Monitoring crop biomass accumulation using multi-temporal hyperspectral remote sensing data

The estimation of the above-ground dry phytomass accumulation is important for monitoring crop growth, predicting potential yield, and estimating crop residues in the context of the carbon cycle. Hyperspectral remote sensing has been proven to be a very effective tool for the estimation of crop variables such as LAI, pigment and water content; therefore it is reasonable to expect that data from hyperspectral remote sensing can show great potential for monitoring crop biomass accumulation, either directly or indirectly through other variables. The objective of this study is to investigate the relationships between optical indices and either crop dry mass or height using multi-temporal, multi-field hyperspectral data. Using the Compact Airborne Spectrographic Imager (CASI), hyperspectral data were acquired in three deployments during the 2001 growing season over corn, soybean and wheat fields in the former Greenbelt Farm of Agriculture and Agri-Food Canada in Ottawa. High correlation was observed between the measured above-ground crop dry biomass and the other two parameters, crop height and leaf area index (LAI). The vegetation index MTVI2, calculated from hyperspectral images, was used to estimate the accumulated absorbed photo-synthetically active radiation (APAR) for the monitoring of crop biomass production. Both dry biomass and crop height were highly correlated with the accumulated APAR. For all the samples from the three dates, the coefficient of determination (R²) between the estimated APAR and crop dry mass was 0.95, 0.99 and 0.76, and R² between the estimated APAR and crop height was 0.90, 0.89 and 0.70 for corn, soybean and wheat, respectively. However, further analysis shows that the correlation between biomass increment and the accumulated APAR during a short period of time is much lower for wheat. This demonstrates that apart from APAR, biomass accumulation is affected by other factors as well