The Department of Meteorology of Sri Lanka receives NOAA-AVHRR images daily. Normalized Difference Vegetation Index (NDVI) can be deduced from this raw satellite data to estimate the photosynthetic activity of chlorophyll. Surface temperature and surface albedo, values based on the AVHRR measurements are used to solve the land surface energy balance equation following the SEBAL methodology. The conversion of absorbed photosynthetical active radiation into biomass production is done through light use efficiency that depends on vegetation water stress. The month-to-month above ground dry biomass production of rice, tea, rubber and coconuts is computed. Values of accumulated biomass-net primary production-have been compared against agricultural census data and the harvest indices have been calibrated. The relationship of yield to remotely sensed biomass production is not statistically significant on a monthly basis because biomass production is related to antecedent growing conditions. The ratio between harvestable, yield and accumulated biomass on an annual basis was used to obtain yield statistics for rice, tea, rubber and coconut from remotely sensed data. Research in Sri Lanka should give more attention to the determination of harvest indices and light use efficiencies. The model is 88% accurate at predicting paddy yield. The frequency distributions of yield and yield gaps are identified. Rice has a yield gap of 2835 kg ha(-1) per year. There seems to be more room to improve the agricultural production of rice and tea, than for rubber and coconut. (C) 2002 Elsevier Science B.V. All rights reserved.
