In the context of wetland ecosystem management, a combination of approaches involving different time and spatial scales must be applied. One primary driver of wetland degradation is agricultural expansion at watershed scale. Wetlands have undergone several hydrological and biological changes as a consequence of increased water inputs from agricultural drainage off the watershed. For the establishment of suitable wetland ecological indicators, watershed scale studies focusing on pressures influencing ecosystem dynamics are necessary. Specific enhanced methods for watershed modelling, wetland mapping and land cover assessment are thus essential tools for wetland monitoring and management. Watershed draining to the Marina del Carmolí semiarid wetland in Murcia Region (SE Spain) was delimited using a digital elevation model. Map algebra operations were applied on the elevation model of the Campo de Cartagena coastal plain to reinforce existing drainage network and to force flow accumulation from all draining areas around wetland perimeter to converge into a single point within the wetland area. Watershed delineation was thus improved.A land use/land cover map of the Campo de Cartagena was then obtained for year 2008-09 by means of supervised classification of Landsat images. A set of four spectral indices were calculated and included in the classification analysis using a combination of bands in order to better discriminate vegetation, water bodies, infrastructures and bare soil. An enhanced classification procedure based on maximum likelihoodand random reselection of train areas was applied. Object-based analysis of the Landsat scenes based on automatic image segmentation diminished the occurrence of isolated pixels in the classification. The proposed classification methodology showed great accuracy, thus improving the results of traditional classification techniques. Wetland plant communities in Marina del Carmolí wetland were mapped in 2008 by means of remote sensing techniques using satellite and airborne images. Characteristic plant communities were first characterized by combining fieldwork sampling of plant species and multivariate analysis. Georeferenced sampling units were further used as training areas for supervised image classification of plant communities. Maps obtained showed great accuracy. However, sensors are adequate for different applications. The proposed set of GIS methodological tools contributes to improve the study of wetland plant as indicators, the mapping and future monitoring of watershed land cover classes, and the study of wetland plant community changes over time. All pieces of software used in the study are free and mainly open source programs, which make it an inexpensive and universal methodology.