A new GIS-based model for automated extraction of Sand Dune encroachment case study: Dakhla Oases, western desert of Egypt

The movements of the sand dunes are considered as a threat for roads, irrigation networks, water resources, urban areas, agriculture and infrastructures. The main objectives of this study are to develop a new GIS-based model for automated extraction of sand dune encroachment using remote sensing data and to assess the rate of sand dune movement. To monitor and assess the movements of sand dunes in Dakhla oases area, multi-temporal satellite images and a GIS-developed model, using Python script in Arc GIS, were used. The satellite images (SPOT images, 1995 and 2007) were geo-rectified using Erdas Imagine. Image subtraction was performed using spatial analyst in Arc GIS, the result of image subtraction obtains the sand dune movement between the two dates. The raster and vector shape of sand dune migration was automatically extracted using spatial analyst tools. The frontiers of individual dunes were measured at different dates and movement rates were analyzed in GIS. The ModelBuilder in Arc GIS was used in order to create a user friendly tool. The custom built model window is easy to handle by any user who wishes to adapt the model in his work. It was found that the rate of sand dune movement ranged between 3 and 9 m per year. The majority of sand dunes have a rate movement between 0 and 6 m and very few dunes had a movement rate between 6 and 9 m. Integrating remote sensing and GIS provided the necessary information for determining the minimum, maximum, mean, rate and area of sand dune migration.