Modelling the vegetation–climate relationship in a boreal mixedwood forest of Alberta using normalized difference and enhanced vegetation indices

Abstract

The present study compared two popular vegetation indices (VIs), the Normalized Difference Vegetation Index (NDVI) from the National Oceanic and Atmospheric Administration (NOAA) Advanced Very High Resolution Radiometer (AVHRR) and the Enhanced Vegetation Index (EVI) from the Moderate Resolution Imaging Spectroradiometer (MODIS) for monitoring the temporal responses of vegetation to climate over a boreal mixedwood forest of central-eastern Alberta. Linear and nonlinear regressions and an artificial neural network (ANN), called the Back Propagation Neural Network (BPNN), were used to elucidate the influence of climatic variables (precipitation, temperature, potential evapotranspiration and aridity index) on the VIs. These climate variables were used either individually or in certain combinations as predictors to these models. It was found that using multiple climate variables as predictors predicts the VI more accurately than using a single climate variable. Furthermore, the BPNN was found to be more efficient than the linear and nonlinear regressions at modelling the VI–climate relationship. For both VIs, BPNN using precipitation, temperature, potential evapotranspiration and aridity index as predictors modelled the VI–climate relationship very accurately in both the calibration and the validation stages. This study demonstrated a promising potential for monitoring the patterns of terrestrial vegetation productivity from climate variables in a boreal mixedwood forest of western Canada.