Progress 08/15/24 to 08/14/25
Outputs
Target Audience:Target audiences are farmers, agricultural engineers,agronomists, and agricultural extension professionals. Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided?This project has provided a valuable opportunity for the training and professional development of the postdoctoral research fellow leading the software implementation and analytical components of the study. Through this work, the fellow has gained experience in programming, data analysis, satellite image processing, and scientific writing, significantly enhancing their technical expertise and research capacity. Additionally, participation in fieldwork has further strengthened the fellow's practical skills and broadened their experience in agricultural research. How have the results been disseminated to communities of interest?We currently have two open-access scientific papers published in Computers and Electronics in Agriculture (https://doi.org/10.1016/j.compag.2025.110086, https://doi.org/10.1016/j.compag.2025.110676), contributing to disseminating the results of the project. In addition, as part of our efforts to develop efficient tools, we developed two open-source Python libraries and made them available to the scientific community. The first library (https://github.com/lbferreira/fastnanquantile) offers an efficient method for computing quantiles in satellite image time series, with over 10,000 downloads, indicating its widespread adoption within the community. The second library (https://github.com/lbferreira/ezgrid) provides an easy alternative for creating sampling points within a field based on various strategies. Lastly, we presented the project results at two conferences (ASABE 2025 Annual International Meeting and AI in Agriculture & Natural Resources Conference 2025), where the audience consisted of agricultural-related professionals and academics. What do you plan to do during the next reporting period to accomplish the goals?In the next reporting period, we will advance the processing pipeline to extract and estimate crop phenology stages and associated climate variables at the field level (objective 2). These phenological metrics will also be evaluated across multiple years. Validation of the crop phenology stages and vegetation indices will be conducted using in situ data collected during the 2025 and 2026 summer seasons (objective 3). We will continue to refine and test our satellite-based agricultural monitoring framework to improve its ability to generate accurate and actionable insights into crop growth dynamics, while maintaining computational efficiency (objective 4).
Impacts
What was accomplished under these goals?
We have successfully automated the extraction of crop field boundaries across the 17 target states by integrating processing workflows on Google Earth Engine and High-Performance Computing infrastructure (Objective 1). The resulting dataset consists of individual polygon vectors delineating crop field boundaries for the year 2023. These boundaries were derived using temporal composites of vegetation indices from Sentinel-2 imagery and the Meta AI Segment Anything model. To assess the scalability and accuracy of our approach, validation was conducted in two 10×10 km agricultural regions in the United States and six additional locations worldwide with manually delineated field boundaries. We also made extensive progress in developing a harmonized Landsat-Sentinel-2 (HLS) time series processing pipeline at the field scale (Objective 2). We collected raw HLS surface reflectance data for each field and implemented gap-filling and interpolation techniques to generate daily synthetic vegetation index time series. These per-field time series will enable the extraction of crop phenology dates. During the summer of 2025, we have been conducting in situ data collection in five agricultural fields at the Mississippi State University farm to validate vegetation indices and crop phenology stages.
Publications
- Type:
Peer Reviewed Journal Articles
Status:
Published
Year Published:
2025
Citation:
Ferreira, L. B., Martins, V. S., Aires, U. R., Wijewardane, N., Zhang, X., & Samiappan, S. (2025). FieldSeg: A scalable agricultural field extraction framework based on the Segment Anything Model and 10-m Sentinel-2 imagery. Computers and Electronics in Agriculture, 232, 110086.
- Type:
Peer Reviewed Journal Articles
Status:
Published
Year Published:
2025
Citation:
Aires, U. R., Martins, V. S., Ferreira, L. B., Huang, Y., Heintzman, L., & Ouyang, Y. (2025). Impact of sampling techniques on crop type mapping using multi-temporal composites from Harmonized Landsat-Sentinel images. Computers and Electronics in Agriculture, 237, 110676.
|