Detection of Land Use Change In Lake Maumelle Watershed Critical Management Area: Image Processing and GIS Integration Approach

Land use in the Lake Maumelle watershed is expected to undergo significant changes in the next several decades, with residential developments replacing forest in many areas. These land use changes have the potential to increase pollutant loads and degrade water quality in the lake. Detecting the land use and land cover changes is therefore a critical requirement for effective land management. Therefore, the research on the cost-effective methodology of detection of land use change is critical and of extreme significance.;The first objective of this study is to develop an effective image classification method using very high resolution remote sensing images. Classification is the key to successful land use change detection through time. Therefore, we proposed a novel multi-feature fusion, object-based classification method using very high resolution remote sensing data for the study area. A total of 42 tiles of one feet resolution color infrared and natural color aerial photos that cover the study area were processed using the proposed classification method. In spite of the complexity and qualities of the very high resolution aerial photos, high classification performance has been achieved.;The second objective of this study is to detect land use changes within the study area. Urban development including roads and buildings and an increase of bare ground within the area of three tiles were identified by post-classification comparison approach. A geoprocessing model was developed to extract image objects in pixels that represent these geographic features. An existing GIS database was used to estimate the true or false changes in actual pixels and land use changes from-to transition. Both pessimistic and optimistic overall accuracies as global accuracy assessment were estimated from the overall classification accuracies.;The research project has proved the effectiveness of the proposed multi-feature fusion, object-based classification method on the ArcGIS platform using very high resolution remote sensing data. The integration of image processing and GIS technologies on a single GIS platform can be an effective approach in the detection of land use changes and updating existing land use and land cover GIS database.