Sensitivity Analysis of Vegetation Reflectance to Biochemical and Biophysical Variables at Leaf, Canopy, and Regional Scales

The objective of this paper is to investigate the sensitivity of reflectance to the variation in biochemical and biophysical variables at leaf, canopy, and regional scales using a modeling approach. The results show that, at the leaf scale, the variations in chlorophyll a+b content, the leaf structure parameter, and the water content dominate the reflectance variance in the visible light (VIS), near infrared (NIR), and short-wave infrared (SWIR) regions, respectively. At the canopy scale, the sensitivity of reflectance to variation in the leaf structure parameter is very slight. For sparse foliage cover (leaf area index ), LAI is the most important variable to the canopy reflectance. As LAI increases, the sensitivity of reflectance to variation in LAI is reduced to a very low value. Moreover, chlorophyll a+b, dry matter, and water content control the variation of canopy reflectance in the VIS, NIR, and SWIR regions, respectively. At the regional scale, the sensitivity of reflectance to variation in vegetation variables is highly influenced by the mixed pixels. Thirty-six vegetation indices (VIs) are chosen in this paper to illustrate the scale dependence of the estimation accuracy of vegetation variables. The results show that the relationships between the VIs and the variables highly depend on the observation scale. For chlorophyll a+b content estimation, transformed chlorophyll absorption in reflectance index (TCARI), Blue Green pigment Index, leaf chlorophyll index (LCI), modified Normalized Difference (mND705), and Plant Biochemical Index at the leaf scale and canopy scale of and TCARI at the canopy scale of are highly related. The correlation between the indices and chlorophyll content in the regional scale is, however, much lower. For water content estimation, disease water stress index (DSWI), leaf water vegetation index 2 (LWVI_2), moisture stress index (MSI), normalized difference infrared index (NDII), normalized difference water index (NDWI), hyperspectral perpendicular vegetation index (RVI), SWIR water stress index (SIWSI), SR water index (SRWI), and water index (WI) are good choices at the leaf scale and canopy scale of , while at the canopy scale of and the regional scale, the correlation between the indices and water content is very low. For LAI estimation, VIs, including the Greenness Index, simple ratio (SR), Normalized Difference VI, modified soil-adjusted vegetation index (MSAVI), modified triangular vegetation index 1 (MTVI1), modified triangular vegetation index 2 (MTVI2), optimized soil-adjusted vegetation index (OSAVI), modified chlorophyll absorption ratio index 1 (MCARI1), modified chlorophyll absorption ratio index 2 (MCARI2), Enhanced VI, LAI Determining Index, renormalized difference vegetation index (RDVI), Spectral Polygon VI, Wide Dynamic Range VI, and triangular vegetation index (TVI), have high correlation with LAI at the canopy scale of while a low correlation at the canopy scale of and the regional scale.