Assessing ecological risk through automated drainage extraction and watershed delineation

Decision makers are frequently involved in projects requiring ecological risk definition, which are inherent to biological conservation process. It is important to recognize these risks in order to invest wisely in the management and protection of biological resources. In this matter, Geographic Information System tools and remote sensing data have been used frequently as important components in planning and management of conservation units, Rabus et al. (2003), Valeriano et al. (2009) and Valeriano et al. (2010) stressed the advantages of using data that were gathered during the Shuttle Radar Topographic Mission (SRTM) for biological and geomorphologic purposes. For Brazil's national territory, the SRTM data were refined (Valeriano, 2008) and offered as free access on the TOPODATA Project website (http://www.dsr.inpe.br/topodata) where geomorphometric information (including elevation data) at a resolution of 30 m are provided. The aim of this paper is to demonstrate an example of how TOPODATA products have been applied in order to determine the ecological risk of the border of a Conservation Unit, located in the State of São Paulo—Brazil, in the Brazilian Atlantic Forest, using automated drainage network and watershed extraction. A comparison between SRTM, TOPODATA, and ASTER DEM was carried out, showing an advantage of TOPODATA drainage network product. The vectors generated using this data are more similar to the official drainage network vectors than the drainage network extracted using ASTER-DEM or SRTM. The network product generated using ASTER-DEM produced many commission errors and the one generated using SRTM produced a poor network, with generalized vectors, less detailed than the others. The results showed that using the TOPODATA Project‘s Digital Elevation Model (DEM) can provide important data for ecological analysis and significant additional information for decision making, regarding drainage networks and watershed features. The produced map for border ecological risk showed to fit perfectly to the field work analyses, produced in other works. Furthermore, the extracted watershed polygons might furnish important information unrevealing best conservation unit boundaries, which means more efficient management and best biological conservation results.