MAPPING CROP PHENOLOGY USING VENµS OBSERVATIONS OVER MARYLAND EXPERIMENTAL SITES

Crop growth stages are critical in scheduling irrigation, fertilization, and harvest operation. For decades remote sensing data have been used to extract vegetation phenology by utilizing mathematical functions to describe crop growth and development. However, results somewhat depend on the availability of earth observations, especially during critical growth stages. Vegetation phenology products, at kilometer resolutions, are commonly available today. Nevertheless, these resolutions are too coarse to monitor crop growth at the field scale and difficult to validate at such coarse resolutions. Moreover, remote sensing phenology and crop growth stages represent different crop properties so they are not interchangeable The VENµS observations with 2-day repeat and 5-m resolution provide an opportunity to map crop phenology at field scales since the fine spatial resolution is more suitable to be validated using ground observations and the PhenoCam photos. In this presentation, we demonstrate a near real-time algorithm for detecting crop emergent dates. Our results using VENµS images over the Choptank River watershed in Maryland, USA, show that crop emergence can be reliably mapped. Different green-up dates of the forest, corn, and soybeans match the PhenoCam observations and agree to the state-level crop progress reports from the USDA National Agricultural Statistics Service (NASS). Separating single and double-crop soybeans, which is a challenge in traditional crop classification, were differentiated in green-up dates. Based on these findings, this study indicates future opportunities to map crop growth stages at field scales using high temporal and spatial resolution remote sensing data.