Remote Sensing and GIS Based Ground Water Potential Mapping of Kangshabati Irrigation Command Area, West Bengal

The Remote Sensing and GIS tools have opened new paths in land and water resource studies, presently. Satellite images are increasingly used in ground water exploration because of their utility in identifying various geomorphic features. In the present study, all the prepared thematic data layers such as slope, relief, soil, geology, geomorphology, drainage, land use and ‘NDVI’ are integrated using the Spatial Analyst Tool in Arc GIS 9.2 implying weighted overlay methods to delineate the Ground Water Potential Zones in Kangshabati Irrigation Command Area (KICA). In weighted overlay analysis rank value assigned for each class of all thematic data layers according to their influence on ground water hydrology and factor weighted values are assigned according to analytical hierarchy process (AHP). Finally, an accuracy study is being performed in ERDAS Imagine Software by ground truth verification of 30 training sites with GPS readings for major land use/ land cover information which states the overall classification accuracy of the present study is 86.66%.


Introduction
Ground water is a dynamic and replenishing natural resource, which forms the core of the ecological system. But in hard rock terrains, availability of Ground Water is of limited extent. Agriculture is the main stay in India because 69 % of the total population depends on it. Poor knowledge about this resource, due to its hidden nature and its occurrence in complex subsurface formation is still a big obstacle to the efficient management of this important resource. The varying nature of ground water potentiality and agricultural drought is a recurrent phenomenon in the western part of West Bengal. Now a day's Agricultural drought is also a frequent phenomena in West Bengal. It occurs when soil moisture and rainfall are inadequate during the growing season to support the healthy crop growth to maturity and cause extreme crop stress and wilt. Such condition is the outcome of lowering of ground water level and its less accessibility to various activities. The study of ground water potentiality of Kangshabati Irrigation Command Area (KICA) will exhibit a clear idea about the spatial distribution of ground water and will contribute the knowledge to formulate and execute a suitable plan to improve agriculture and others allied activities.
The integrated remote sensing and GIS based study has facilitated to delineate the ground water potential zones by analyzing various phenomena related to land and water resources. According to Saraf et al. [1] GIS helps to integrate conjunctive analysis of large volumes of multidisciplinary data, both spatial and non-spatial. Jones [2], Sinha et al. [3], Chi and Lee [4], Bahuguna et al. [5] and Kumar et al. [6] studied and also integrated different thematic data layers such as topography, lithology, geological structures, depth of weathering, extent of fractures, slope and drainage pattern with the help of geographic information techniques to delineate ground water potential zones. The Digital Elevation Model (DEM) provides different thematic data layers namely slope, drainage, relief, structural features etc. which are obtained more easily, less subjective and provides more reproducible measurements than traditional manual techniques applied to derive topographic maps [7]. Over the last two decades, digital representation of topography has facilitated a lot to analyze various surfaces and subsurface geomorphic and geo-hydrologic features at different scales. In the field of geologic and geographic research RS and GIS has brought a new horizon by measuring and evaluating topographic data set more conveniently. The geographic information system is very much helpful in delineation of ground water prospect and deficit zones [8,9]. In the present study of preparing the ground water potential mapping of Kangshabati Irrigation Command Area (KICA), various thematic maps namely slope, relief, soil, drainage, geology, geomorphology and land use/land cover were reclassified on the basis of weightage assigned and brought into the 'raster calculatior' function of spatial analyst tool for integration. The weightage for different thematic data layers are assigned considering the work done by Rao and Jugran [10] and Krishnamurthy et al. [11]. But at the time of integration of all the data layers a simple arithmetical model is adopted by averaging the weightage.

Data Source and Methodology
To carry out the study and to prepare ground water potential map assessing the accuracy remote sensing data, others ancillary data, GIS Software (ERDAS IMAGINE 9.0, ARC MAP 9.1), Mathematical Software (MATLAB), and GPS are used in the right sense. The specifications of the satellite and others ancillary data are in detail in the following Table1. The specifications of the satellite and others ancillary data used in the present study are in detail in the following Table 1. The different types of Geomorphological features are found in the study area such as Floodplains, Upland Plains, Badlands, Duricrust, Paradeltaic fan surfaces, Pediments, Pediplains, Ridges and Hills. Drainage network is delineated by using satellite images to visualize the areas of sheet flow/channel flow and the area is classified into different drainage density classes, viz. very high, high, moderate, low and very low. After the extraction of different blocks of the study area from the satellite image (FCC) with the help of AOI tools in the ERDAS IMAGINE software a supervised classification is carried out to obtain various land use/land cover classes i.e., cropland, wet land, barren land, dense forest, degraded forests, and sandbank etc. These pre-field classifications are made to plan the field survey for land use/ land cover data collection.

Creation of thematic data layers
It is universally accepted that satellite derived NDVI is an important index to assess crop stage /condition. Crop condition at any given time during its growth is influenced by complex interactions of weather, soil moisture, and soil and crop types. The analysis of NDVI is regarded as the rough estimation of vegetation amount present and ground water prospect over the space. The NDVI map is prepared from the Land sat TM images of the year 2010. The NDVI involves a non-linear transformation of visible or red (R) and near infrared (NIR) bands of satellite images (Rouse et al., Jackson et al. [12] and Tucker et al.). NDVI value is derived using the following equation.
To estimate prioritized factor/criteria rating value, Analytical Hierarchy Process (AHP) after Saaty [13] is applied developing a consistent couple comparing matrix (Table 2) in which each factor is rated against every other one by assigning a relative dominant value ranging between 1 and 9 using MATLAB Software quantifying consistency ratio (CR) of the matrix.

Ground Truth verification:
The field survey was carried out over a 2-day period at the beginning of 22 nd May 2010. Field work encompasses a thorough study of the area in the satellite imagery, SOI topo-sheets and the classified (supervised) imagery to ensure representative site identification for land use/ land cover data collection. Major land use/ land cover information in the area were obtained from different geographical locations. Total of 30 training sites with GPS readings for various land use/ land cover information have been collected for accuracy assessment.

Application of the model to delineate ground water potential zones
The Ground Water Potential Zones are obtained by integrating all the entire thematic maps in a linear combination model (Equation 2) using the spatial analyst tool in Arc GIS 9.2. During the weighed overlay analysis the ranking values are assigned for each classes of individual thematic map according to the influence of the different parameters on ground water potentiality (Table 3).

Ground Water Potentiality Zones
Kangshabati Irrrigation Command area is divided into 4 ground water potential zones (Figure 3 (14), Keshiary (16), and Garbeta -III (12). It is assumed that the Command area is characterized as Good to Excellent Ground Water (Table 4). North Eastern and South Central part are providing Good Ground Water Potentiality due to the existence of adequate drainage networking system.

Drainage Density and Ground Water Potentiality
Drainage density is high in the Western part of Irrigation Command Area covering 80% of Jhargram, Raipur, Khatra and Simlapal; 90% of Jamboni, 60% of Binpur II, Salboni, Garbeta II, Taldangra and Sarenga. High drainage density more confluence points, active channel erosion and consequently soil loss from the area. So, Western and South Western part of Kangsabati area are dominated by soil erosion. On the other hand East and South East and North East marginal part covering the blocks of Sankrail (80%), Kharagpur-II (60%), Ghatal (60%), Keshiary (80%), Chandrakona (85%), and Goghat (65%) are attributed as low drainage density and good ground water potentiality (Figure 5f). Middle and Southern part shows moderate level of drainage density with moderate ground water potentiality.

Land use / cover and Ground Water Potentiality
Kangsabati irrigation Command area which classified in to major eight Land use pattern i.e. crop land, Scrub forest, Dense forest, Open forest, Medium dense forest, Mixed forest, Barren land, Wet land, Gulley, River, and Sand bank etc. The maximum area is dominated by dense forest which is 23.44 % of the total dense forest area, ( Table  7). The middle most part is basically covered by dense forest, Medium dense forest and Open forest. Extreme Eastern and south Eastern part of the area is experienced as the land of Dry crops, Scrub forest, mixed

Accuracy Result
The basic idea is to compare the predicted classification of each pixel with the actual classification and the basic goal of the accuracy study is to quantitatively determine how effectively pixels were grouped into the correct land cover classes. In the accuracy analysis, dense forest and crop land, mixed forest and medium dense forest, open forest and degraded forest and bared surface were considered as excellent, good, moderate and poor ground water potentiality respectively. The random points are compared with the classification map. When the random points and classification match, then the classification of that pixel is considered accurate. Classification accuracy in a broad sense refers to the correspondence between classification of remotely sensed data and actual observations on the field. The classification accuracy of the present work is 86.66% (Table 8).  Overall classification Accuracy = 86.66%

Conclusion
The calculated prioritized factor rating value of land use, geomorphology, geology, drainage, slope, soil, relief and NDVI are 0.378, 0.205, 0.146, 0.103, 0.069, 0,048, 0.031 and 0.024 respectively that indicate geomorphology, land use, geology and drainage have dominant impact on ground water distribution in the study area. Large part of the study area suffers from severe drought condition and the ground water zonation map of the Command Area will contribute a lot of help and knowledge about the hydrology to the concerned authorities engaged in Land use planning. Basically, western part of the Command Area consisting the blocks of Salbani, Sarenga, Binpur, Raipur, Nayagram, Kotalpur, Khatra, Onda, Jaipur, and some parts of Garbeta-I, Garbeta-II, Garbeta-III with moderate to low ground water potential condition should be immediately paid much attention by Water Resource Planners for ensuring diversified agricultural practice based on ground water prospect. Besides, water resource preservation policies/technique should also be applied to get rid of the problems of water during deficient rainfall year evaluating the ground water potential zones map of the KICA. Overall study concludes that the floodplains and paradeltaic fan surfaces contribute much ground water prospect part of the Command Area. As the Kangshabati Irrigation Command area is an undulating terrain, the low lying area may provide suitable sites for the construction of reservoir which can supply water to water deficient area through proper irrigation system during summer/ dry season for agricultural practice. Besides, the drainage network analysis indicates that the area is fit for the construction of check dams at the confluence point of several streams.