses used to meet science objectives with additional interpretations by water-resource decision makers to meet management or policy objectives (Conrad, 2004).

In the South African context, groundwater occurs predominantly in fractured aquifer settings, with a low percentage of groundwater occurring in primary porosity type settings. Thus it is important to take into account the degree of fracturing in assessing groundwater vulnerability. This is a difficult factor to quantify in an indexed GIS approach, due to the large number of associated variables and it can be accommodated for in such a method as a qualitative estimating, based on specialist opinion. Deterministic GIS-based algorithms may be better suited to calculate the effect of fracturing, should the extent of fracturing be quanitifiable.

1.7.2 Selected GIS-based approaches to vulnerability assessment

In the introductory session the two main different approaches to vulnerability assessment were briefly discussed. The “index or subjective rating method” is relatively easily addressed within a GIS framework. The cell-based layer approach facilitates the assignment of ratings and weights and rapid achievement of a final result of groundwater vulnerability. This approach also means that the algorithm can easily be repeated as new or more detailed data sets are obtained or if ratings and weightings need to be adjusted as a result of a sensitivity analysis for example. The most well known “index or subjective rating method” is the so-called DRASTIC method. This approach has been revised for the assessment of groundwater vulnerability within the South African context of hydrogeological conditions.

GIS has developed to be able to accommodate successful integration with process-based models that simulate physical processes in the environment. Assessments of pollution from organic compounds in urban environments were identified as one of the key research areas in UK and from such a study the GIS based UGIf model originated. UGIf is meant for urban recharge pollutant flux estimation, dealing with volatile organic compounds (BTEX) in urban environments (Thomas, 2001), so it was chosen as one of the GIS based algorithms for modeling pollutant concentrations and travel times of organic contaminants in urban environments. To make it suitable for South African conditions it was adapted to include various vulnerability assessment methods. It is being tested on the Cape Flats Aquifer, which is a primary aquifer.

DRASTIC is an index model capable of dealing with intrinsic vulnerability whereas UGIf is a process based model based on analytical approaches which can deal with contaminant specific vulnerability assessments.

1.8 Purpose of the Guideline Document

The Water Research Commission (WRC) has awarded a project to the CSIR and collaborating organizations, entitled “Improved methods for aquifer vulnerability assessments and protocols for producing vulnerability maps, taking into account information on soils”. This project is addressing the vulnerability in different levels such as the unsaturated zone (soils and regolith) and the saturated zone.

There are two cross-cutting components within this project, and these are:

(i) GIS based algorithms and Guidelines for vulnerability assessment

(ii) Decision Support Framework.

Two different GIS based approaches or algorithms viz. UGIf and ReSIS have been developed from this research for groundwater vulnerability assessment. This document particularly addresses the “Guidelines for developing and compiling groundwater vulnerability maps’ using these GIS based algorithms. The intention of these guidelines is to serve as guide to develop and compile groundwater vulnerability maps using these two methods.

1.9 Organisation of the Guidelines and Procedures

This document has four sections. The first section describes the concept of groundwater vulnerability, approaches or methods for assessment of vulnerability and the general characteristics of GIS based algorithms. The second section is devoted to describe the application and further modification of the UGIf model in developing and compiling groundwater vulnerability maps. The third section focuses on how to apply the ReSIS method for developing and compiling groundwater vulnerability maps. The fourth section includes the appendices which are instructions for running the UGIf model and its data preparation (a manual for the UGIf model) as well as steps to follow in running the ReSIS model.

1.10 References

1. Aller, L., Bennett, T., Lehr, J.H., and Petty, R.J., 1985. DRASTIC – A standardized system for evaluating groundwater pollution potential using hydrogeologic settings. U.S. Environmental Protection Agency Report EPA/600/2-85/018, 163 p.

2. Conrad, J.E. 2004. Literature review on GIS-based vulnerability assessment methods used to date and related data uncertainty and error propagation. Deliverable 1.2. WRC project “Improved methods for aquifer vulnerability assessments and protocols for producing vulnerability maps, taking into account information on soils” (K5/1432).

3. Foster, S.S.D., 1998. Groundwater recharge and pollution vulnerability of British aquifers: a critical overview. In: Robins N.S. (ed.) Groundwater Pollution, Aquifer recharge and Vulnerability. Geological Society, London Special Publications, 130, 7-22. http://books.nap.edu/books/0309047994/html/R1.html#pagetop

4. Haley, J. L., B. Hanson, C. Enfield, and J. Glass (1991) Evaluating the ef-fectiveness of groundwater extraction systems. Ground Water Monit.Rev., 11, 119-124.

5. Hamerlinck, J.D., and Arneson, C.S., eds., 1998, Wyoming ground water vulnerability assessment handbook: Volume 2. Assessing ground water vulnerability to pesticides: University of Wyoming, Laramie, Spatial Data and Visualization Center Publication SDVC 98-01-2, variable pagination.

6. Mackay, D. M., and J. A. Cherry. 1989. Groundwater contamination: pump-and-treat remediation. Environ. Sci. Technol. 23(6):630-636.

7. NRC, 1993. Ground Water Vulnerability Assessment: Predicting Relative Contamination Potential Under Conditions of Uncertainty. Committee for Assessing Ground Water Vulnerability, National Research Council. 224 pages, 6 x 9, 1993, ISBN 0-309-04799-4.

8. Thomas, Abraham, 2001. A Geographic Information System Methodology For Modelling Urban Groundwater Recharge And Pollution. Ph. D. Thesis. The School of Earth Sciences, The University of Birmingham, Birmingham, United Kingdom.

9. US EPA, 1998. Vulnerability ground water to contamination. http://www.epa.gov/seahome/groundwater/src/quality3c.htm

10. Worral F, Besien T and Kolpin DW (2002). Groundwater vulnerability: interactions of chemical and site properties. The Science of the Total Environment, 299, 131-143.

SECTION 22 GUIDELINES FOR DEVELOPING AND COMPILING GROUNDWATER VULNERABILITY MAPS USING UGIf MODEL

2.1 OVERVIEW OF UGIf MODEL

UGIf is a GIS based urban recharge pollutant flux model written in the Avenue programming language within ArcView GIS (ver. 3.x) and is primarily meant for the estimation of groundwater recharge pollutant fluxes of specific pollutants viz. BTEX (benzene, toluene, ethyl benzene and Xylene), nitrate and chloride to an urban unconfined, primary aquifer. UGIf accommodates the following processes (with their calculation method indicated in brackets):

· Infiltration, runoff and recharge (using the NRCS curve number method and water balance calculation based on estimates of evapotranspiration and soil moisture deficit for a given time)

· Interflow (also called lateral flow) of infiltrated water (empirical index approach);

· Volatilization of BTEX compounds in recharge water (Henry’s law);

· Sorption of BTEX compounds (distribution coefficient); and

· Degradation of BTEX compounds (first order decay).

The UGIf model was originally developed from a PhD research in United Kingdom (Thomas, 2001). Details of its implementation in ArcView GIS 3.2 for recharge and pollutant flux simulations and the applications of this model to the Birmingham (U.K.) unconfined aquifer can be found in Thomas et al. (2001; 2006); and Thomas and Tellam (2004; 2005).

2.1.1 Model structure

The basic structure of the UGIf model is shown in Figure 2.1. A land use / land cover classification allows the production of a land cover map. To each land cover class attributes are assigned which relate to permeability, runoff and water quality. With meteorological data and the land cover related runoff characteristics, an estimate of ‘potential recharge’ for each of the land use classes can be made, where potential recharge is defined here as ‘actual recharge’ (i.e. the water reaching the water table) plus interflow. An estimate of the potential mass flux (i.e. flux before interflow, evapotranspiration, and reaction) at the water table can also be made using runoff water quality data associated with the land cover classes. Interflow is estimated from geological maps and used to convert the potential recharge estimates into ‘actual recharge’.

The time taken to pass through the unsaturated zone can be estimated using the actual recharge estimates, hydraulic properties related to the geological units, (reversible) sorption properties related to each geological unit, and unsaturated zone thickness as calculated from land surface and water table maps. Using pollutant-related reaction properties and the estimates of time taken to pass through the unsaturated zone, the decay of degrading pollutants can be calculated, and hence the pollutant mass flux at the water table. Solute concentrations in recharge waters, corrected for evapotranspiration where necessary, are also calculated. A simplified flow chart of estimating pollutant mass fluxes reaching the water table is given in Figure 2.2.

Land Use

Classification

Leakage

Recharge

Source

Standard

Soil Moisture

Balance

Potential

Recharge from

e

ach Land Use

Class

Potential Recharge

Map

Land Use Map

Interflow

Indices

Drift

Classes

Drift Map

Drift

Reaction

Actual

Recharge

Chemical

Concentration

in each Landuse

Chemical

Data

Inter Flow

Indices

Drift

Reaction

Potential Mass

Flux

Poll

utant Mass

Flux From Drift

Mass Flux

In Recharge

Reaction

Term

*1

Rivers / Canals

Mains / Sewers

Septic Tanks

Landfill Leachate

Fuel Tank Spillage

*2

Horticulture

Industrial Landfills / Dumps

Domestic

Road

River / Canal

Sewer / Mains

Attribute Table Da

ta

and /or Calculation

Map

*2

*1

Figure 2.1Structure of UGIf model (drift = superficial deposits).

Figure 2.2Simplified flow chart for estimating pollutant mass fluxes at the water table.

2.1.2 Modified UGIf model

This model was revised further to make it more suitable for South African conditions and for groundwater vulnerability assessments using process based approaches. Apart from the travel time model for BTEX compounds, three new screening level models for vulnerability assessment and a simple approach to assessing intrinsic vulnerability of conservative contaminants are incorporated in the model. These models are: 1) the Attenuation Factor model of Rao et al. (1985), 2) the Leaching Potential Index Model of Meaks and Dean (1990) and 3) the Ranking Index Model of Britt et al. (1992).

Screening level models are relatively simple, easy-to-use, require very little input data and provide a management decision support. Their major areas of application are: 1) management of water resources (regional planning as related to groundwater control); 2) formulation and implementation of regulatory policies (zoning, land use alterations and practices that protect groundwater quality); 3) identification of “hot-spots” and selection of pollution abatement strategies; and 4) design and management of groundwater monitoring programs (Tim et al., 1996).

2.1.3 Various Menus available in UGIf

The various menus available in UGIf are:

· Direct Recharge

· Indirect Recharge

· BTEX NPS Pollution

· BTEX Petrol

· Groundwater Vulnerability Assessment

These menus are available in the View document of the project file (Figure 2.3). Figure 2.3 shows the interface of the UGIf model for the assessment of Non-Point Source (NPS) pollutant fluxes of BTEX.

2.1.4 Capabilities and Applicability

With respect to vulnerability assessments, the UGIf model can be categorized as a process based model using analytical equations and empirical approaches. Its main capabilities are prediction of pollutant concentration and mass fluxes exiting the vadose zone, pollutant transport velocity and travel time in the vadose zone for specific pollutants viz. BTEX. UGIf shows promise for use in providing input for regional groundwater solute transport models; in identifying gaps in knowledge and data; in determining which processes are the most important regarding urban groundwater quantity and quality; in evaluating existing recharge models; in planning, for example in investigating the effects of land use or climate change; and in assessing groundwater vulnerability.

Figure 2.3Interface of the UGIf model for assessing NPS BTEX pollutant fluxes.

2.2 Data Requirements and Outputs

The input data required for modeling direct recharge and non-point source (NPS) pollutant fluxes in runoff and recharge using UGIf are the following:

· Meteorological data (rainfall, evapotranspiration and soil moisture deficits);

· Land use/ land cover map

· Soil map or Hydrologic soil group map;

· Geological map with hydraulic and geochemical attributes;

· Event mean concentration (EMC) values for each land use type/class; and

· Topographic and water table depth data in grid form.

The hydraulic and geochemical attributes needed are the following:

porosity, bulk density, specific retention, presence of clay (clay index), horizontal and vertical hydraulic conductivity values, fraction of organic carbon content, half lives of BTEX compounds etc.

Standard model outputs include the following:

· Distribution of surface runoff;

· Cumulative infiltration;

· Potential recharge;

· Ground level slope;

· Interflow;

· Actual recharge;

· Pollutant fluxes in surface runoff;

· Travel times of each pollutant through the unsaturated zone; and

· Pollutant fluxes and concentrations at the water table.

2.2.1 Data Requirement for Vulnerability Assessments using Modified UGIf

For assessing vadose zone travel time and retarded velocity of BTEX compounds the model needs inputs of direct recharge, geology, vadose zone depth, hydrologic parameters such as soil texture, porosity, hydraulic conductivity, fraction of organic carbon content, and geochemical properties of BTEX compounds (half life period).

For vulnerability assessments using the three screening level algorithms of UGIf, it requires a combined grid containing attributes of average recharge rate (m/day), soil moisture or volumetric water content, vadose zone depths (m), and the retardation factor values. The input required for vulnerability assessment for a conservative contaminant is a combined grid containing attributes of average recharge rate (m/day), soil moisture or volumetric water content and vadose zone depths (m).

2.3 Limitations, Prediction Accuracy and Reliability

2.3.1 Limitations

The model takes into account the principal processes involved, and, as it is incorporated in a GIS, it allows the complexities of spatial heterogeneity to be investigated. A major limitation is the way time is being dealt with. It is assumed that land use and land use-related properties do not vary within the ‘time-slice’ or period being considered by the model. Daily recharge estimations are summed over the user-specified period. Within this period, steady-state conditions are assumed for the movement of water and solutes through the unsaturated zone. Thus, individual recharge pulses are not tracked: residence time in the unsaturated zone is calculated on the basis of the averaged recharge rate, but it is only used, with a delay arising from any sorption, to estimate degradation/decay of the pollutant concentration. Without incurring considerable computation times, it would be difficult to track individual recharge pulses simultaneously.

In Non-Point Source pollution assessment, all units of the same land use type are assumed to have the same Event Mean Concentration (EMC) value regardless of their spatial location within the urban area. However, in reality the concentration of pollutants in recharging water will vary depending on the soil type and vadose zone chemistry. Pollution is often related to past human activities, e.g. accumulation of waste material. Sometimes the background concentration in the subsurface may significantly contribute to the pollution. At present, this contribution is not accounted for in the model due to the often lacking information on the background chemistry and heterogeneity of the urban hydrogeological system. Another weakness is the lacking relationship between the water table elevation and recharge.

2.3.2 Prediction Accuracy and Reliability of Assessment

The different sub-models in UGIf make use of many input parameters (both spatial and non-spatial data) and the accuracy of their predictions is dependent on the assumptions made in each of the sub-models and the accuracy of the input data used.

Testing the UGIf model for a particular region in South Africa requires a variety of spatial and non-spatial inputs. Testing the UGIf model for a particular site, for example a portion of the Cape Flats area (e.g. Coastal Park area), requires the availability of local data. Scarcity of input data (e.g. land use information, evapotranspiration and soil moisture deficit data, hydraulic properties and geochemical parameters like fraction of organic carbon, etc.) limits testing and applicability of the model.

Environmental models are simplified representations of real systems, and uncertainty is always associated with their representations. In many cases the systems, especially urban groundwater systems, are heterogeneous, where a wide range of parameters with a wide range of possible values for them control the complex behavior of the system. In the case of recharge and solute transport simulations of urban environments, the hydraulic and transport parameters are generally not known in sufficient detail. Most of the input parameters in the present GIS based urban pollutant flux models are derived from literature values and therefore predictive runs and the results obtained from them are subject to much uncertainty in relation to the complex heterogeneity of the urban system being modelled. There may also be additional uncertainty relating to whether the conceptual model with simplified analytical equations is fully applicable to the field situation in an urban area (Thomas, 2001).

2.4 Vulnerability Assessment Using UGIf Model

UGIf currently estimates groundwater vulnerability of primary unconfined aquifer (from NPS BTEX pollution). For groundwater vulnerability assessment of primary unconfined aquifer from non-point source (NPS) BTEX pollution, the user has to open the project file (ugif_vulnerabilty.apr), input various data and run each program (submenus) in three main menus (viz. Direct Recharge, BTEX NPS Pollution and Groundwater Vulnerability Assessment).

2.4.1 Steps involved in vulnerability assessment for BTEX compounds

Various steps involved in vulnerability assessment for BTEX compounds using the UGIf model are the following:

1. Creation of input data folder (for example, C:\vulnerability) and copying of ArcView 3.x compatible format input data;

2. Creation of a working directory or folder (e.g., C:\ vulnerability\work) for storing subsequent GIS files created while running the model.

3. Opening of the ArcView GIS project file (ugif_vulnerabilty.apr).

4. Estimation of direct recharge (grid format)

5. Assessment of Non Point Source Pollution in Recharge (initial concentration)

6. Preparation of vadose zone depth map (grid format)

7. Preparation of geology grid map

8. Combining grids of recharge, BTEX concentration, geology and vadose zone depth

9. Assigning of soil textures, Clap and Hornberger constants, porosity, saturated hydraulic conductivity and calculation of vadose zone volumentric water content using the Clap and Hornberger methd

10. Calculation of soil-water partitioning coefficient values for BTEX

11. Assigning of bulk density and calculation of vadose zone retardation factors

12. Running of three programs of vulnerability assessments such as:

a. Calculation of vadose zone travel time of BTEX compounds (running of last program called ‘Vadose Zone BTEX Travel Time’ under the menu ‘BTEX NPS Polution’).

b. Running of the program for the assessment of groundwater vulnerability for conservative pollutants (first program under menu ‘Groundwater Vulnerability Assessment’).

c. Running of the program for the assessment of groundwater vulnerability for BTEX compounds (second program under menu ‘Groundwater Vulnerability Assessment’).

The instructions for opening the project file for vulnerability assessment (ugif_vulnerabilty.apr) and running of all the programs under various submenus of UGIf model form quite a lengthy document and one may find it tiring or boring to go through each step. These instructions are given as a manual form in Appendices 1 to 4.

The instructions for opening the project file (step 3) are described in Appendix 1. On opening the project file, a start up script will initiate by which the user will be asked to specify a data folder, a working folder and select input maps of land use and hydrologic soil group. After adding these two data layers the user can select the menu of Direct Recharge and run various programs in this menu which help in estimating recharge. The procedure for preparing the input data for recharge estimation and the instructions for running the programs under the Direct Recharge menu (step 4) are illustrated in Appendix 2.

After assessing direct recharge one can go for the other programs in the model (such as the menus of ‘BTEX NPS Pollution’ and’ Groundwater Vulnerability Assessment’). The modelling steps/procedure for the assessment of initial recharge pollutant fluxes (step 5) is given in Appendix 3.

A vadose zone depth map (step 6) can be generated from water table depth map and elevation grid. Use ‘Map Calculator’ available in Spatial Analyst extension to create a vadose zone depth map by subtracting the water table depth grid from the Elevation grid. A geology grid map (step 7) can be prepared from a shape file of geology using the Convert to Grid submenu. While converting to a grid one has to specify the same area extent and the same cell size of the other input maps.

Combining grids of recharge, BTEX concentration, geology and vadose zone depth (step 8) is done using the submenu ‘Combine Grids’ under BTEX NPS Pollution menu (Appendix 3). Before running this program the user has to first add these themes into a View and while running this program the user will be asked to select appropriate grid and its attribute field for combining. Steps 9, 10 and 11 are straightforward using the respective submenus under BTEX NPS Pollution menu.

The instructions for running vulnerability assessment programs (submenu under BTEX NPS pollution menu (viz. ‘Vadose Zone BTEX Travel Time’) and the two programs under the Menu Groundwater Vulnerability Assessment) are described in Appendix 4. The following section illustrates different types of vulnerability maps generated using these programs.

2.4.2 Preliminary results of the revised UGIf approach

The revised UGIf vulnerability algorithms were tested on Coastal Park data and preliminary results obtained for benzene are shown in Figures 2.4, 2.5 and 2.6. The predicted values are not the true representation as some of the model inputs, especially hydraulic property values are assumed values based on literature.

Comparison of the attenuation factor values and leaching potential index values shows that the results look more or less similar in their spatial distribution. The pools around the sewage treatment plant (influent type having a small leakage rate because of lining) have lower values in both maps indicating lesser vulnerability values or zones mainly because of lesser amount of recharge from these areas. The open ground/grass land areas have higher attenuation factor values and leaching potential index values indicating higher vulnerability in those areas because of higher recharge in these areas. The residential areas have medium vulnerability in both maps, which also corresponds to the medium recharge rates in these areas.

A comparison of the ranking index map with that of the attenuation factor map and the leaching potential index also reveals the same behaviour. Comparison of the travel time of a conservative contaminant (chloride; Figure 2.7) with the previous maps also indicates a more or less similar distribution of vulnerability. Preliminary results thus suggest that for all four predictions there is a consistent pattern of vulnerability of groundwater to organic pollution at the Coastal Park landfill site.

Figure 2.4Attenuation Factor Values for Benzene in Recharge Waters of Coastal Park.

Figure 2.5Leaching index potential values for Benzene in Recharge Waters of Coastal Park.

Figure 2.6Ranking Index for Benzene in Coastal Park.

Figure 2.7Distribution of travel time in years for a conservative contaminant.

2.5 References

1. Britt, J.K., Swinell, S.E. and McDowel, T.C. 1992. Matrix decision procedure to assess new pesticides based on relative groundwater leaching potential and chronic toxicity. Environmental Toxicology and Chemistry. Vol. 11, pp. 721-728.

2. Clapp, Roger B. and George M. Hornberger. 1978. Empirical equations for some soil hydraulic properties. Water Resources Research 14: 601-604.

3. Meeks, Y.J. and Dean, J.D. 1990. Evaluating ground water vulnerability to pesticides. Journal of Water Resources Planning and Management. Vol. 116. No. 5, pp. 693-707.

4. Rao, P.S.C., Hornsby, A.G. and, R.E. Jessup. 1985. Indices for ranking the potential for pesticide contamination of groundwater. Proceedings of the Soil and Crop Science Society of Florida, Vol. 44, pp.1-8.

5. Thomas, A. and Tellam, J.H. 2005. Modelling of Recharge and Pollutant Fluxes to Urban Groundwaters. Full paper accepted in Mar 2005 for publication in the special issue of the International Journal: The Science of the Total Environment.

6. Thomas, A. and Tellam, J.H. 2004. Development of an ArcView GIS Based Petrol Station BTEX Pollution Model for Assessing Groundwater Pollution from Small Scale Petrol Spills. Proceedings of 32nd International Geological Congress, Florence, Italy (Aug 20 - 28, 2004). Abstract Vol., Part 1, Abstract 98-16, p. 437.

7. Thomas, A. and Tellam, J.H. 2005. Development of A GIS Model For Assessing Groundwater Pollution From Small Scale Petrol Spills. Paper accepted for publishing in the Matthias Eiswirth Memorial Volume- the proceeding of the 32nd International Geological Congress held at Florence, Italy, August 20-28, 2004.

8. Thomas, A., C.S. Cheong, and Tellam, J.H. 2006. Development of a GIS Based Model for Assessment of Groundwater Contamination through Sewage Networks in Urban Environments. Paper accepted for publication in the International Journal of Pollution Research. Vol 1, 2006.

9. Thomas, A., Tellam, J.H. and Greswell, R. 2001. Development of a GIS Based Urban Groundwater Recharge Pollutant Flux Model. Proceedings of the Twenty-first Annual ESRI International User Conference, Individual paper presentation session (30 minutes slot). July 9-13, San Diego, California, USA. http://gis.esri.com/library/userconf/proc01/professional/papers/pap293/p293.htm

10. Thomas, Abraham., 2001. A Geographic Information System Methodology For Modelling Urban Groundwater Recharge And Pollution. Ph. D. Thesis. The School of Earth Sciences, The University of Birmingham, Birmingham, United Kingdom.

SECTION 33 GUIDELINES FOR DEVELOPING AND COMPILING VULNERABILITY MAPS USING THE ReSIS METHOD

3.1 Introduction

The first deliverable for the GIS component of the AVAP project was the “Development of GIS-based algorithms for vulnerability assessments”. The algorithm by the name of ReSIS was developed as part of this first deliverable. This report constitutes the second deliverable and focuses on “Guidelines for developing and compiling vulnerability maps”, with particular emphasis on the ReSIS method. For a full explanation of the ReSIS method, please refer to Deliverable 1.3 GIS-based Algorithms for vulnerability assessment methods (Conrad and Thomas, 2005). Only an abridged version of the method is presented here.

When developing a vulnerability map, related information that can be measured and mapped is used, since vulnerability itself cannot be measured. Combining information into one map of vulnerability requires modelling, statistics, case studies and the intuitive judgment of experts. The challenge then exists in that numerous data sets need to be compiled and a resultant data set presented, ideally with associated levels of certainty; in a manner that is relatively easily understood by the non-specialist decision maker.

Two main types of vulnerability assessments can be done:

· Intrinsic – focuses on hydrogeological settings and the natural protection provided by physical characteristics

· Specific – focus on the properties of specific contaminants and their behaviour in the sub-surface environment.

The type of vulnerability assessment then also has significant bearing on the final vulnerability map.

The starting point for any investigation is to consider all groundwater vulnerable. The approach chosen should always consider why the vulnerability map is needed, the scale of interest (i.e. local or regional, short-term or long-term) and how the final product is to be presented.

Groundwater vulnerability assessment is a dynamic process requiring the full cooperation of the relevant policy makers, resource managers and technical experts. For a successful assessment, clear objectives should be set early on. To ensure their needs are accommodated, the people who will use the completed assessment should be involved from the start and consulted regularly as the assessment proceeds – it makes little sense to use a complex method and model, the output of which can only be interpreted by technical experts. Because of the inevitable lack of precision, vulnerability assessments are best described as “approximations”. In most cases vulnerability assessment methods determine only the likelihood of the vertical movement of contamination in shallow groundwater. Lateral migration is an additional possibility that needs consideration. This type of information also needs to be conveyed with the final product.

This deliverable forms a subset of a more extensive deliverable on Vulnerability Map Guidelines, and focuses on how to apply the ReSIS method. Please note so far the deliverables have focussed on developing a robust methodology for vulnerability assessment and a specific tool has not been fully developed. The ReSIS method can be used outside of a GIS environment on point specific data, with vulnerability being calculated in a spreadsheet. To take spatial variability into account, a GIS based approach is recommended. Since the ReSIS method uses a rated and weighted layer approach, it can be GIS platform independent. For this project ArcView 3.2 scripts have been developed to test and refine the methodology. The final ReSIS product in ArcView, to be developed during the course of the project, can be distributed to users that have ArcView 3.2 and the Spatial Analyst extension. However, rigorous attention will have to be paid to data structures and data formats for the product to be run.

3.2 The ReSIS layer method

The ReSIS layer method is an adaptation of the well-known DRASTIC method (Aller et al., 1985,) in which the 7 layers are reduced to 3 significant layers corresponding to three impact zones, i.e. the

· soil zone (S),

· intermediate zone (I) (beneath the soil zone and above the saturated zone), and

· saturated zone (S).

A fourth component, the amount of rainfall converting into groundwater recharge, termed natural recharge (Re), has been added. ReSIS focuses on the vertical movement of contaminants.

The ReSIS model is based on a rated and weighted approach and provision is made for scalability of the data. An uncertainty factor is assigned to the vulnerability assessment which defaults to the rate of the coarsest data set used. For each of the input layers, the recommended input is field data collected by specialists. In the absence of sufficient field data, or in preparation for field visits in determining sampling sites, the model makes provision for the inclusion of coarser resolution data sets. National scale assessments are not recommended.

3.3 Elements of ReSIS layer model

Figure 3.1 describes the process in assessing groundwater vulnerability using the ReSIS model.

· Each layer (soil, intermediate and saturated zone) is processed in its entirety before combination of the layers

· The Vulnerability Rating (V RATE) is calculated for the Hydraulic and Chemical attenuation per soil (S), intermediate (I) and saturated (S) zone layer

· The Vulnerability Weighting (V WEIGHT) is then varied according to role played by each layer e.g. for underground tanks, soils weight will be low. The weighting is based on the relative importance of each of the three zones.

· A preliminary vulnerability rating is then obtained by applying a weight factor to each layer (VR PRELIM = (V RATE * V WEIGHT))

· The final vulnerability rating is obtained by applying a preferential flow rate multiplier (VR FINAL = VR PRELIM * PFM). This preferential flow multiplier depends on the degree of fracturing and preferential flow paths within the study area.

Figure 3.1Flowchart of ReSIS layer model

3.4 Required layers

The ReSIS model requires input data for the following layers:

· Natural recharge (Re)

· Soil zone (S)

· Intermediate zone (I)

· Saturated zone (S).

The requirements for each of these layers is described in detail in Conrad and Thomas (2005).

3.5 Assigning a vulnerability rating using the Zone Vulnerability Matrix

The Vulnerability Rating (V RATE) is calculated for the Hydraulic and Chemical attenuation per soil, intermediate and saturated zone layer using the Zone Vulnerability Matrix (ZVM). In the ZVM, a Hydraulic Vulnerability Rating (HVR) and a Chemical Vulnerability Rating (CVR) are assigned per zone/layer (soil, intermediate or saturated zone), regardless of the scale of the input data.

Specialist input is required to create the ZVM for each layer. Even though these ratings appear to be subjective, they are based on physical characteristics associated with the hydraulic attenuation or the chemical attenuation of the particular zone under investigation. The HVR gives an indication of the intrinsic vulnerability of the particular zone. The CVR gives an indication of the specific vulnerability of the particular zone. By combining the HVR and CVR, the Vulnerability Rating for a zone can be derived.

3.6 Deriving a vulnerability weighting

The Vulnerability Weighting (V WEIGHT) is varied according to the role played by the zone e.g. for underground tanks, soils weight will be low. The weighting is based on the relative importance of each of the three zones combined with the thickness of the zone. A thick zone implies a higher weight, while a thin zone implies a lower weight. Assigning a weight to each of the zones is currently a subjective process, based on specialist judgement.

3.7 Intrinsic vulnerability

The preliminary intrinsic vulnerability rating is obtained for each zone by applying the weight and rate factors described before and then combining all four zones. Accelerated flow rates due to preferential flow paths need to be taken into account. A preferential flow rate multiplier expressed as a percentage, is applied to the preliminary intrinsic vulnerability rating.

3.8 Quantitative assessment of uncertainty

Due to the fact that the ratings and weights assigned for the ReSIS model remain subjective, it is not possible to calculate the uncertainty of data and accuracy of the model quantitatively. A qualitative assessment can however be made based on the scale of the data and the experience of the specialists assigning HVR and CVR ratings. Qualitatively, the level uncertainty and accuracy will be associated with the coarsest dataset. A cumulative system can be devised to address the compounded effect of adding the four layers together.

3.9 Conclusion

The ReSIS method follows the approach of rating and weighting three component layers; i.e. the soil zone, the intermediate zone (beneath the soil zone and above the saturated zone) and the saturated zone. The method can be applied to both intrinsic and specific vulnerability assessments. The weightings are based on the relative importance of the various zones. It takes into account preferential flow conditions. The method is scaleable and can be used with field data and if this is not available, coarser data sets can be used. The level of uncertainty associated with the final results is also assigned, thereby assisting decision makers.

Future work on the ReSIS model will include:

· Accounting for lateral migration of contaminants.

· Determining a reliable weighting formula.

· Automating scalability of input data sets in GIS.

· Validation and sensitivity testing.

3.10 References

1. Aller, L., Bennett, T., Lehr, J.H., and Petty, R.J., 1985. DRASTIC – A standardized system for evaluating groundwater pollution potential using hydrogeologic settings. U.S. Environmental Protection Agency Report EPA/600/2-85/018, 163 p.

2. Conrad, J.E., and Thomas, A., 2005. GIS-based Algorithms for vulnerability assessment methods. Deliverable 1.3. WRC project “Improved methods for aquifer vulnerability assessments and protocols for producing vulnerability maps, taking into account information on soils” (K5/1432).

3. Vegter, J.R., 1995. An explanation of a set of national groundwater maps. (WRC report no. TT74/95). Pretoria: Water Research Commission.

SECTION 4 - APPENDICESAppendix 1

4 Opening of ArcView Project for Vulnerability Assessment

4.1 Preliminary steps

The preliminary steps involved in opening UGIf model are as follows:

1. Creation of input data folder (for example, C:\vulnerability) and copying of ArcView 3.x compatible format input data such as land use and soil hydrologic soil group;

2. Creation of a working directory or folder (e.g., C:\ vulnerability\work) for storing subsequent GIS files created while running the model.

3. Opening of the ArcView GIS project file (ugif_vulnerabilty.apr), browsing to a directory of input data, and inputting of land use and hydrologic soil group maps;

4.2 Opening of ArcView Project file of UGIf model

After copying land use map and hydrologic soil group or soil map to a specified location (for example, C:\vulnerability), open ArcView Project file (e.g., ‘ugif_vulnerabilty.apr’). On opening the project, a start up scrip is executed which actually assist in specifying the data and working directories and browse to the specified folder for adding the input data. The user gets a message window stating that this program needs ArcView Spatial Analyst extension. The start up scrip will look for whether the Spatial Analyst extension is installed or not, and if this extension is installed in the user’s PC, then the user gets the following message window:

Click OK.

Now the start up scrip will ask to enter your data file directory or folder for your input data. The user gets the following window interface.

In the text window specify (or type in) the path of your data folder (e.g., c:\vulnerability\inputdata). After entering the data path click OK.

If the specified path is correct then the user gets a window to specify the working directory.

Enter your working directory path, for example c:\vulnerability\work.

If the path for your data directory and the working directory is right, then the start up script will prompt to browse to the data folder and select the input maps of land use and hydrologic soil group.

Specify the ‘Data Source Types’ which could be shape file or a grid data source. Select the land use map and hydrologic soil group map and click OK. On adding the input data files the start up script creates a View which will have the following appearance with various menus of the UGIf model.

If you do not select any data for adding to the project the following warning message appears, and the user can add the later.

On clicking OK, a blank view is created in the project which may have the following interface”

If the specified directory for input data and /or work directory (e.g., c:\ava) does not exist in the PC, an error message appears stating that the specified path is not a directory and the start up scrip quits without completing its remaining program to browse to the specified folder and adding the input data.

In such a case the project file is opened which has the following look.

Now the user can create a new View by clicking on the ‘New’ button and add a land use and a hydrologic soil group map and proceed either to prepare the input map for modeling or use other menus in the model.

Appendix 2

5 Procedure for Running the Direct Recharge Model

The Direct Recharge model has certain menus through which the user has to go through in order get estimates of direct recharge. The direct recharge calculation is done using grid input maps of land use and hydrologic soil group.

The preliminary steps involved for running this model are:

1. Preparation of land use map and hydrologic soil group map in grid file format acceptable to the model;

2. Preparation of input data tables such as curve number (text file or dbf file) and meteorological data (text file) in a specified format needed for the model;

Once these input data are ready one can directly go to the ‘Direct Recharge’ model for estimating recharge through its various menus.

The steps or submenus involved in direct recharge estimation are the following:

1. Preparation of a curve number map (submenus 1 to 4)

2. Estimation of runoff and potential recharge (submenu 9 / submenu 10 / submenu 11)

3. Estimation of interflow (submenus 12 to 17)

4. Estimation of actual direct recharge (submenu 19)

5.1 Preparation of land use map in grid format acceptable to the model

The runoff-recharge model needs land use data in grid format having predetermined land use classes represented with code values as shown in Tables 1 or 2. The land use grid map preparation is achieved through the sub menu ‘Land Use Grid Map Preparation’ under the ‘Direct Recharge’ menu. This program can accept either shape file or grid file format land use map. The shape file format land use data should have one attribute field either ‘Type’ or ‘Luse_code’. The ‘Type’ field should have the specified land use classes as shown in Tables 1 or 2. If field ‘Luse_code’ is not present in its attribute table, the land use classification script/program will create such a field and write the code values based on the text values available in its field ‘Type’. Finally, the land use classification script will convert the shape file format data into a grid map having the code values and land use types as shown in Table 1.

Table 1. Typical land use codes and land use types acceptable to UGIf model.

Luse_code

Type

1000

Commercial/Business

2000

Industrial

3000

High Density Residential

4000

Medium Density Residential

5000

Low Density Residential

6000

Car Park

7000

Transportation

8000

Recreation Ground (Grass)

9000

Agricultural/Horticultural/Farm

10000

Woodland/Shrub

11000

Cemetery/Graveyard

12000

Open Ground/Grassland

13000

Reservoir/Lake/Pond

14000

River

15000

Canal

16000

Freeway/Motorway

17000

A Road

18000

B Road

19000

Minor Road

20000

Railway

21000

Pavement Asphaltic

22000

Pavement Concrete

23000

Pavement Brick

24000

Recreation Ground (Hard Surface)

25000

Industrial Light

26000

Industrial Medium

27000

Industrial Heavy

28000

Bare Land

29000

Shrub/Scrub Land

30000

Forest/Woodland

31000

Nature Conservation

32000

Canal/Open Channel (Lined)

33000

Canal/Open Channel (Unlined)

34000

Dam/Reservoir (Lined)

35000

Dam/Reservoir (Unlined)

36000

Lake/Pond/Vlei/Wetland

37000

River/Drain/Open Channel (Gaining)

38000

River/Stream/Open Channel (Loosing)

39000

Effluent Pond

40000

Landfill/Waste Dumping

41000

Major Road/Arterial

42000

Derelict Built Up Land ('Brown Fields')

43000

Park/Garden

44000

Golf Course

45000

Play Ground/Sports Fields

The model can also handle land use map having the classes as shown in Table 2 and a grid map having the classes as shown in Table 1 will be generated. If the land use map has a ‘luse_code’ field, the program will add a ‘luse_type’ to its attribute table and will write the above land use classes based on the code given in the program. If the map has land use types as listed in Table 2, the corresponding code in Table 1 will be written to its attribute table.

Table 2. Acceptable land use types and their code values for use in UGIf model.

Luse_code

Type

1000

Commercial

1000

Commercial/Business

2000

Industrial

3000

High Density Residential

4000

Medium Density Residential

5000

Low Density Residential

6000

Car Parks

6000

Car Park

6000

Parking

7000

Transportation

8000

Recreation Ground

8000

Recreation Ground (Grass)

8000

Pasture/Park/Play Ground (Grass)

8000

Pasture

8000

Park

8000

Play Ground (Grass)

9000

Agriculture

9000

Agricultural Farm

9000

Agricultural Field

9000

Allottment Garden

9000

Horticulture

9000

Horticultural Farm

9000

Horticultural Field

9000

Agricultural/Horticultural/Farm

10000

Woodland/Shrub

10000

Woodland

10000

Shrub

11000

Cemetery/Graveyard

11000

Cemetery

11000

Graveyard

12000

Open Ground/Grassland

12000

Open Ground/Heath

12000

Open Ground

13000

Reservoir/Lake/Pond

13000

Reservoir

13000

Lake

13000

Pond

13000

Lake/Pond

14000

River

15000

Canal

16000

Motorway

16000

Freeway/Motorway

16000

Roads

16000

Road

17000

A Road

18000

B Road

19000

Minor Road

20000

Railway

21000

Pavement Asphaltic

22000

Pavement Concrete

23000

Pavement Brick

24000

Recreation Ground (Hard Ground)

25000

Industrial Light

26000

Industrial Medium

27000

Industrial Heavy

28000

Bare Land

29000

Shrub/Scrub Land

30000

Forest/Woodland

31000

Nature Conservation

32000

Canal/Open Channel (Lined)

33000

Canal/Open Channel (Unlined)

34000

Dam/Reservoir (Lined)

35000

Dam/Reservoir (Unlined)

36000

Lake/Pond/Vlei/Wetland

36000

Vlei/Wetland

36000

Wetland

36000

Vlei

37000

River/Drain/Open Channel (Gaining)

38000

River/Stream/Open Channel (Loosing)

39000

Effluent Pond

40000

Landfill/Waste Dumping

41000

Major Road/Arterial

42000

Derelict Built Up Land ('Brown Fields')

42000

Brown Fields

43000

Park/Garden

44000

Golf Course

45000

Play Ground/Sports Fields

If the above mentioned land use types are not present in the added land use map the user has to edit its attribute table and make a field of ‘type’ (character type) and reclassify the features based on the classes shown in Table 1. Alternatively a field of ‘luse_code’ (number type) can be made and the code values as shown in Table 1 can be given to each type of features present in the map.

5.2 Land Use Grid Map Preparation

On clicking on the sub menu ‘Land Use Grid Map Preparation’ the user gets the following interface to select the land use map from the themes available in its View.

Select land use theme and click OK. The user will be asked to give a grid file name (default is ‘land_use’).

Brows to a path for storing the grid file and write its preferred file name and click OK.

When land use grid map preparation is over the user get the following message window:

Click OK.

The newly generated land use grid map will be added and displayed and the project’s View interface will be maximized to its full view on the screen so that all menus are visible in full.

The added land use grid theme will have an attribute table as shown below, which shows that all grid cells are appropriately coded based on the land use types available.

Examine the land use codes available in the table displayed in the above interface and note down the available land use codes. These codes are needed for preparing an input table NRCS curve number values in a text file format.

If a shape file is selected while running the program under Land Use Grid Map Preparation’ sub menu, the user has to specify the grid extent and output grid cell size. The default option for the ‘Output Grid Extent’ is ‘Same as Display’ and an appropriate grid cell size will be shown for selection.

Select the option of ‘Same as the selected shape file’s extent’ and select a preferred grid cell size. The grid cell size controls the resolution of the map and select an appropriate grid cell size (for example 10 m or more) in which all narrow features such as railway lines or minor road network are properly represented in the output grid for a regional scale study on 1: 50000 scale. If a very small grid cell size is selected the grid map will have a large file size and subsequent analysis (especially using combining of grids) in the model will not be faster.

A grid map will be generated by this program and it will be added to the View and will be displayed based on the land use codes generated.

The example below shows an unprocessed land use map in shape file format having an attribute table with different types of land uses under its field ‘Type’. The land use types are not exactly matching with the types shown in Table 2.

The grid map generated from this program has only two land use types (commercial and industrial) acceptable to its code as shown in the following diagram and the other land use types are represented with a code value of zero. This gird map is not a good input as most area is not represented with appropriate code values.

5.3 Preparation of hydrologic soil group map

The runoff-recharge model needs hydrologic soil group map in a specified format (either shape file format or in grid format) having hydrologic soil group classes represented with code values as shown in Table 3.

Table 3. Code values for hydrologic soil groups.

Hydrologic Soil Group

(Soil_HSG)

HSG Code

(Recls_Code)

Soil Texture

A

1

Sand, loamy sand, and sandy loam

B

2

Silt loam and loam

C

3

Sandy clay loam

D

4

Clay loam, silty clay loam, sandy clay, silty clay, and clay

The hydrologic soil group map can be prepared from a soil map or a geological map and based on the texture or lithology of the features present. These units can be classified to generate hydrologic soil group classes as shown in Table 3.

The second sub menu entitled ‘Assigns Hydrologic Soil Group in Soil / Geology Map’ (available under the Direct Recharge menu) helps to assign hydrologic soil group values in the soil map.

Click OK. The user will be asked to later select a soil theme.

The user has to select a texture or lithology field based on which the subsequent interfaces allows assigning the HSG values as A, B, C, and D.

Based on the texture description available in the map the user can enter HSG values A, B, C, and D.

Click OK. The attribute table will be queried for each type of texture description and the corresponding values of HSG will be assigned and finally the user gets the following message stating that HSG names are assigned.

The above prepared data has to be converted into a grid map, which will be combined with the land use grid for use in the direct recharge calculation.

5.4 Hydrologic Soil Group Grid Preparation

The third menu entitled ‘Hydrologic Soil Group Grid Preparation’ automatically converts the above shape file or the grid file into a grid file having HSG code values in numeric form (1,2,3 and 4) in a field named’ Recls_code’. On selecting this menu the user has to input the soil theme.

After selecting the soil theme the user has select the HSG field.

Now the user will be asked to specify a grid name and the location wherein to store the converted grid map (this step is similar to the land use grid preparation).

Specify the same extent and cell size of the land use map.

Finally the user gets a message which states that the soil theme is ready for use in subsequent programs.

The soil grid map will be added to the View and displayed.

On opening the attribute table of hydrologic soil group map it may look like this:

5.5 Combine Grids and Assign NRCS Curve Numbers

The next step is to combine the two input maps (land use grid and soil grid) using the next submenu ‘Combine Grids and Assign NRCS Curve Numbers’. While running this program the user has to input a table of land use codes represented by the underlying HSG type and the corresponding Curve Numbers (CN) in a tab limited text file format. This input table should have two fields viz. ‘LU_SOIL_CD’ and ‘CN’. The last digit in the values of LU_SOIL_CD’ field or column represents the hydrologic soil group code values (1 or 2 or 3 or 4). If land use code is 1000 and it is underlain by HSG value A (which means HSG code 1), then the corresponding code value for this column is 1000 + 1 = 1001. The same style has to be used for all other land use types underlain by different HSG types.

If the study area has all four HSG types for a particular land use type (e.g., commercial/business area represented as 1000), then the input CN table should have four land use codes for this land use type such as 1001, 1002, 1003 and 1004 and corresponding CN values also. This file can be generated in Excel spreadsheet having these two columns and save as Tab limited text file. Before saving the columns and its values should be selected in Excel spreadsheet, so that there will not be any blank rows in the saved file.

Table 4. Example Input table of CN values for a set of land use classes / codes having four different HSG conditions.

LU_SOIL_CD, CN

1001,89

1002,92

1003,94

1004,95

2001,89

2002,92

2003,94

2004,95

3001,85

3002,91

3003,93

3004,95

4001,81

4002,87

4003,91

4004,93