Correlation Between NDVI and Sentinel-1 Derived Features for Maize

Operational agricultural applications of remote sensing, such as crop monitoring or irrigation scheduling, often rely on the Normalized Difference Vegetation Index (NDVI) obtained from multispectral observations. Yet, cloud cover limits its availability, so the possibility to estimate it from SAR data is appealing, as it would enable a complementary monitoring of crops. The objective of this article is to evaluate the correlation of NDVI with several SAR features obtained from Sentinel-1 data over maize. Eighteen maize fields, located in the province of Navarre (Spain), were analyzed in two agricultural campaigns. Nine SAR features were evaluated, including: the backscattering coefficients in VH and VV polarizations, their ratio, product, sum and difference, as well as the Radar Vegetation Index (RVI), the Vertical Dual De-Polarization Index (VDDPI) and the Normalized Difference Polarization Index (NDPI). The correlations obtained in linear and dB units were compared, as well as the influence of temporal smoothing. The highest correlation was obtained with VH backscatter expressed in dB.