Data-model fusion for improving LAI mapping: a case study over China's land mass

A simple data-model fusion method is developed to improve leaf area index (LAI) mapping using satellite data. The objective is to overcome two issues with satellite-derived LAI maps: (1) optical remote sensing data are often seriously affected by the atmosphere due to clouds, and in some areas no reliable data are obtained in the whole growing season, and (2) seasonal variations in conifer LAI derived from satellite data are often distorted by the seasonal variations in leaf greenness (pigments), the background vegetation and snow cover, etc., and the derived LAI reflects the overall greenness rather than the actual forest leaf area present in a pixel. These shortcomings of satellite measurements can be greatly alleviated when an ecological model is used to simulate the LAI in the absence of reliable remote sensing data and to estimate the seasonal variation of LAI according to ecological principles. The usefulness of this fusion method is demonstrated through improving a China-wide LAI map series in 10-day intervals at 1 km resolution using Satellite Pour l'Observation de la Terre (SPOT) VEGETATION (VGT) data.