•  
  •  
 

Keywords

high-cadence, maize, phenology, classification

Abstract

Crop phenology can be defined as the study of biological processes such as emergence, flowering, and senescence that are associated with and affected by environmental growing conditions. The ability to reliably detect crop phenology and its spatial-temporal variability is critical for farmers, policymakers, and government agencies, since it has implications for the entire food chain. Currently, two methods are the most used to report crop phenology. Land surface phenology provides insight into the overall trend, whereas USDA-NASS weekly reports provide insight into the development of specific crops at the regional level. High-cadence earth observations may be able to improve the accuracy of these estimations and bring more precise crop phenology classifications closer to what farmers need. The use of robust classifiers (e.g., random forest, RF) to manage large data sets is required to successfully achieve this goal. This study compared the output of an RF classifier model using weather, two different satellite sources (Planet Fusion; PF and Sentinel-2; S-2), and ground truth data to improve maize (Zea mays L.) crop phenology classification during the 2017 growing season in Kansas. Our findings indicate that high-cadence (PF) data can enhance crop classification metrics (f1-score = 0.94) as compared to S-2 (f1-score = 0.86). This study emphasizes the significance of very high temporal resolution (daily) earth observation data for agricultural crop monitoring and decision-making tools.

Creative Commons License

Creative Commons Attribution 4.0 License
This work is licensed under a Creative Commons Attribution 4.0 License.

Share

COinS