geo-spatial approach for soil salinity mapping in sego irrigation farm, south ethiopia

Soil salinization is a major problem affecting productivity of irrigated lands in arid and semi-arid areas. Managing salinity to minimize environmental impact is a prerequisite for sustainable irrigated agriculture. The objective of this study was to assess the level of salinity in Sego irrigated farm, Ethiopia and to map temporal and spatial distribution of salt affected soils to support management programmes. The study employed normal image classification, and developing models from ECe vs NDSI and thematic layers to map soil salinity, using geospatial tools. Multi-temporal Landsat TM images of 1984, 1995 and 2010 were used to detect salinity affected areas in the study area. Results of the study revealed that around 6 ha of the area gets moderately to strongly salinized per year. Empirical model developed from ECe vs NDSI of 2010 image using regression analysis revealed the coefficient of relation as 66%. The model was extended for the whole area, and revealed that 2.0% of the study area was strongly saline and 54.7% of the area was non-saline. Overlay model developed from water table, land form, land management type and land-cover has revealed that 11.2% of the study area was non-saline, whereas 47.2% and 3.0% was moderately and strongly saline, respectively. Validation of the models was made to test their predication capability and the overlay model has revealed better correlation coefficient of 67.3% to the measured ECe. Most of the salt affected areas were on shallow water table. Cambisols and fluvisols were greatly affected by salinity. The results indicate that long-term irrigation activities would affect agricultural potentiality of the area in the future, and geospatial tools are efficient and feasible for detecting salt affected areas from satellite images and thematic maps.