Flooding: The effect of water depth on the spectral response of grass canopies

Vegetation indices generated from remotely sensed data have been widely used to estimate biophysical characteristics of natural vegetation and agricultural crops like aboveground productivity, leaf area index or absorption of the photosynthetically active radiation. However, in flooded environments, such as grasslands or rice crops, alterations in the spectral response of canopies may happen due to the presence of surface water. The objective of this study was to analyse these alterations in flooding environments. Spectral response at high resolution was measured in grass canopies inside a tank with varying levels of water between 0 and 25 cm, resulting in different above-water biomass proportions. Reflectance data were acquired using an OceanOptics lnc((c)) USB2000 visible and near-infrared spectroradiometer. Spectral indices like the Normalized Difference Vegetation Index (NDVI) were calculated for each flooded situation. As flooding level increased, absorption in red wavelengths decreased and reflectance in near infrared decreased. NDVI did not show differences between flooding levels lower than 5 cm, where more than 60% of biomass was above water. From 5 cm, NDVI decreased with the decreasing proportion of emerged biomass. These results evidence not only the alterations of spectral response data under flooded situations but also the conditions limiting vegetation index as a reliable estimator of plant biophysical characteristics. (c) 2006 Elsevier B.V. All rights reserved.