Estimating Crop Leaf Area Index at Different Spatial Scales Based on Statistical Distribution Theory

摘要

Estimating leaf area index (LAI) of crop canopies with regional optical remote sensing observations is the normal way for crop growth monitoring and yield estimation in the research field of agriculture quantitative remote sensing. One fundamental difficulty for multi-scale crop LAI estimation is the spatial scale effects of LAI, due to the spatial heterogeneity in crop canopies and the non-linearity of LAI estimation model. In this study, we proposed a new data upscaling scheme to estimate crop LAI at different spatial scales based on statistical distribution theory, aiming to reduce the influences of scale effects and enhance the LAI estimation precision. Numerical experiments based on the observations in farm underlying surface and suburban underlying surface were conducted to compare and analyze the effectiveness and applicability of the normal upscaling scheme and the new proposed scheme. Experimental results indicated that the estimated multi-scale LAI based on the new proposed data upscaling scheme were significantly more reduced in scale effects and more enhanced in estimation precision under non-uniform underlying surface.