Cent. Eur. J. Geosci. • 1(1) • 2009 • 120-129 DOI: 10.2478/v10085-009-0008-5

Central European Journal of Geosciences

Groundwater potential zonation for basaltic watersheds using satellite remote sensing data and GIS techniques Research Article

Biswajeet Pradhan∗ Faculty of Forestry, Geo and Hydro-Science, Dresden University of Technology, 01062 Dresden, Germany

Received 11 January 2009; accepted 12 February 2009

Abstract: This paper summarizes the findings of groundwater potential zonation mapping at the Bharangi River basin, Thane district, Maharastra, India, using Satty’s Analytical Hierarchal Process model with the aid of GIS tools and remote sensing data. To meet the objectives, remotely sensed data were used in extracting lineaments, faults and drainage pattern which influence the groundwater sources to the aquifer. The digitally processed satellite images were subsequently combined in a GIS with ancillary data such as topographical (slope, drainage), geological (litho types and lineaments), hydrogeomorphology and constructed into a spatial database using GIS and image processing tools. In this study, six thematic layers were used for groundwater potential analysis. Each thematic layer’s weight was determined, and groundwater potential indices were calculated using groundwater conditions. The present study has demonstrated the capabilities of remote sensing and GIS techniques in the demarcation of different groundwater potential zones for hard rock basaltic basin. Keywords: groundwater • watersheds • GIS • remote sensing basin © Versita Warsaw

1. Introduction The growing demand for our “most precious resource” will never cease. As our population increases, the need for potable water also increases. Agriculture and rapid industrialization has only added to the growing demand, especially for groundwater. However, just like any other “precious” commodity, it has to be conserved for potential use. The watershed management has become of major importance for the development and progress of all ∗

120

E-mail: [email protected]

regions. The present study highlights the usefulness of remotely sensed data and knowledge based geographical information systems (GIS) for the collection, integration and analysis of the existing data in GIS environment. The systematic study of geological, hydrogeomorphological, hydrogeological and remote sensing [1–9] parameters have been chosen in quantifying the groundwater resource and also predicting the groundwater potential in Bharangi River basin, Maharastra. The study area, the upper Bhatsai basin, lies in between the latitudes 19°30.9’ and 19°13.18’ and longitudes 73°14.55’ and 73°21.43’, covering an area of approximately 59.5208 km2 . The location of the study area is shown in Figure 1. Geomorphologically, the study area comprises

Biswajeet Pradhan

Figure 3. Figure 1.

Figure 2.

Lithological map.

Location map of the study area.

Hydrogeomorphological map.

a typical horizontal basalt, namely mesas and buttes, which constitutes 15 percent of the study area. There are steep scarps present in the north-western boundary of the drainage basin (Figure 2). The region is mainly hilly with an extensive plateau to the west and small plateaus to the north and south. The valley itself is quite narrow and elongates, with steep ridge slopes that grade into a portion of terraces that, in turn, gently slope towards the main river channel. The Bharangi River basin experiences seasonal summer from March to June and winters from November to the following February. The nights during the summer months are quite warm, with an average temperature of apprioximately 24°C, while the average temperature during the day is about 41°C. During the winter months the mean daily temperature is around 30°C and the average night temperature is approximately 12°C with December being

the coldest month. Humidity is very high during the summer months with an average of about 70-80% relative humidity. The climate is reported to vary as one travels from the upper reaches of the valley to the lower reaches. Consequently, there is a difference in the land use pattern and soil type in the entire valley, possibly on account of variable climatic conditions. For instance, the soil in the study area is predominantly black cotton soil with the thickness of the soil layer varying from 0.5 meter in the slopes to about 1.5 meter in the plains. The study area falls within the Deccan Basaltic Province which has an areal extent of over 500,000 km2 ; covering parts of Maharastra, Karnataka, Andhra Pradesh, Madhya Pradesh and Gujarat and has a maximum thickness of about 2500 m. The “Lithological Map” of the study area is shown in Figure 3 [10–14]. Due to high rainfall and large variation in seasonal temperature, the basalts are found to be weathered up to a maximum depth of about 10 meter. Initial spheroidal weathering has given rise to lateritic soils in the area which has been observed in locations such as Shil and Airavil (Figure 4). Red boles have been found near the Belapur Station which may be a product of the weathering of basalt (Figure 5).

2. Data used The following data was used for the study: (i) Remotely sensed data, viz. IRS 1B LISS III with 23.5 meter spatial resolution; LANDSAT TM with 30 meter spatial resolution (ii) The Survey of India toposheet 47E/7scale 1:50,000, and (iii) field data, viz. vertical electrical sounding data. Six factors (slope, drainage density, lineament density, land use/land cover, lithology geomorphology) were considered for the groundwater zonation analysis, which were extracted from the constructed spatial database. The factors were transformed into a vector-type spatial database using GIS, and groundwater-related factors were ex121

Groundwater potential zonation for basaltic watersheds using satellite remote sensing data and GIS techniques

was then used to rasterize the elevation point clouds to make up the DEM. Using this DEM the slope angle was calculated. In the present study, substantial attention has been given for slope conditions. Slope configuration and steepness played an important role in conjunction with the lithology. The contour height information and elevation data were extracted from Survey of India toposheets and the drainage map was prepared from the Survey of India toposheet & satellite data. All the prepared primary input maps (hydrogeomorphology, lineament, slope and drainage) were digitized in the Arc/View GIS software package, and the slope map was prepared from the DEM. The DEM was then used to describe geomorphological and hydrological processes in the landscape. Figure 4.

Lateritic soil profile along the Bharangi river bank.

In the present study, the DEM’s are assigned as the cell based data with 20×20 meter resolution and registered to the UTM coordinate system, the same as that of the satellite data used in this study.

2.2.

Thematic layers and their analysis

2.2.1. Remote sensing data

Figure 5.

Amygdaloidal basalt showing red boles and spheroidal weathering.

tracted using the database. In order to demarcate the groundwater potential zones of the study area, different thematic maps on 1:50,000 scales were prepared from remote sensing data, topographic maps and resistivity data. The thematic maps on hydrogeomorphology and lineaments were prepared using IRS 1B LISS-III satellite data by visual interpretation on 1:50,000-scale.

2.1.

DEM generation

A digital elevation model (DEM) was created first from the topographic database. Contour and survey base points that had elevation values were extracted from the 1:50,000-scale topographic maps. The contour lines and elevation data clouds were interpolated using triangulated irregular networks (TIN). The TIN-mesh was built to represent the facets of the point clouds by using Delaunay triangulation [15]. The TIN-based interpolation method 122

Remote sensing formed an integral part of the present work, as it provided the maximum information on the various geomorphic parameters like landforms, structures, etc. Thematic maps such as lineament maps, land use/ land cover maps, etc have been prepared from the remote sensing data. In this case, Landsat TM and IRS 1B LISS III satellite data were used. The Landsat TM is a multispectral scanning system that records reflected/emitted electromagnetic energy from the visible, reflective-infrared, middle-infrared and thermal-infrared regions of the spectrum. TM has swath width of approximately 185 km from a height of approximately 705 km. The spatial resolution of TM is 30×30 meter for six bands and 120×120 meter spatial resolution for the thermal band (band 6) with 8 bit radiometric resolution [16, 17] projection. The Landsat TM digital data was used for digital image processing and different images were prepared. The image processing work started with the extraction of the digital data by windowing data of 600 columns and 450 rows. Using the Erdas Imagine 8.7 image processing software, more than 50 ground control points (GCPs) were selected from the rectified image and paired with their uncorrected corresponding points in the raw image. Then polynomial transformation was applied to georectify the image. The calculated RMS error was acceptable in the range between 0.127 and 1.032. For resampling, the algorithm of nearest neighbour was used. The radiometric value of the output pixel was set equal to the value of the nearest input pixel in the original geometry. This algorithm preserves the original digital numbers and it is good for

Biswajeet Pradhan

Table 1.

Figure 6.

Landsat TM False Colour Composite (FCC) image draped over DEM.

following classification work [18]. The Cubic convolution algorithm produces a better-looking image, but, on other hand changes the original digital numbers. The rectification was necessary for using the image of the study area with other topographical data. Feature selection plays an important role in choosing the right band combinations used in visualization and classification. According to Skidmore [19], band combinations are created by combining bands of spectral data to enhance the particular land form or land cover of interest. A significant advantage of multispectral imagery is the ability to detect important differences between surfaces by combining spectral bands. Within a single band, different materials may appear virtually the same, but by selecting particular band combinations, various materials may be contrasted against their background by using colour. The visible bands (RGB on channels 3, 2, 1) were used for a first reconnaissance of processing image. True colour composite shows cultural features such as road, but this composite has low contrast and appears very flat. False colour composite (RGB on channels 4, 3, 2) was used initially to detect the influence of soils by vegetation. This composite appears similar to an infrared photograph, for instance, vegetation appears red, and water appears navy blue or black. False colour composite (RGB on channels 7, 5, 3) data was used to solve the problem (Figure 6).

2.2.2.

Hydrogeomorphology

A hydromorphological map (Figure 2) was prepared by digitizing the topographical maps through inputs from the field observations. The Landsat TM FCC was used in determining the various morphological features in the study area.

Karade sub-basin

Chandroti sub-basin

Jamchelpada sub-basin

Bendodi sub-basin

Mean Bifurcation Ratio

3.43

3.37

4.46

3.11

Weighted mean bifurcation ratio

4.37

4.66

4.91

4.89

Mean length ratio

1.96

2.1

0.78

2.46

Weighted mean length ratio

1.84

1.93

0.74

1.72

mon drainage patterns observed in the study area were dendritic and sub-dendritic. A drainage pattern (Figure 7) map was prepared from the topographical maps. Table 1 shows the results of drainage characteristics for the study area. The map was further used for the preparation of other thematic maps like drainage density (Figure 8) and drainage frequency. The drainage density map depicts 3 classes in terms of km2 , low (0-2), medium (24) and high (>4) drainage density. Jamlechapada subbasin falls in the low drainage density class; Bendodi comes under medium class while major portion of Karade and Chandrotic sub-basin indicates high drainage density class (Table 2). Similar to the drainage density, the drainage frequency was also prepared by taking into account the number of streams in each grid. Length of overland flow was also calculated for four sub-basins [20, 21]. The drainage density and drainage frequency values were reclassified into three groups (Figure 8 and 9).

2.2.4. Water divided zones Water divide zones play an important role for the non percolation of groundwater. Water divide zones were delineated from the drainage map showing as the zone of high run-off and less percolation. Rainfall data for the past 10 years of the area were plotted to make a correlation with the net recharge of the basin (Figure 10).

2.2.5. 2.2.3.

Drainage

Sub-surface hydrological condition of any area is controlled by the drainage characteristics of the basin that leads to decipher the groundwater condition. The com-

Results of drainage characteristics.

Lineaments

Lineament plays a vital role for the development of the groundwater zone. Lineaments were mapped in terms of dykes and fractures (Figure 11) from the satellite images. Landsat TM FCC and IRS LISS 1B III were used to delin123

Groundwater potential zonation for basaltic watersheds using satellite remote sensing data and GIS techniques

Figure 9. Figure 7.

Drainage frequency (km per km2 .).

Drainage pattern map.

Figure 10.

Figure 8.

Drainage density (km per km2 .).

Rainfall plot of Bharangi river basin.

low as 0°. Keeping the slope vector map as the base map, the slope elements were grouped into three classes (Figure 13) by keeping in mind the hydrological characters of the basin. The classes are: (i) >15°, (ii) 15° -5°, (iii) <5°.

eate the lineaments by applying vertical gradient filters. Field experience was essential in delineating the various linear fractures [22, 23]. These lineaments were classified into four groups: major, minor, faults and thrusts; depending on the nature and aerial extent. Keeping the lineament map as the base map, a lineament density map (Figure 12) was prepared. In the preparation of the lineaments density map, the values were grouped into three classes represented on a chloropleth map by (i) low, (ii) medium and (iii) high.

2.2.6.

Slope

The slope of the land varies between 1% and 50% with basaltic ridges on three sides surrounding the Bharangi River basin. An actual slope of the area was prepared from the toposheet of Survey of India. Contours and spot heights were measured from the topographical map in terms of magnitude and direction of the slope. The slope values were found to vary considerably, from 85° to as 124

Figure 11.

Lineament pattern map.

Biswajeet Pradhan

Figure 12.

Figure 13.

2.2.7.

Figure 14.

Landuse/Land cover map draped on DEM.

Figure 15.

Graph showing water table layer vs. surface layer.

Lineament density map (km per km2 ).

Slope classes map (in degrees).

Land use and Land cover

A land use and land cover map (Figure 14) was based on the field data in conjunction with the inputs from topological maps and IRS LISS 1B FCC. The false colour composite (FCC) was very useful in identifying structures such as joints, and landforms such as forests, water bodies, etc. The standard FCC generated using bands 1, 2, 3 projected on blue, green and red, respectively, provided useful information on vegetation, water bodies, land use, land cover, rock types and structure. The standard FCC was used for selecting training sites for the supervised classification. Classes were selected and at least five training sets were chosen for all the three groups. A classified image was generated using these training areas on the three principal components [24]. An edged enhanced image for the area was generated using the ORTHOPHOTO package and it was found useful in structural and geomorphological interpretations. Supervised classification was carried out by using the training sites. The Idrisi software was used to develop a statistical characterization of the reflectance for each information class “Signature analysis”. After statistical characterization was done for each information class, the image was classified by examining the

reflectance for each pixel and making a decision about which of the signatures is resembled.

2.2.8.

Aquifer zone

The vesicular-amygdaloidal portions of simple and compound basalt flow/units, whenever sheet jointed, form the main aquifers in the area. Sometimes, the underlying compact basalt in hydraulic connection with these vesicular amygdaloidal portions also acts as a secondary portion of the basalt aquifer. The shallow aquifer of Bharangi River valley is mostly tapped through large diameter dug wells. The wells have diameters between 3 to 7 meter and 9 to 13 meter deep. Usually, in most of the wells the upper 1.5-2.8 meter portion is lined with masonry. Borehole data was taken from the various locations of the area and water table layer was drawn showing a clear demarcation between aquifer zone and soil cover (Figure 15).

3.

GIS modeling

A groundwater potential zonation map was prepared by different geological and hydrogeomorphological parameters such as slope, land use and lineament density, etc. The methodology adopted in this research was obtained from Goyal et al. [25]. The map overlying technique was 125

Groundwater potential zonation for basaltic watersheds using satellite remote sensing data and GIS techniques

applied in order to reveal the hydrological condition of the region. In the present analysis, drainage density and drainage frequency maps show close relationship and hence only the drainage density map was selected. The selected thematic maps were converted into digital format by scanning and digitization of the existing maps (vector). After rasterization, all the analytical steps were carried out in a GIS environment. The individual maps were then reclassified and each class was assigned to a weight in accordance with its influence on the occurrence of groundwater. The resulting reclassified and weighted maps were integrated by using map addition module. The resultant map had various values and was reclassified into a map with three classes with distinct ranges of values. This classification was done on the basis of abrupt changes in the histogram of values of the map produced by integration. These three different classes represented zones of low, medium and high groundwater potential. Satty’s Analytic Hierarchy Process was applied for the selection of the weightage of the individual maps. In this method, a pair wise comparison matrix was prepared for each map using Satty’s nine point importance scale [25, 26], and this matrix was solved using Eigen Vector method. In this method the basic input is the pair wise comparison matrix ”A” of order n×n constructed based on Satty’s scaling ratios. X = [xij ] ,

where

i, j = 1, 2, 3, . . . , n.

(1)

The matrix ”A” has generally the property of reciprocity and also the consistency. This is explained as Xij =

1 . aij

(2)

Mathematically, the matrix equation can be expressed as (A − II)x = 0,

(3)

where: I = identity matrix of order of n×n, 1 = eigen value, x = eigen vector. The result of the above equation (3) gives the individual class weight. Similarly the map scores were also calculated by the same procedure. Thus, finally the map scores were multiplied with the weights and applied to linear summation equation. Y = Ywi ,xi , Y = weighted map, wi = weight of individual class, xi = weight of map score. 126

(4)

Figure 16.

Assigned weights and map score.

The results show that the weight value ranges from 0.021 to 0.456 (Figure 16). The lowest weight found in the study area is 0.021. By utilizing the above mentioned model, the groundwater potential zonation map was prepared.

4. Terrain mapping units and groundwater potential zonation Terrain mapping units and groundwater potential zonation involves the grouping of the terrain into different units based on the different geomorphic parameters such as slope, drainage density and lineament density. The terrain mapping unit approach of Meijerink [27] was used in the present study. The procedure involved the identification of units at hierarchical levels, namely terrain mapping complex (TMC), terrain mapping units (TMU) and terrain mapping sub-units (TMS). The TMC’s were identified based on the geomorphic complexes. In this study, side slope, flat ridge crest, and pediments and plain were identified as the three TMC’s. The TMC map was overlaid on the lithology map, and the unique combinations were recorded to delineate TMU’s. Thus, the side slope with massive basalt represents one of the TMU’s. The TMU map has been overlaid on the slope and drainage density maps to identify 54 TMU’s. A pediment in massive basalt with high gentle slope and low drainage density is an example of TMS (Figure 17). A three digit coding system was used to identify each TMS. The first, second and third digits refer to the TMC, TMU and TMS respectively. The TMS map was converted to a Groundwater Potential Zonation (GWPZ) map by reclassifying the TMS into three groups namely high, moderate and low potential based on the understanding of the terrain. However, in each case, it is imperative that the geomorphologist should formulate a set of decision rules to explain the reason for assigning a particular potential to a TMS. The final qualitative groundwater potential zonation

Biswajeet Pradhan

omorphology, geology, lineament, land use/ land cover, slope, water depth and aquifer media. The results showed that geomorphology contributed most significantly to the groundwater potential zonation with a highest score weight of 0.485. Next to geomorphology, geology played the second most important role for the groundwater delineation with a score weight of 0.188. This could be due to presence of numerous faults/ dykes and presence of various stratigraphic controls. The presence of amygdloidal basalts also supported a good groundwater regime.

Drainage density was measured for the sub-basins and ranged from 4.23, 4.05, 2.60 and 3.13 km/km2 , respectively (Table 2). The slightly lower drainage density of the Jamlechapada sub-basin indicated relatively higher permeability and the presence of more fractured rock. This also indicated less surface-run-off in the sub-basin. The constant channel maintenance was calculated from the drainage density value. The comparatively larger value in case of the Jamlechpada sub-basin indicated that due to the comparatively more permeable nature of this region, a larger area was required to maintain a stream length of 1km. Constant channel maintenance values for the other three sub-basins were found to be similar. The stream frequency value of the Jamlechapada sub-basin was lesser than that of other sub-basins pointing lower surface runoff and ground slopes.

Figure 17.

Terrain mapping sub-units (TMS).

Figure 18.

Groundwater potential zonation (qualitative).

In order to validate the model, borehole yield data was collected during the field work for the study area which reflected the actual groundwater level. A comparison study between the borehole yield data and groundwater potential zones prepared by the model was made to check the validity of the proposed model. The results showed that both sets of data complemented each other. Table 3 further shows a good agreement between yield data and estimated groundwater potential zonation map.

map is shown in Figure 18.

5.

Results and discussion

In the present study the main factors responsible for the delineation of groundwater potential zonation were ge-

Table 2.

Morphometric parameters of Karade (a), Chandroti (b), Jamlechapada (c) and Bendodi (d) sub-basins. Order of Streams in subbasins

Sub- basins Karade (a) Chandroti (b) Jamlechapada (c) Bendodi (d)

5 5 4 5

Number of streams

Sub-basin Area (km2 )

Total length of streams (km)

N

A

L

97 133 49 63

10.34 12.81 14.38 13.09

43.81 51.95 37.5 41.1

Drainage density (1/km)

Stream frequency (1/km2 )

Constant channel of maintance

Length of overland flow (km)

Dd=L/A

F=N/A

C=1/Dd

Lg=1/(2D*d)

4.23 4.05 2.60 3.13

9.38 10.38 3.40 6.24

0.236 0.246 0.384 0.319

0.118 0.123 0.192 0.159

127

Groundwater potential zonation for basaltic watersheds using satellite remote sensing data and GIS techniques

Table 3.

Actual bore well yield data for validation of the model.

Site

Name of the

Groundwater

location

place/ village

potential zones

obtained from

achieved by

bore hole (lpm)

No.

the model

6.

1

Bendodi

High

2

Karade

High

678 587

3

Jamelchapada

Moderate

380

4

Chandroti

Moderate

420

5

Shil

Low

80

6

Airavil

Low

137

7

Shahpur

Low

203

Conclusion

In the present study, qualitative analysis was carried out using map overlying techniques. The output map was found to have three classes, with high groundwater potential classes occurring in plain areas with the presence of highly fractured amygdaloidal basaltic rock. We can therefore conclude that high potential zones are found within very low drainage density and moderate to low lineament density areas. This region is dominated by basalt, which being a hard rock is generally considered to be a poor aquifer. However, since it is highly jointed in nature and has suffered a high to moderate degree of weathering it has been converted to a fairly good aquifer. Geomorphological studies have indicted that the northern part of the basin is fairly immature and is the run off zone. The drainage area is mature and is the saturated zone with a large amount of infiltration. Thick zones of weathering were observed in the area indicating groundwater potential in this aquifer. This study has demonstrated the capabilities of using remote sensing and GIS for demarcation of different groundwater zones, especially in diverse geological setup. This gives a more realistic groundwater potential map of an area to be formed, which may be used for any groundwater development and management program.

Acknowledgements: Thanks are due to Dr. Lucian Dragut and several anonymous reviewers for their critical comments and suggestions that helped to bring the manuscript to the current form.

128

References

Actual yield

[1] Saraf A.K., Chaudhary P.R., Integrated Remote Sensing and GIS for Groundwater Exploration and Identification of Artificial Recharges Sites, Int. J. Remote Sens., 1998, 19, 1825-1841 [2] Krishnamurthy J., Venkataesa K., Jayraman N. V., Manivel M., An Approach to Demarcate Groundwater Potential Zones through Remote Sensing and GIS, Int. J. Remote Sens., 1996, 17, 1867-1884 [3] Murthy K. S. R., Groundwater Potential in a Semiarid Region of Andhra Pradesh: A Geographical Information System Approach, Int. J. Remote Sens., 2000, 21, 1867-1884 [4] Edet A. E, Okereke C. S., Teme S. C., Esu E. O., Application of Remote Sensing data for Groundwater Exploration: A Case Study of the Cross River State, Southeastern Nigeria, Hydrogeol. J., 1998, 6, 394-404 [5] Waters P., Greenbaum D., Smart P.L., Osmaston H., Application of Remote Sensing to Groundwater Hydrology, Remote Sensing Reviews, 1990, 4, 223-264 [6] Noriani S, and Geoff T., Groundwater Quality Assessment using Remote Sensing and GIS Techniques in Selangor and Kuala Lumpur, Malaysian Journal of Remote Sensing & GIS, 2002, 3, 39- 49 [7] Shankar M.N., Mohan G., Assessment of the groundwater potential and quality in Bhatsa and Kalu river basins of Thane district, western Deccan Volcanic Province of India, Environ. Geol., 2006, 49, 990-998 [8] Rokade V.M., Kundal P., Joshi A.K., Groundwater Potential Modelling through Remote Sensing and GIS: A Case Study from Rajera Taluka, Chandrapur District, Maharashtra, J. Geol. Soc. India, 2007, 60, 5-12 [9] Pandey A.C., Mondal Md. S., Garg R.D., Groundwater prospects evaluation based on hydrogeomorphological mapping using high resolution satellite images: A case study in Uttarakhand, Journal of Indian Society of Remote Sensing, 2008, 36, 69-76 [10] Ravindra K., Fundamental of Historical Geology and Stratigraphy of India, Wiley Eastern Limited, New Delhi, 1988 [11] Carter H. J., Geography of the South West Coast of Arabia, Journal of Bombay Branch of the Royal Asiatic Society, 1851 [12] Peshwa V. V., Mulay J. G., Kale V. S., Fractured Zones in Deccan Traps of Western and Central India: A Study Based on Remote Sensing Techniques, Journal of Indian Society of Remote Sensing, 1987, 15 [13] Sukeshwala R. N., Sethna S. F., Deccan Traps and Associated Rocks of Bassein Area, J. Geol. Soc. India, 1962, 3, 125-136

Biswajeet Pradhan

[14] Thompson R. D., Sketch of Geology of the Bombay Islands, Madras, Journal of Lit. Science, 1836, 5, 159175 [15] Golias N. A., Dutton R. W., Delaunay triangulation and 3D adaptive mesh generation, Finite Elem. Anal. Des., 1997, 25, 331-341 [16] Lillesand T. M., Kiefer R. W., Remote Sensing and Image Interpretation (3rd Edition), John Wiley & Sons, New York, 1994 [17] Schultz G. A., Remote Sensing in Hydrogeology, J. Hydrol., 1988, 100, 239-265 [18] Dobrovolný P. Dálkový pr˚ uzkum Zemˇe – Digitální zpracování obrazu, Masaryk University, Brno 1998 [19] Skidmore A. K., Remote Sensing of Soils in Eucalypt Forest Enviroment, Int. J. Remote Sens., 1997, 18, 3956 [20] Horton R.E., Erosional Development of Streams and Their Drainage Basins, Hydrophysical Approach to Quantitative Morphology, Bulletin of The Geological Society of America, 1945, 56, 275-370 [21] Strahler A. N., Quantitative Analysis of Watershed Geomorphology, Transactions American Geophysical Union, 1957, 38, 913-920 [22] Moore G., Waltz F.A., Objective Procedure for Lineament Enhancement and Extraction, Photogramm. Eng. Rem. S., 1986, 49, 641-647 [23] Drury S. A., Image Interpretation in Geology, Allen and Unwin, London, 1987 [24] Jensen J. R., Introductory Digital Image Processing: A Remote Sensing Perspective. Prentice Hall Series in Geographic Information Sciences, New Jersey, 1996 [25] Goyal S., Bharadwaj R.S., Jugran D.K., Multi criiteria analysis using GIS for groundwater resource evaluation in Rawasen and Pili Watershed, U.P., GIS Development (Online), 1999, www.gisdevelopment.net/application/nrm/water/ground/watg0004pf.htm

[26] Voogd H., Multi-criteria Evaluation for Urban and Regional Planning, Pion Limited, London, 1983 [27] Meijerink A. M. J., Data Acquisition and Data Capture through Terrain Mapping Units, International Institute for Geo-Information Science and Earth Observation Journal, 1988, 1, 23-44

129