{Reference Type}: Journal Article
{Title}: Integrating remote sensing classification techniques for land use mapping in semi-arid regions: a case study of the Tamlouka basin, Algeria.
{Author}: Bouchagoura L;Medjani F;Zahi F;Djidel M;
{Journal}: Environ Monit Assess
{Volume}: 196
{Issue}: 7
{Year}: 2024 Jun 8
{Factor}: 3.307
{DOI}: 10.1007/s10661-024-12757-1
{Abstract}: Worldwide, the majority of countries are actively devising strategies to address the challenges associated with unregulated and unmanageable development, the decline in environmental quality and the depletion of valuable agricultural land. This has led to a growing emphasis on understanding land use and land cover. In order to determine a better land use policy, legislators and planners need to know the current distribution of agricultural and urban lands, as well as information about changes in their proportions. Our approach combines data centred on main four themes-geology, slope gradient, hydrographic network and land use-in order to exploit classifier complementarities in our targeted agricultural study area of Tamlouka Basin, Algeria. Landsat 8 OLI-TIRs multispectral imagery and Shuttle Radar Topography Mission (SRTM-1arc v3) were used experimentally for classification and Digital Elevation Model (DEM) analysis. The classification's accuracy is confirmed by comparing the results of the decision tree classification with the validation samples. Results of the combination of several maps of classifications from the different methods show that the Tamlouka alluvial plain, having an area of 19,300 ha and an average slope gradient of less than 2°, drains the elevated reliefs that surround it via hydrographic network. The plain occupies 37% of the total basin area, with over of 60% being used for crop cultivation, regardless of fallow land areas in agricultural rotation at that time. The slope has been identified as a crucial factor determining land use patterns in the study area. This result can be used in prospective watershed management.