ENVIRONMENTAL DECISION SUPPORT MODELING FOR THE RED SEA COASTAL ZONE, EGYPT

 

H. I. El-Gamily*, S. Nasr** And M. El-Raey**

*National Authority for Remote Sensing and Space Sciences (NARSS)

 **Institute of Graduate Studies and Research, University of Alexandria, Egypt.

hgamily@idsc.gov.eg

ABSTRACT:

The Red Sea coastal zone is a unique and sensitive coastal environment. It is characterized by a diversity of natural resources such as coral reefs, seagrasses,  mangroves, sandy beaches and many wildlife. Hurghada, as well as Ras Abu Soma is threatened by numerous anthropogenic activities. The diving and snorkeling in addition to other touristic activities and oil industry, are the main sources of danger to these areas. The adverse impact of the tourist activities and potential oil pollution, in addition to municipal disposal, may damage ecosystems in this fragile environment of the Red Sea coastal zone.

The main objective of this study is to build GIS-based environmental models, to be used in the decision-making processes of managing and protecting the coastal zone. In this study, two models are introduced. The first environmental sensitivity model is designed to set systematic procedures for mapping environmental sensitivity of shoreline and foreshore natural resources and habitats. The second one is an analytical tool for determining the most suitable sites, for building new tourism villages and hotels. Thus, the power of GIS, as an analytical tool, is used to handle some environmental issues, related to the coastal zone, in the form of cartographic models.

KEYWORDS:

Environmental Sensitivity Index Map, Cartographic Model, Spatial Decision Support System (SDSS), Spatial Moduler

BACKGROUND

The Red Sea coastal zone is considered to be one of the unique and most sensitive coastal environments. It is characterized by a diversity of natural resources such as coral reefs, seagrasses, turtles, fish, birds,  mangroves, sandy beaches and many  wildlife species, as well as clear water and sky (IUCN/UNEP, 1988). Hurghada, as well as Ras Abu Soma is threatened by numerous anthropogenic activities. The diving and snorkeling in addition to  other touristic activities and  oil industry, are the main sources of danger to these areas. The adverse impact of the tourist activities and potential oil pollution, in addition to municipal disposal, may damage ecosystems in this fragile environment of the Red Sea coastal zone. The major sources of danger, throughout the Red Sea coastal zone can be summarized as follow:

(1) different touristic activities; (2) the landfilling activities; (3) potential oil and phosphate pollution;  (4) municipal disposal; and (5) dynamite fishing.

The main objective of this study is to build GIS-based environmental models, to be used in the decision making processes of managing and protecting the coastal zone of study area. The power of GIS, as an analytical tool, is used to handle some environmental issues, related to the coastal zone, in the form of cartographic models.

The study area, shown in Figure (1), is located in the northern western part of the Red Sea coastal zone. It lies between 33° 42˘ to 34° 01˘ E, and 26° 48˘ to 27° 18˘ N. The study area includes two sectors, Hurghada and Ras Abu Soma. The northern sector, Hurghada, is already developed without any sustainable planning. The southern sector, Ras Abu Soma, is the most promising one for touristic development.

The word “model” is generally used for many things. In GIS, there are two common uses. The first is the data model, which deals with the scheme of  organizing data about real world. Recently, the word “model” means the symbolic representation of the relationships between spatial objects and their attributes. Thus, modeling is a part of the analytical process of discovering, describing, and predicting spatial phenomena (Bonham-Carter, 1994) [s1] .

There are two different approaches, to model the environmental processes. The first is the cartographical modeling, that works within the GIS. The other is the mathematical modeling, which uses the GIS as a cooperating technology (Kemp, 1997). The cartographic modeling is a general methodology for analysis and synthesis of geographical data (Tomlin, 1990 and Berry, 1993).

There are three levels of integration between GIS and environmental models. The lowest level involves a simple exchange of files. The next higher level has an interface program, which manages the file format conversions. At the highest level of  integration, the model becomes one of the analytical functions inside the GIS, or the GIS is an option in the file management and output components of the model (Fedra, 1993).

Dangermond (1993), concluded that  the GIS-based environmental models will become a part of the decision support systems, for  environmental managers and decision makers. Also, models will be used to predict, both the consequences for the environment of human activities, and the impacts on humans of environmental processes. The power of GIS for any ecological research lies in its ability to analyze spatially distributed data, whether the GIS is used alone or linked with more sophisticated models (Johnston, 1993). Goodchild (1993), mentioned that the display of any environmental model results, in map form, using GIS can indeed move environmental modeling and policy formulation closer together.

2. METHODOLOGY

Several techniques of modeling are used to carry out the above mentioned task.

2-1. Procedures Of Modeling:

A model can be defined as any representation of reality (Kemp, 1997). Two models will be discussed:

(a) Environmental sensitivity model.

(b) Site suitability model.

These models can be described as cartographic decision support models (Karen Kemp, Personal Communication, 1997).

Environmental Sensitivity Model:

Red Sea coastal zone has sensitive natural resources and habitats. This model is designed to set systematic procedures for mapping environmental sensitivity index of the shoreline and foreshore natural resources and habitats. Based on Pavasovic (1993), the following issues will be taken into consideration, during performing of the environmental sensitivity and site suitability models:

1.        Selection of criteria used in modeling.

2.        Determination of weights for used criteria.

3.        Representation of criteria in geographic layers.

According to Abdel Kader et al. (1997), the following criteria are used to rate the shoreline sensitivity, for the potential oil spills:

1.        Nature of the shoreline: The sensitivity of the shoreline increases by increasing the permeability of its sediments.

2.        Slope: Steep shoreline is less sensitive than flat one.

3.        The grain size: The shoreline sensitivity increases by increasing the grain size of shore sediments.

4.        Exposed and sheltered shoreline: Sheltered shoreline is more sensitive than the exposed one.

5.        Cleaning and restoration: The shoreline sensitivity increases by decreasing the possibility and capability of cleaning and restoration.

Hence, the following criteria are used to determine the degree of sensitivity of the foreshore natural resources and habitats, for the potential oil spills, as well as the other anthropogenic activities:

1.        The speed and extend of damage.

2.        Duration of damage.

3.        Possibility of restoration and preservation of the natural resources and habitats.

4.        The rate of growth of the natural resource.

5.        To what extent, other species in the community are dependent on the exposed natural resource.

In addition to the shoreline and foreshore sensitivity criteria, the following criteria are taken into consideration, to produce the final environmental sensitivity index map. These criteria can be summarized as follow:

1.        Cultural and social value;

2.        Economic and recreational value;

3.        Scientific value; and

4.        Environmental considerations.

b- Site Suitability Model:

The site suitability model is an analytical tool for determining most suitable sites, within the study area, for building a tourism villages and hotels. The interview with native individuals, environmental specialists in the coastal zone management, and consultation of published papers, are main sources of determinative criteria used in this model.  The following criteria are chosen to determine optimal sites, for building new tourist villages and hotels:

1.        Slope of the shoreline area.

2.        Distance from the main roads.

3.        Distance from the existing urban communities.

4.        Elevation of the main land.

5.        Distance from the shoreline sensitivity classes.

6.        Area of the selected sites.

7.        Distance from the boarders of main wadi.

8.        Distance from the main faults.

Data Description:

The following data is used, as input data, in the environmental sensitivity model.

Shoreline  morphology.

The morphological map of the study area (Hurghada and Ras Abu Soma), is produced based on the following:

·         Satellite image interpretation; and

·         Field surveying.

The morphological maps are converted into a digital vector format, and transformed into UTM projection. These maps are converted into raster format and entered to the database.

Slope image.

The contour lines coverages (layers), of the study areas, are manipulated, to produce the Digital Elevation Model (DEM). The DEM is used to produce the slope images. The slope image, inside a specific buffer from the shoreline, is entered in the model.

 

 

Exposed and sheltered shorelines.

The same buffer, which is used in the slope image, is used with the shoreline coverages of the base maps. These coverages are converted into raster format. The raster files are classified into sheltered, semi-sheltered and exposed classes.

Land-use/ land-cover map.

The land-use/ land-cover maps of 1984, for Hurghada and Ras Abu Soma sectors, are used in this model.

Foreshore natural resources and habitats.

The above mentioned land-use/ land-cover map is used to produce the foreshore natural resources and habitats layer. All land-cover classes are recoded into the following ones; (a) intertidal flat; (b) shallow coral reef; and  (c) mangrove forests and seagrasses.

Oil sensitive wildlife layer.

The oil sensitive wildlife species are distributed  geographically on the base map. Mammals, fish, reptiles and birds are the main wildlife categories in this layer. Field observations, as well as the search on papers and reports (Barratt, 1982, Heathcoat et al., 1984, Mohamed and Abul-Azm, 1992 and CRI, 1992), are the main sources of the wildlife data.

The following is used, as input data, in the site suitability model:

1.        Digital Elevation Model (DEM).

The digital elevation model of the study area is created from the available topographic maps. It is used as a source for the elevation values of the main land.

2.        The slope layer of the main land.

The slope layer is extracted from the DEM of the study area. This layer is classified into two classes, suitable and unsuitable. The class is suitable, when the slope is less than 5 degrees, and it is unsuitable, when the slope is greater than 5 degrees.

3.        The main roads layer.

This layer is extracted from topographic maps of the study area. Then, it is converted into raster format, before its use in this model.

4.        Existing urban communities layer.

The land-cover / land-use map extracted from the LANDSAT-TM scene of 1984, is recoded to produce the existing urban communities layer. The extracted layer is in raster format. The existing urban communities of 1984 is suitable to present the urban areas before the unplanned tourism rush in the Red Sea. Thus, it will enable us later on, to check the compatibility of new urbanization after 1984, with the site suitability criteria.

5.        The shoreline sensitivity layer.

This layer is one of the most important layers, required for the site suitability model. It is produced from the environmental sensitivity model. The shoreline sensitivity layer is recoded to produce three categories of sensitivities, low (up to 30%), medium ( up to 60%), and high ( more than 60%).

6.        The main wadi layer.

This layer is extracted, by recoding the classified images of the study area, by recoding the wadi to value one and others to zero.

7.        The main faults layer.

The main faults is digitized from the geological maps of the study area. Then, it is converted into raster format.

Model building:

The environmental sensitivity model aims to produce the environmental sensitivity index map, to support decision makers and planners in the sustainable development process, for environmentally sensitive areas. It is created using the Spatial Modeler Language (SML) of ERDAS IMAGINE  spatial modeler. The model maker is a unique graphical editor for creating GIS and remote sensing models, by using a palette of tools to place icons on a plank page. It is a component of spatial modeler module.

Figure (2), illustrates the flow chart of the environmental sensitivity model. In this model, the  environmental  sensitivity index map is produced  by  overlaying  these main layers. These layers are; the shoreline sensitivity; foreshore sensitivity; and oil sensitive wildlife.

The shoreline sensitivity map is produced by applying the sensitivity criteria, respectively on the shoreline geomorphology layer, as well as the slope image and the shoreline exposure layer (Figure 2). The morphological units of the shoreline is experienced to the different shoreline sensitivity   criteria, to  produce    various   shoreline   sensitivity  layers. 

Table (1) illustrates weighting factors for the shoreline sensitivity criteria, which were used to rank the sensitivity of various morphological units of the shoreline. The resultant layers are added, to produce morphologic units sensitivity layer of the shoreline. The slope image and the exposure nature, of the shoreline layer, are used, to produce the slope and exposure sensitivity layers. These two layers are multiplied  to produce the slope/ exposure sensitivity layer, which represents the total effect of the slope and exposure nature, of the shoreline, on its sensitivity. Table (2) shows the expected results from the overlaying of the slope layer with the exposure one. There are nine values, from 1 through 9, in resultant layer. The addition of the slope/ exposure sensitivity layer  to the morphological sensitivity units , produces the final shoreline sensitivity layer.

The land-use/ land-cover map of 1984, is recoded to produce the foreshore natural resources and habitats layer. Then, the foreshore sensitivity criteria are applied, to produce various foreshore sensitivity  layers. These  layers  are  added,  to  produce  the foreshore sensitivity layer. Table (3) illustrates foreshore sensitivity criteria for the oil pollution and other anthropogenic activities. The weighting factors for the foreshore sensitivity criteria (Table 4), are used to produce foreshore sensitivity layers.

The oil sensitive wildlife layer is produced by defining geographical locations of endangered species, using the base map. This layer is produced, based on field observations, as well as information extracted from the published papers and reports.

Finally, the shoreline sensitivity, foreshore sensitivity, and oil sensitive wildlife layers are overlain, using the union option, to produce the overall  environmental sensitivity index map. This map will be helpful for decision makers and planners, to protect and sustain the development processes, throughout the study area.   

The main objective of the site suitability model is to select the optimal sites of  building a tourist village. Figure (3) illustrates the flow chart of the site suitability model. This model is created using the Spatial Modeler Language (SML) of ERDAS   IMAGINE  spatial  modeler. It is a component of spatial modeler module.

All the input data are thematic raster layers. Using “ search” function of the spatial modeler, a buffer zone is created around the following thematic layers:

1. Main roads;                                      2. Existing urban communities;

3. Low sensitive shoreline;                 4. Medium sensitive shoreline;

5. High sensitive shoreline;                6. Main faults; and

7. Main wadi (as a watercourse for flood torrents).

 

The deterministic distance of the buffer zone should be as high as we can. The function of  “Search” operation is to perform a proximity analysis on the input thematic layers, and to create a new output layer. The spatial modeler function is defined as (ERDAS, Inc., 1994):

Search(<Input thematic layer>,<Buffer distance>,<Class no.>, ….)

The resultant output layers, are stored as temporary layers (Figure 3).

Then, the “Conditional” function, of the spatial modeler language (Figure 3), is used to apply the deterministic site suitability criteria on the temporary layers, in addition to the slope and DEM ones. The spatial modeler function, “conditional”, is defined as (ERDAS, Inc., 1994):

Conditional{((<input L1> GT<specific value>) AND (<input L2>GT <specific value>) AND (< input L3> LT <specific value>) AND (<input L4> LT <specific value>) AND (<input L5> GT <specific value> AND < input L5> LT <specific value>) AND (<input L6> GT <specific value> AND <input L6> LT <specific value>) AND (<input L7> GT <specific value> AND <input L7> LT <specific value>) AND (<input L8> GT <specific value>) AND (<input L9> GT <specific value>))}

where:

Lx

input layer

GT

greater than

LT

less than

AND

GIS operator (intersection)

Input  L1

main faults temporary layer

Input  L2

main wadi temporary layer

Input  L3

DEM layer

Input  L4

slope layer

Input  L5

low sensitive shoreline temporary layer

Input  L6

medium sensitive shoreline temporary layer

Input  L7

high sensitive shoreline temporary layer

Input  L8

main roads temporary layer

Input  L9

existing urban temporary layer

 

The result of the “conditional” function is stored in a temporary thematic layer. The “clump” function (Figure 3), is used to categorize the pre-selected sites into groups of pixels in a new temporary GIS layer. The spatial modeler function, “clump”, is defined as (ERDAS, Inc., 1994):

Clump (<input thematic layer>, 8) where: (8) =  is the connectivity radius in  pixels.

The clumped temporary thematic layer, is experienced to the “sieve” function. It eliminates clumps of specific sizes, to meet with the requirements of users. The sieve sub-model (Figure 3), is constructed as follows:

1.        The input raster thematic layer, is the clumped temporary layer.

2.        The spatial modeler function is defined as:

Sievetable (<threshold value>, <histogram>, (<clumped layer>))

3.        Custom sieve table is created to show threshold value.

4.        Histogram lookup table, created from input clumped layer, is used to perform “sieve” operation.

The result of this function, is the final output layer. It represents the optimal sites for building new tourist villages and hotels, according to specified criteria.

3. RESULTS AND INTERPRETATION:

The environmental sensitivity and site suitability models produced the following important results:

 

 

 

 

 

3-1  Environmental Sensitivity Model:

The environmental sensitivity model (Figure 2), was designed to produce the following sensitivity layers:

(1) shoreline morphology units;

(2) shoreline exposure;

(3) shoreline sensitivity;

(4) foreshore sensitivity; and

(5) environmental sensitivity index, as permanent layers.

The other layers would be temporal layers.

The shoreline morphology layer, from Hurghada to Ras Abu Soma, was produced by applying various shoreline sensitivity criteria on the shoreline geomorphology map. In this layer, there are five different morphological units. These are gravel beaches; sandy beaches; coralline sea cliff; sabkha; and sandy spits. The shoreline exposure layer shows three different categories:

1)       exposed;

2)       semi-sheltered; and

3)       sheltered shorelines.

The slope temporal layer was multiplied by the shoreline exposure layer, to produce the slope/ exposure sensitivity, as a temporary layer. It was added to the shoreline morphology layer to produce the shoreline sensitivity layer (Figure 4). In this layer, the relative shoreline sensitivity (RSS) ranges from 1 through 9 . The higher (RSS) of the shoreline, is noticed at the following sites:

1)       To the west of Sheraton reef;

2)       Sahel Hasheesh, to the west of Deshat El-Dabaah;

3)       Abu Al-Makhadeg bay;

4)       Sharm Al-Arab;

5)       Sharm Al-Naqa; and

6)       South Ras Abu Soma bay.

The lowest shoreline sensitivity was concentrated in the following sites:

1)       Hurghada embayment;

2)       Deshat El-Dabaah; and

3)       Ras Abu Soma cliff.

The foreshore sensitivity layer  is produced by applying foreshore sensitivity criteria, on the foreshore natural resources and habitats layer. The relative foreshore sensitivity (RFS) ranges from 10 through 12. The highest (RFS) appears in red color. The natural resources in Hurghada marine sector; Sahel Hasheesh bay; and south Ras Abu Soma bay, have the highest (RFS) value.

The final environmental sensitivity index map (Figure 5), is produced by overlaying three layers; shoreline sensitivity; oil sensitive wildlife species; and foreshore sensitivity. The relative sensitivity index (RSI) varied from 1 through 12, where the lowest (RSI) had the white color and the highest (RSI) had the red color. Various oil sensitive wildlife species are represented in different symbols. The highest (RSI) is centered on the following sites:

1)       The marine sector around the coral islands, to the east of Hurghada;

2)       Sheraton reef and the nearby sites;

3)       Sahel Hasheesh bay; and

4)       South Ras Abu Soma bay.

This layer should be  a deterministic layer, in any development process, throughout the shoreline from Hurghada to Ras Abu Soma.

3-2. Site Suitability Model:

The site suitability model is an open model. By changing the value of used criteria, the selected best sites will be changed. Table (5) illustrates the criteria used and their values. To check the model, these criteria  were applied on the study area from Hurghada to Ras Abu Soma.

 

Table (5): The used criteria and their values

 

 

Criteria

Value

1

Slope of the shoreline area.

< 5 degree

2

Distance from the main roads

> 5 pixels ( > 150 m)

3

Distance from the existing urban communities.

> 5 pixels (> 150 m)

4

Elevation of the main land.

< 70 m

5

Distance from the shoreline of low sensitivity

>2 &<150 pixels (>60&<4500 m)

6

Distance from the shoreline of medium sensitivity

>3 &<150 pixels (>90&<4500 m)

7

Distance from the shoreline of high sensitivity

>7&<150 pixels (>210&<4500 m)

8

Area of the selected sites.

> 1800 m2

9

Distance from the boarders of main wadi

>20 pixels (>600 m)

10

Distance from the main faults

>20 pixels (>600 m)

 

Figure (6) shows the best selected sites in Hurghada sector, for tourist villages. The best sites appear in red color. The total area of  selected sites is equal (1.2 km2). The best sites are located to the north of Hurghada region, and to the west of Hurghada embayment. A very small site is shown to the west of Sheraton reef. Also, Figure (6)  illustrates the existing tourist villages and hotels (yellow). It appears that the existing tourist villages and hotels were not selected, as  best sites, using the model. The total area of existing tourist villages and hotels is calculated as  (2.4 km2).

Figure (7) shows the best selected sites for tourist villages, in Ras Abu Soma study area. The total area of the best sites (red) is equal (18 km2). It means that Ras Abu Soma sector is a promising site for tourism industry. The best sites are located on the following areas:

(1) Ras Abu Soma peninsula;

(2) To the north of Sharm Al-Naqa;

(3) Around Sharm Al-Arab; and

(4) Deshat El-Dabaah and Sahel Hasheesh region.

The visual interpretation, of overlaying this layer on the environmental sensitivity index layer (Figure 6), shows that the best sites are far away from the highly sensitive areas. Also, visual interpretation of the best site layer, with the enhanced LANDSAT-TM image, shows that the best sites are far away from major wadis. The model has excluded any sites, on front of main wadis. Also, the distance of the selected sites from the shoreline, dependent on the environmental sensitivity of the shoreline and foreshore natural resources and habitats.

 

4. CONCLUSION AND RECOMMENDATIONS:

After the serious environmental impacts which occurred due to landfilling, oil pollution, dynamite fishing and spearfishing (El-Gamily, 1998), throughout the coastal zone of Hurghada, Ras Abu Soma sector should be carefully developed. The environmental sensitivity model, as well as the site suitability one, are valuable tools to support decision-makers and planners, in the development process of the environmentally sensitive areas, such as Hurghada and Ras Abu Soma coastal zones.

The following recommendations are given to support the decision makers and planners, to protect and develop the Red Sea coastal zone.

(1) The environmental sensitivity index maps for the Red Sea coastal zone should be produced.

(2) The site suitability model is an efficient decision support tool, in the process of sustainable development, of the Red Sea coastal zone.

ACKNOWLEDGMENT:

This work has been carried out in cooperation with The Cabinet, Information and Decision Support Center (IDSC), Cairo, Egypt.

 

 

REFERENCES:

Abdel-Kader, A. F., Nasr, S. M., El-Gamily, H. I. And El-Raey, M., 1997. Environmental sensitivity analysis of potential oil spill for Ras Mohammed coastal zone, Egypt. Journal of Coastal Research, Vol. 13 (in print): 9 p.

 

Barratt, L., 1982. Scientific report of the joint services expedition to the Egyptian Red  Sea. Coral Reef and Tropical Marine Res. Unit, Univ. York, UK: 47 P.

 

Berry, J. K., 1993. Cartographic modeling: the analytical capabilities of GIS. In: Environmental Modeling with GIS. Goodchild, M. F., Parks, B. O. and Steyaert, L. T. Eds., New York, Oxford University Press: 58-74.

 

Bonham-Carter, G. F., 1994. Geographic Information Systems for Geoscientists: Modelling with GIS. Geological Survey of Canada, Love Printing Service Ltd, Ontario, Canada: 398 P.

 

CRI, 1992. Vulnerability assessment to accelerated sea level rise, case study: Egypt. Final report, Coastal Research Institute, Alexandria, Delft Hydraulic.

 

Dangermond, J., 1993. The Role of software vendors in integrating GIS and environmental modeling. In: Environmental Modeling with GIS. Goodchild, M. F., Parks, B. O. and Steyaert, L. T. Eds., New York, Oxford University Press: 51-56.

 

El-Gamily, H. I., 1998. Remote sensing and GIS techniques for environmental sustainable development of the Red Sea coastal zone. Institute of Graduate Studies and Research, Alexandria University: 189 p.

 

ERDAS, Inc., 1994. Geographical models reference guide. ERDAS, Inc., Atlanta, Georgia, USA: 102 P.

 

Fedra, K., 1993. GIS and environmental modeling. In: Environmental Modeling with GIS. Goodchild, M. F., Parks, B. O. and Steyaert, L. T. Eds., New York, Oxford University Press: 35-50.

 

Goodchild, M. F., 1993. From modeling to policy. In: Environmental Modeling with GIS. Goodchild, M. F., Parks, B. O. and Steyaert, L. T. Eds., New York, Oxford University Press: 315-316.

 

Heathcote, P., Parr, D., Jennings, M. and Frost, B., 1984. Oil pollution of the Egyptian Red Sea and Gulf of Suez and its effects on birds. Scientific report, Tropical Marine Res. Unit,   Univ. York, UK: 72 P.

 

IUCN/UNEP, 1988. Coral Reef of the World. Vol. 2: Indian Ocean, Red Sea and Gulf. UNEP, Nairobi, Kenya, IUCN, Gland, Switzerland and Cambridge, UK.

 

Johnston, C. A., 1993. Introduction to quantitative methods and modeling in community, population, and landscape ecology. In: Environmental Modeling with GIS. Goodchild, M. F., Parks, B. O. and Steyaert, L. T. Eds., New York, Oxford University Press: 276-283.

 

Kemp, K., 1997. Modeling the environment in GIS I: mathematical models of environmental processes. Regional Workshop on: Integrated Environmental Information Systems. Cairo, Egypt: 11 P.

 

Mohamed, H. and Abul-Azm, A., 1992. Environmental impacts of Ras Abu Soma development project, northern Safaga bay, the Egyptian Red Sea. Environmental assessment report (unpublished): 108 P.

 

Pavasovic, k., 1993. Land suitability models. Workshop on Geographical Information Systems in Integrated Coastal Management, Alexandria, Egypt.

 

Tomlin, C. D., 1990. Geographic Information Systems and Cartographic Modelling. Prentice Hall, Englewood Cliffs, New Jersey: 249 P.

 [s1]