Estimation of winter wheat LAI using hotspot-signature vegetation indices

Leaf area index (LAI) is often retrieved from mono-angle remote sensing, the main weaknesses of which are the saturation limits at intermediate values of LAI and lacking structure information. Given the above deficiencies, two new vegetation indices, Soil Adjusted Hotspot-signature Vegetation Index (SAHVI) and Enhanced Hotspot-signature Vegetation Index (EHVI), were proposed for a better quantitative estimation of LAI. To obtain the new indices, we adjusted at-nadir Enhanced Vegetation Index (EVI) and Soil Adujsted Vegetation Index (SAVI) to incorporate Hot-Dark Spot (HDS) index respectively that represents spatial distribution pattern of leaves. Next, the red, near-infrared, blue and green bands were exploited to calculate the respective HDS indices. Four HDS indices were compared for correlation with increasing LAI and those relatively more sensitive to LAI variability were then selected to construct SAHVI and EHVI. At last, the linear relationships between the new indices and LAI were investigated based on in-situ measurements of bi-directional reflectance and LAI from winter wheat. It was found stronger correlations between SAHVI, EHVI, NHVI and LAI than between EVI, SAVI, NDVI and LAI. Better resistance to saturation limits were both observed for SAHVI and EHVI.