In the arid and semi-arid tropics, low annual rainfall together with high intensity rains has resulted in excess runoV, soil erosion and low moisture intake leading to poor crop yields. Therefore, adoption of soil and water conservation measures is necessary for the optimal utilization of natural resources and to increase the productivity of land on a sustainable basis. Remote sensing and G IS techniques can be used for generating development plans for the watershed area in consonance with the production potential and limitation of terrain resources, and can also be used for assessing the impact of these measures before actual implementation in the eld. This paper describes a case study for the Jasdan taluka (district ) of Rajkot in G ujarat, India. The aims are to prioritize watersheds on the basis of runoV generated, expressed as yield, due to existing land use conditions; to suggest soil and water conservation measures; and to evaluate the hydrologic response of these measures on runoV. The Soil Conservation Service Curve N umber (CN ) method was used for computing the runoV; subsequently runoV yield in percentage was calculated for prioritizing the watersheds. Satellite and other collateral data were used to identify the problems and potential in the watersheds and recommend measures for soil and water conservation. The impact of these measures was assessed by computing runoV under alternative land use and management practices. It was found that the runoV yield decreased by 42.88% of the pre-conservation value for the watershed.*Currently on leave from ISRO . P resent address: D epartment of G eography, 217-1984, West M all,

No abstract

Water management simulation model DRAINMOD-S was calibrated (1995-96) and validated (1997) using 3-year experimental field data (1995-1997) from the installed subsurface drainage system at 1.8 m drain depth with 40, 60, and 80 m drain spacing at Golewala watershed, Faridkot, Punjab, India. Sensitivity analysis of the model parameters revealed that lateral saturated hydraulic conductivity, drain depth, and drain spacing are the most effective parameters in changing the model output. The root means square error, efficiency, and coefficient of determination between observed and simulated soil salinity ranged from 0.01to 0.06 dS.m-1, 0.647 to 0.834 dS.m-1, and 0.957 to 0.999 dS.m-1 for three drain spacings (40, 60, and 80 m), respectively, during calibration and validation period. The calibrated and validated model was used to predict the soil salinity (EC) for five consecutive years (1998-2003). The average soil salinity of root zone (300-600 mm), (600-900 mm), and (900-1200 mm) decreased from January 1998 to December 2003. The predicted values of soil salinity were found to decrease from 2.23, 2.34, and 1.92 dS.m-1 to 1.68 dS.m-1, 1.70, and 1.42 dS.m-1 for 40 m drain spacing at root zone depth of 300-600 mm, 600-900 mm, and 900-1200 mm, respectively. Similarly, the salinity values for the same period and root zone depth were found to decrease from 2.20, 2.31, and 1.90 dS.m-1 to 1.75,1.78, and 1.74 dS.m-1 for 60 m drain spacing; and 2.21, 2.31, and 1.93 dS.m-1 to 1.80,1.82, and 1.48 dS.m-1 for 80 m drain spacing, respectively, at the end of five years. DRAINMOD-S model was reliably applicable for predicting soil salinity under sub-surface drainage system in arid and semi-arid region of Punjab, India