Estimating wheat yield: an approach for estimating number of grains using cross-polarised ENVISAT-1 ASAR data

Abstract

In this paper an attempt to model wheat yield is made by exploiting characteristic interaction of cross-polarised SAR with wheat crop. SAR backscatter from a crop field is affected by the density, structure, volume and the moisture content of various components of plant (viz. head, stem, leaf) alongwith soil moisture. Hence, to effectively handle the influence of each of these components of the plant on SAR backscatter, a plant parameter, termed as Interaction Factor (IF) is conceptualised by combining volume, moisture, height for each of the component and density of plant. For this purpose, detailed experiment over farmers' fields was carried out in synchrony with SAR acquisition involving in-depth measurements on volume, moisture content and height of various components of wheat plant, number of grains, plant density and soil moisture. Stepwise regression analysis revealed that IFHead significantly affects the shallow incidence angle, cross-polarised C-band SAR backscatter. IF_Head is also highly correlated to the number of grains. This is attributed to the fact that parameters of the wheat head from which IF_Head is calculated, namely moisture, volume and height, determine eventual number of grains. The study offers an approach for estimating wheat yield by retrieving number of grains from shallow incidence angle cross-polarised SAR data.