Free advanced modeling and remote-sensing techniques for wetland watershed delineation and monitoring

摘要

Watershed scale studies focusing on hydrological pressures influencing freshwater ecosystem dynamics are necessary for the establishment of suitable wetland ecological indicators. Enhanced and reproducible methods for watershed modeling and land-cover assessment are thus essential tools for wetland monitoring and management. However, few integrated studies propose advanced open source methodologies for watershed modeling and assessment. In this study, a set of GIS methodological tools was applied and further developed in order to delineate wetland watersheds and map their land-cover changes over time. Watersheds draining to 11 semiarid Mediterranean saline wetlands were delimited and map algebra operations were applied on the digital elevation model in the Campo de Cartagena coastal plain to enhance watershed delimitation, land-use/land-cover maps of wetland watersheds were obtained for years 1987 and 2008 by means of supervised classification of Landsat images. A set of four spectral indices was included in the classification analysis using a combination of bands in order to improve the discrimination of vegetation, water bodies, infrastructure, and bare soil. An iterative classification procedure based on maximum likelihood and random selection of training areas was applied. Contextual information based on automatic image segmentation of Landsat scenes was also included as ancillary layers. Watershed areas obtained ranged from 70 to 17,000 ha and delineation was improved in the Campo de Cartagena coastal plain. The proposed image classification methodology showed high accuracies and improved standard classification techniques. The proposed methodology is based on free and open source tools, which makes it broadly applicable.