Estimation of the fraction of canopy cover from multispectral data to be used in a water soil erosion prediction model

• David González Lanteri ^[1] ; Alfredo Huete ^[2] ; HoJin Kim ^[2] ; Kamel Didan ^[2]
  1. [1] Universidad de Concepción

     Universidad de Concepción

     Comuna de Concepción, Chile

     
  2. [2] University of Arizona

     University of Arizona

     Estados Unidos

     
• Localización: Gayana, ISSN-e 0717-6538, ISSN 0717-652X, Vol. 68, Nº. 2, 2004, págs. 239-245
• Idioma: inglés
• Enlaces
  • Texto completo
• Resumen

  • Using the 16 day MODIS (aboard the EOS Terra satellite) 250m NDVI and ground biophysical and spectral measurements we established simple relationships between these parameters and the canopy cover. The canopy cover is used in water erosion models to estimate the amount of soil loss under precipitation events and specific geographic conditions. Two transects, in the grassland part of the Walnut Gulch Experimental Watershed (WGEW) located near the town of Tombstone in Arizona, were established for ground data collection. Ground measurements were performed every 16 days, to coincide with the Terra Satellite overpass. Erosion, in desert environment is a contributing factor to soil degradation and subsequently desertification. Erosion is strongly related with canopy cover, soil parameters, topography, and climate variables. Although ground point estimates of canopy cover are usually used in erosion models, their temporal and spatial variability need to be accounted for. Using MODIS NDVI data, calibrated with field measurements, we were able to estimate the canopy cover using regression analysis. This technique is very simple and properly accounts for the spatial and temporal variability of the canopy cover. We tested this technique with the WEPP erosion model and we found it to be very valuable