Satellite Data Applied to Hydrologic Models for Regional Watersheds: A Case Study, Apure Llanos, Venezuela.

Abstract

<p> Satellite data from GOES and LANDSAT where evaluated as a source of information for hydrologic distributed models applied to large watersheds. Three basins within the Llanos area of the Orinoco River basin, Venezuela, were selected as study areas. The specific objectives of the study were; (1) To test the applicability of meteorological satellite data for improving information on the temporal and areal distribution of precipitation,as well as estimates of amount over large areas. (2) To investigate photographic and digital LANDSAT data as a source of land surface information for hydrologic distributed models. The satellite and ground data used in this research were: (1) GOES WEFAX electrostatic facsimiles, (2) LANDSAT photographic and digital data, (3) Reports and maps on soil studies by Desarrollo Industrial Agricola C.A (1958) and Comerma and Luque (1971). </p> <p> The analysis of the data was carried out by visual analysis on the photographic products of GOES and LANDSAT using r·egular photo-interpretation techniques. GOES photographic data allowed the analysis of temporal and areal distribution of precipitation over large areas. Follansbe's (1973) method for estimating precipitation using satellite imagery was found potentially applicable to hydrologic distributed models. Variations to the method are suggested. </p> <p> The visual analysis of a single LANDSAT image allowed the mapping of broad land-cover classes and some soil characteristics in the study area. Analysis of the multidate imagery was found very useful in detecting seasonal and non-seasonal changes. </p> <p> Digital analysis of LANDSAT data was carried out on the Image 100 system at the Canada Centre for Remote Sensing in Ottawa. Contrast stretched images and breakpoint enhancement supervised and unsupervised classifications were produced.The results showed that LANDSAT digital analysis either by unsupervised or supervised classification can be used for the extraction of land-use/land-cover information for application in hydrologic distributed models. </p>