CNN-Based Downscaling of SIF Data for Field-Scale Cotton Yield Estimation

This study introduces a novel approach for estimating cotton yield at the field scale by leveraging the power of convolutional neural networks (CNNs). Recognizing the limitations of coarse-resolution SIF data, we propose a CNN-based framework to downscale SIF products from a spatial resolution of 0.05ﾰ to a highly detailed 0.0005ﾰ. This significant enhancement in resolution allows for the accurate estimation of cotton yield at the field level. Our method addresses the challenges posed by the spatial mismatch between readily available SIF data and the need for fine-grained yield information, providing a valuable tool for precision agriculture and crop management in the cotton industry.