Relationship between leaf area index of wheat crop and different spectral indices in Punjab

Abstract

This study was carried out to estimate the suitability of IRSP6LISS-III data for estimating LAI in wheat crop in Punjab conditions. LAI of 45 placescovering three agro-climatic regions of Punjab were surveyed where two-thirds of the data (30 cases) were allocated by random sampling to the modeling set and one-third (15 cases) to the validation set. The empirical relationships between wheat-LAI and satellite acquired spectral reflectance data were studied using correlation analysis, linear and non-linear regression analyses. Useful spectral features included single band reflectance inIR, logarithmic transformation of IR band reflectance and several spectral vegetation indices like RDVI, DVI, NDVI, SR, MSAVI2 and MSI. Amongst the LISS III bands, relationship between IR reflectance and the LAIwas the strongest (in polynomial function, r = 0.86; RMSE = 0.31i.e. 7.4 % of observed mean). However, LAI could be predicted most accurately by RDVIusing linear function (R2(r)= 0.78 (0.88); RMSE, 0.27 i.e. 6.3% of observed mean). Keeping in view the high accuracy of estimates, 24 regression models developed through this study can be employed for wheat LAI estimation in the Punjab region of India.

Keywords

• Wheat, LAI
• Spectral reflectance
• Vegetation-Indices
• Satellite
• Punjab