Leaf-shedding phenology in tropical seasonal forests of Cambodia estimated from NOAA satellite images

Forest seasonality is an important element to ecosystem functions. Eco-hydro models describing the Indochina bioregion under the seasonal tropical climate need regional phenological information about leaf dynamics. We examined the utility of remote sensing technology to leaf phenological research in Cambodian lowland forests. For this purpose, we aimed to detect any difference in leaf-shedding phenology between evergreen forests and deciduous forests. We analyzed the NOAA/AVHRR normalized differential vegetation index (NDVI) obtained from May 2001 to April 2002. The Local Maximum Fitting (LMF) processing combines the time series filtering and the functional fitting was used for creating cloud/noise-free 10- day composites 1.1 km pixel data. Firstly, we estimated seasonal changes in the NDVI dominated by lowland evergreen forests and by lowland deciduous forests. Secondary, we identified local minimum points and the antecedent local maximum points of the fitting trigonometric function curve to each pixel as indicators of leaf-shedding events. The heterogeneous seasonal changes in the NDVI were well detected. Deciduous forests demonstrated drastic and uniform leaf phenology; while evergreen forests did spatially and temporally heterogeneous one during the dry season. It indicated the difficulty in getting information of regional forest seasonality; thus it displayed the utility of remote sensing for phenological investigation. Leaf-flushing was detected during the dry season both in evergreen forests and in deciduous forests. It suggested that leaf phenology was not completely governed by drought stress. Leaf-shedding and leaf flushing in twice within a single evergreen pixel was the most striking findings in this study. In summary, remote sensing technology was of great service to getting phenological information that was considerably different between evergreen forests and deciduous forests distributed in the seasonal tropical zones.