Assessment of runoff response to landscape changes in the San Pedro subbasin (Nayarit, Mexico) using remote sensing data and GIS.

Results on runoff estimates as a response to land-use and land-cover changes are presented. We used remote sensing and GIS techniques with rainfall time-series data, spatial ancillary information, and the curve-number method (NRCS-CN) to assess the runoff response in the San Pedro subbasin. Thematic maps with eight land-cover classes derived from satellite imagery classification (1973, 1990, and 2000) and hydrologic soil-group maps were used as the input for the runoff calculation. About 20% to 25% of the subbasin landscape has changed since 1973, mainly as consequence of the growth of agriculture. Forest is the main cover, although further analyses indicate that forest is degrading from good to poor conditions when evaluated as a function of the spectral response. Soils with low infiltration rates, classified as the hydrological soil-group "C", were dominant in the area (52%). The overlaying of all the hydrological soil groups with the land-use map produced a total of 43 hydro-group and land-use categories for which runoff was calculated using the curve-number method. Estimates of total runoff volumes (26 x 10(6) m3) were similar for the three dates analyzed in spite of landscape changes, but there were temporal variations among the hydro-group and land-use categories as a consequence. Changes are causing the rise of covers with high runoff potential and the increase of runoff depth is expected, but it can be reversed by different management of subbasin hydro-groups and land-use units.