Abstract:
Present research deals with the problem o f monitoring saline soils o f Hambantota district, southern Sri Lanka, based on Remote Sensing data o f TM sensor o f Landsat 7 satellite, using ILWIS 3.2 'functions. Effectiveness o f saline indices viz. Salinity Index (SI) and Normalized Differential Salinity Index (NDSI) fo r the identification o f soil salinity in the area were tested. Since salt-affected soils can also be characterized by stressed vegetation, Normalized Differential Vegetation Index (NDVI) is also used.
Supervised classification was used to isolate the area coverd with Acacia leucophloea. Five landuse classes namely water surfaces o f tanks, swamps and canals [Water], scrublands [Scrub], Acacia leucophloea [Acacia], Paddy cultivation [Paddy] and other vegetation [Other vegetation] were isolated fo r the study. The ground truth data and 1:50000 topographic maps were used to picking up training sites fo r all classes.
After geometric correction, Histogram equalization stretch and filtering operations in ILWIS, the image was visually interpreted in individual bands and various band combinations. Supervised classification was employed using the identified five classes with maximum likelihood, minimum mahalanobis distance and minimum distance methods including the bands 2, 3 and 4. Results o f the analysis helped to produce SI, NDSI and NDVI maps o f the Hambantota District. Supervised classification helped to produce land cover maps in respect to Vegetation (Acacia and Other vegetation), Paddy fields, Scrub lands and Water bodies. The results revealed that 35% o f land area o f Hambantota District was covered with Acacia leucophloea and its associates which appeared to be a good indicator o f salt-affected soils, using GIS and Remote Sensing techniques. The saline indices o f SI and NDSI used in arid areas to identify saline soils cannot be used in Hambantota district due to the growth o f Acacia leucophloea vegetation and its associates.
