SEDIMENT LOAD CRITERIA FOR PRIORITISING WATERSHEDS FOR RESOURCE DEVELOPMENT PROGRAMMES

Abstract

Watersheds constitute a viable base for planning resources development and management programmes for the communities concerned. The country has been delineated into 3237 watersheds in the fourth order with size ranging from 20,000 hectares to 300,000 hectares. As these watersheds are the source of goods such as food, fodder, fibre and fuel wood and services like water for the local population, their scientific management is essential for sustainable development and to meet the increasing demand for larger biomass. But over-exploitation of the land resources due to the pressure of human as well as livestock population has resulted in degradation of the watershed in terms of land production, work opportunity and income as well as fall in the availability of water in quantity & quality. Thus Soil & Water Conservation (SWC) treatment measures have been a necessity to restore and enhance its productivity. A number of watershed based programmes are being implemented in the country with the objective of reducing erosion, improving productivity, rebuilding ecosystems and generating employment & income. A methodology for identifying the criticality of these watersheds would help in facilitating investment decision and making best use of the available resources. The critically eroding sub-watersheds in the River Valley Project (RVP) catchments are identified by the All India Soil & Land Use Survey (AISLUS) Organization based on Silt Yield Index (SYI) method. Although this system is being used extensively in the RVP catchments of India, the linkage of SYI iv with the actual Sediment Production Rate (SPR) of the sub-watershed need further study. Moreover, a methodology is needed to quantify the level to which the SYI, being the main objective of the scheme, could be reduced if SWC treatment measures are introduced in the watersheds and the finances that would have to be moblized for reducing the erosion to a permissible level. The response to soil and water conservation measures in moderating erosion would be different for different physiographic attributes and land use. Thus it is not only necessary to know the state of erosion of the watershed but it is equally important to quantify the rate of erosion from different land use within a watershed and the level of treatment measures needed for moderating erosion from different categories of land. Keeping this in view, an attempt has been made in this research study to quantify the sediment production rate from different land use categories of Haharo sub-catchment and explore financial implications for maintaining different levels of sediment yield and sediment production rate. The Haharo sub-catchment, or the watershed 2A2H3 (as per National Watershed Atlas) with an area of 498 sq.kms has been delineated into 21 sub-watersheds of various priority categories. Thirteen of these sub- watersheds come under very high and high priority categories having SYI above 1301. Soil and water conservation measures were being given in the identified priority sub-watersheds since 1978. An evaluation study was carried out during 1990-91 by an independent agency to determine the effect of the SWC treatments in the watersheds. METHODOLOGY The statistical properties of the sediment load data recorded for four watersheds were utilized to determine the sediment yield and sediment production rate for different return periods. Chi Square test of the data for two watersheds revealed the suitability of a Pearson Type-Ill distribution. Hence the Pearson Type-Ill distribution was fitted to the data. Multivariate regression analysis was performed to study the relationship between the area treated in uplands, forest lands, gullied lands along with the climatic input of rainfall. The relationship thus developed was used to estimate the sediment load of the watersheds with progressive coverage with SWC measures. Available rainfall data were used and where rainfall data was not available, the projections were obtained through Auto-Regression-1 (AR-1) model. Like the model for individual mapping units, a model for the whole watershed was developed which could account for the variation of the sediment load vis-a-vis the area under various land uses and the areas treated with SWC measures. A constrained regression analysis was carried out in the form of grid search method for this purpose. A linear relationship was maintained in the model for its use in a Linear Programming model for optimizing the cost of treatment. The investments required to bring down the sediment yield from a higher level to lower level was studied using the current cost of treatment (1989 level) for different land use in the Upper Damodar Valley area. The optimization was also performed with and without the net return constraint. This was carried out for 18 sub-watersheds within four watersheds of Haharo sub-catchment. vi The cutoff limits for different priority category was determined on the basis of estimated SPR for 18 sub-watersheds of the Haharo sub-catchments and studying its frequency distribution. RESULTS As soil and water conservation measures involving construction of masonry structure for gully control, water harvesting etc. are mostly designed for 50 years frequency, considering a return period of 50 years, the sediment production rate was highest for watershed No.4/2 followed by watersheds Nos. 4/4, 4/3 and 4/1. The regression analysis on sediment yield vs soil and water conservation treatment on different land use along with rainfall provided a relationship which explained 49% of the variations in sediment yield. The relationship was utilized to estimate the sediment yield when more areas were brought under soil and water conservation treatment. This helped to generate 46 data points covering areas with various levels of treatment under all categories of land use in the gauged watersheds. Assignment of coefficient to various land uses in the form of grid search analysis resulted in another relationship having a standard error of 6739.2 tonnes and correlation coefficient of 0.803. The relationship thus obtained was used as one of the constraints for the Linear Programming model. The optimization analysis revealed that in the case of watershed No.4/1, the SPR could be brought down from 3.64 tonnes/ha./year to 1.48 tonnes/ha./year at an investment of Rs. 412.3 lakhs. The study also revealed that the investment required for bringing down per unit of sediment yield varied between Rs.16,400 (watershed vii No.4/1) to 20,800 (watershed No.4/3) per tonne. The minimum and maximum estimated SPR for 18 subwatersheds of the Haharo sub-catchment varied from 0.61 to 6.68 tonnes/ha. per year. Based on frequency analysis for the estimated SPR of 18 sub-watersheds, two sub-watersheds came under very high and high priority category which had sediment production rate greater than 5 tonnes per ha. Priority classification based on standardized SPR gave a better distribution of sub-watersheds between various priority categories. After this transformation four sub-watersheds came under very high and high priority categories. The study also helped to quantify the investments that would be required to reduce the sediment production rate from very high and high priority level to medium priority level