Change Detection of Land Cover and Salt Affected Soils at Siwa Oasis, Egypt

Taher, M. H. Yossif

Constant and accurate change detection of land cover/land use features is highly significant to interpret the correlation between human activities and natural phenomena in order to make better decisions and for investigating the processes and patterns of landscape changes over time and the impact of these changes on sustainable development. Soil salinization is a major problem affecting productivity of irrigated lands. Managing salinity to minimize environmental impact is a prerequisite for sustainable irrigated agriculture. The current study aims at detecting, classifying and  mapping  the current status for land cover types, salt affected soils and their changes in quantity that have taken place in Siwa Oasis during the period from 1987 to 2017, based on satellite imagery data, GIS facilities, statistical analysis, ground truth and laboratory analysis.  An approach of multi-temporal mid resolution satellite images of TM 1987, and OLI 8 2017 combined with visual interpretation, supervised classification and new remote sensing indices namely NDVI to enhance accuracy of land cover classification was used. The study employed normal image classification, and developing model from ECe vs NDSI to produce maps for salt affected soils during the period under study. Results showed that the study area has different land cover / land use units; i.e. Salt marches, Sabkha, Cropland, Grassland, Bare Land, Urban, Lakes, Sand dunes, and Hills. During the study period, a very severe land cover change has taken place as a result of mismanagement of land resources. Change detection at the study area explains the rapid increase of saline lakes and salt marshes, and the consequent hazards to the cultivated lands, roads and archeological sites and urban areas over the last three decades. These changes in land cover led to land degradation and water logging in many parts of the study area. In 1987, very slightly saline soil area was the largest in extent 34.1% of the total area whereas strongly saline, slightly saline and moderately saline soils covered 27.0%, 17.4% and 17.1%, respectively, of the total area. In 2017, strongly saline area increased as compared to 1987 and contributed for 39.1% of the total area, while very slightly, slightly saline and, moderately saline areas represented 18.9% and 13.2% and 14.9% respectively of the total area. Most of the salt affected areas were on shallow water table. The results indicate that long-term irrigation activities would affect agricultural potentiality of the area in the future. RS, GIS, NDVI and NDSI tools are efficient and feasible for detecting land cover/land use and salt affected areas from satellite images.
Keywords: Land cover/Land use; Land cover change; Salt affected soils; Remote Sensing; GIS; Salinity Index; Siwa Oasis.