IWRA Proceedings

< Return to abstract list

W²DVI - A WATERSHED INDEX BASED ON GIS AND SATELLITE IMAGE DATA

IWRA World Water Congress 2003 Madrid Spain
IWRA WWC2003 - default topic
Author(s): Marcos FERREIRA

Marcos FERREIRA

Institute of Geosciences, Campinas State University – UNICAMP, Campinas – SP, Brazil macferre@ige.unicamp.br macferre@uol.unicamp.br



Keyword(s): watershed index; GIS, vulnerability to pollution; water quality; satellite image, Brazil
Article: PDF

Abstract

Indexes reflecting vulnerability to pollution of watersheds have been developed and used by environmental agencies. The indices are based on Theory of Multiple Atributtes and used in decision analysis and watershed planning. A watershed index is formed by addition of information about indirect water quality parameters, translating it in a single index. Water quality parameters are integrated in a linear sum of functions, to produces a single index of watershed as a whole. The quantities of parameters depending on temporal and spatial dimensions and reflecting environmental factors influencing on water quality of the watershed. Numerous problems have been highlighted concerning some characteristics of this type of index, like delimitation of spatial and temporal units, choice of weights and use of subjective indices. In sense to minimize the above cited limitations of some watershed indicators, we present in this paper a new watershed index based on GIS and satellite image data: the Watershed Water Degradation Vulnerability Index – W²DVI.

The W²DVI uses as database the following watershed parameters: urbanization coefficient (UBC), vegetation index (VGI) and topographic index (TPI). The UBC coefficient reflects constructed and impervious areas density existing in the watershed. The VGI index indicates green biomass over watershed terrain, and identifies the relative infiltration capacity and erosion suitability of watershed. The VGI index is calculated applying the NDVI vegetation index algorithm in TM-Landsat 7 satellite images. The UBC index is obtained by digital classification of land use of color composite images. The TPI index is related to the flood risk, soil loss and potential energy of watersheds, and uses the following watershed morphometric data: first order stream density, relief ratio and circularity index of watersheds

The indexes values are equalized on a scale of common values, with intervals between 0 and 1 interval. The scale of values of the three indices are used to construct a three-dimensional space of data in a cubic format, representing the W²DVI index. This cube is divided in eight blocks, each related to a category of watershed degradation vulnerability. The watershed are classified according to the block to they correspond, depending on the UBC, VGI and TPI values. The W²DVI index methodology was applied in Sao Francisco watershed, located in the northeast region of Brazil, and is associated to a water quality monitoring program in semi-arid areas, supported by National Water Agency of Brazil.