GIS and RS integration: Application of geostatistical techniques and environmental changes in the coastal zone in Kenya

摘要

Understanding the dynamics of land cover change has increasingly been recognized as one of the key research imperatives in global environmental change research. Scientists have developed and applied various methods in order to find and propose solutions for many environmental world problems. From 1986-1995 changes in Kenya coastal zone landcover, derived from the post-classification TM images, were significant with arid areas growing from 3% to 10%, woody areas decreased from 4% to 2%, herbaceous areas decreased from 25% to 20%, developed land increased from 2% to 3%. In order to generate the change probability map as a continuous surface using geostatistical method-ArcGIS, we used as an input the Generalized Linear Model (GLM) probability result. The results reveal the efficiency of the Probability-of-Change map (POC), especially if reference data are lacking, in indicating the possibility of having a change and its type in a determined area, taking advantage of the layer transparency of the GIS systems. Thus, the derived information supplies a good tool for the interpretation of the magnitude of the land cover changes and guides the final user directly to the areas of changes to understand and derive the possible interactions of human or natural processes.