Development of a GIS based application for selection of villages for sodic...

Development of a GIS based application for selection of villages for sodic land reclamation

SHARE

L. I. M. Rao, Niva Kiran Verma, N. K. Srivastava, and A.N. Singh

Introduction
In the state of Uttar Pradesh, nearly 1.2 million hectares area is under saline and sodicland (logically ‘USAR’). For reclamation of sodicland in the 10 districts of Uttar Pradesh viz., Aligarh, Allahabad, Etah, Etawah, Hardoi, Fatehpur, Mainpuri, Pratapgarh, Raebareli and Sultanpur, a pilot project was taken up with the assistance of the World Bank. Selection of villages for reclamation in each district through conventional methods is time consuming and involves lot of calculations. Collaborative decision problems such as site selection can be analyzed and supported efficiently with user-friendly computer systems, combining remote sensing, GIS multicriteria analysis techniques and environmental modeling (Greetman & Toppen 1990, Fedra & Jamieson 1996, Fedra et al 1996). Remote sensing and GIS have emerged as powerful tools in isolating favorable zones from unfavorable ones. Spatial data generated on vegetation, land cover, soils, geology and hydrogeology using remote sensing can be effectively used in conjunction with ground information in GIS domain for different reclamation activities such as site selection, soil monitoring, groundwater monitoring in a more scientific manner. Because GIS has capability to capture, store, manipulate and analyze various types of spatial & non-spatial information into a single spatial framework, which is very important for any further monitoring and analysis.

The present paper describes how effectively GIS has been utilized in selection of suitable site for sodicland reclamation activities to be carried out based on certain criteria. The analysis work was done for Raebareli district for optimum selection of villages for sodic land reclamation.

Different thematic map layers viz. Sodicland (land use/Wasteland), Drainage, Electrical Conductivity (EC), Residual Sodium Carbonate (RSC), Depth to Groundwater and Village boundary have been converted to Arc/Info GIS coverages. Each of these layers in GIS was analyzed based on set criteria by assigning a weight and suitable rank values. Cumulative weightage was calculated by multiplying the weight values with multiplication factor decided based on rank. Outputs of all the five maps were integrated to produce the final output, which indicates the villages satisfying, set criteria in four ranges. These ranges indicate the priority of villages for reclamation.

Criteria Table

      Suitability Ranking      
S.No Description Weight Rank1 Rank2 Rank3 Rank4
1. Sodic area (percentage) 2 >30 25-30 15-25 2.0
3. Electrical conductivity (dS/cm) 2 £2.0 >2.0
4. Residual Sodium Carbonate (meq/l) 2 £1.5 >1.5
5. Depth to groundwater (m) 2 3-8 8-10 2-3 10
  Multiplication factor (MF) 3 2 1 0  

The villages for reclamation were selected based on the following criteria:

  1. The villages should have a high density of sodiclands, preferably more than 30% of geographical area
  2. Average size of patches should be > 4ha (40000 sqm),
  3. A main drain should be present near by preferably 0.5 – 1 km and
  4. Quality of groundwater should be acceptable for irrigation.

Integration of thematic maps showing the information relevant to these criteria are being done manually. However, when integration has to be done on 1: 50,000 or at larger scale the process involves handling large number of maps in each district. Preparation of the required thematic maps on the same scale and overlaying is time consuming. With the advent of GIS, handling large amount of data of varied scales has become possible and data analysis faster. A GIS analysis procedure was developed and tested for one district Raebareli.

Study Area
The district Raebareli falls between latitudes 25° 45′ to 26° 45’and longitudes 80° 30′ to 81° 45′. The area is a part of Gangetic alluvium having gentle slopes.

Data Used
The village boundary map, drainage map from SOI toposheets and Sodic land map prepared using IRS satellite data during 1987 prepared on 1:50,000 scale were used. The EC, RSC in groundwater and Depth to Groundwater contour maps were collected from Central Groundwater Board, Lucknow (Singh 1988-89). These maps were implemented in Arc/Info GIS.

Methodology
All the maps were digitized, projected to polyconic projection and attribute information was added. The maps were then analyzed as indicated in the following flowchart:

Flow chart showing the process of integration and analysis of each layer in GIS

Data processing / analysis

Sodic land map:
Sodic land map of the study area was implemented in GIS. This layer consists only barren sodic lands mapped from Satellite data in 1987. All the polygons of sodic land were assigned single class. The sodic land map was integrated with Village boundary map through union. From the resultant map, village wise sodic land area was computed. Based on percent of sodic area to the geographic area each village was assigned a rank. Final sodic suitability layer was prepared by product of rank and multiplication factor for each village.

The resultant map (fig. 2) shows all the villages classified in four categories as per their suitability for reclamation. Villages having more than 30% of sodic area are 128, 25-30% are 149, 15-25% are 234 and rest are not suitable (table 2).

Drainage map:
From the Drainage layer, buffer zones were generated at a distance of 500 m, 1000 m and 2000 m from each drain. The village boundary layer was integrated with buffer layer through union. The village boundaries intersecting 2000 m buffer were assigned rank 3 and rest of villages were assigned rank 4. Within rank 3 villages, those intersecting 1000 m buffer were assigned rank 2 and within rank 2 villages those intersecting 500 m buffer were assigned rank 1. The final drainage proximity layer (fig.3) was prepared by calculating product of rank and multiplication factor for each village.

Based on the analysis criteria 1112 villages are falling in rank1, 212 in rank2, 228 in rank3 and the rest in rank4 (table 2).

Depth to Groundwater:
The depth to groundwater contour map has been classified into four groundwater zones as per the criteria mentioned in table 1. Integrated groundwater layer with village boundary layer through union. Village wise statistics were computed and suitable rank value was assigned to each village based on maximum area falling in one of 4 groundwater zones. The final groundwater suitability map was prepared by calculating the product of rank and multiplication factor. The resultant map shows total 1190 villages falling in rank1, 286 in rank2, 129 in rank3 and rest in rank4.

EC and RSC data:
The contour maps of Electrical Conductivity and Residual Sodium Carbonate (in groundwater) were reclassified each into two categories. In EC map the area having EC value less than or equal to 2dS/m in class1 and the rest in class2 (>2.0 dS/m). Similarly the

RSC map was assigned class1 for the area falling in RSC value less than or equal to 1.5 meq/l and the rest in class2.

These two EC & RSC layers were integrated with village boundary layer separately. The two layers thus obtained were assigned rank and weight values for each village depending on their maximum area falling in rank1 or rank2. Finally, each village has been assigned a weightage value (product of rank and multiplication factor).

All the polygons with EC less than or equal to 2dS/m and polygons with EC greater than 2 dS/m were merged to obtain EC layer with two ranges. This layer was integrated with village boundary layer through union. From the union coverage assigned, all the villages that intersect with EC polygon less than or equal to 2 dS/m were assigned rank1 and rest were assigned rank4. The selection thus made is shown in table2.

Residual Sodium Carbonate:
The contour map of residual sodium carbonate has been converted into polygon coverage. All the polygons with RSC less than or equal to 1.5 meq/l and polygons with RSC greater than 1.5 meq/l were merged to obtain RSC layer with two ranges. This layer was integrated with village boundary layer through union. From the union coverage, all the villages that intersect with RSC polygons less than or equal to 1.5 were assigned rank1 and the rest of villages were assigned rank4. The selection of villages on the basis of RSC is shown in table2.

Rank Weightage Number of villages satifying criteria
Sodicland area Drainage distance(m) Depth to groundwater (m) Electrical Conductivity dS/m Residual Sodium Carbonate meq/l
1 6 319 1155 1215 694 421
2 4 67 233 301
3 2 130 237 152
4 0 1230 121 78 1052 1325

From the results of the above analysis the weightage values of each village for Drainage distance, Depth to groundwater, Electrical Conductivity and Residual Sodium Carbonate were summed. The sum of weightages of each village were then multiplied by the weightage values of Sodicland area to obtain final selection of villages for sodicland reclamation. The details are given in table3 in four ranges.

Class Weightage Range No. of villages selected Priority
Unsuitable 0-10 1232 4
Moderate 11-48 372 3
Good 49-88 91 2
Best 89-144 51 1

Results and Discussion:
In this analysis through GIS integration of various layers became easy and it provides flexibility to change the option. All the five parameters were given equal weight, the weight values were then multiplied with their corresponding multiplication factor and thereby obtained weightages. Since the selection is limited to sodicland area, final weightage values of Drainage, Groundwater, EC and RSC were summed and it was multiplied with sodicland weightage. Thus arrived with each sodicland village with its corresponding weightage. 507 villages were selected in the district which follow the first three priorities. The result of the selection using GIS matches by about 85% when compared to the manual method.


Acknowledgements:
The work was carried out using the GIS facilities of RSAC-UP under UPSLR Project. The authors are thankful to the Scientists of this project for providing necessary data inputs and thankful to Sri Sushil Chandra, Systems in-charge for his cooperation from time to time in executing this work.

References:

  • B.K. Singh, 1988-89, Report on Hydrogeological Frame work and Groundwater Resource Potential, Raebareli district, Uttar Pradesh, C.G.W.B
  • Murruthachalam. M, 1999, A report on GIS Strenghening, Remote Sensing Applications Centre, U.P.