Skip to main content
ARS Home » Southeast Area » Stoneville, Mississippi » Crop Production Systems Research » Research » Research Project #446481

Research Project: Identify Agronomic Interventions Need for Climate-smart Agriculture Using Remote Sensing, Cloud Computing, Machine Learning and Big Data Approach

Location: Crop Production Systems Research

Project Number: 6066-22000-081-008-S
Project Type: Non-Assistance Cooperative Agreement

Start Date: Jul 1, 2024
End Date: Jul 1, 2029

Objective:
1. Develop high-resolution remote sensing-based data products related to corn and soybean phenology. 2. Estimate crop yields specifically for the Lower Mississippi River Valley (LMRV) watershed region.

Approach:
Understanding how phenology (the timing of recurring biological events) varies over space and time and its relationship with climatic factors is essential. This knowledge helps inform agronomic practices for climate adaptation and mitigation. Throughout the crop growing season (from 2010 to 2023), Sentinel-2 (10 m) and Landsat 7 and 8 (30 m) will be used to develop normalized difference vegetation index (NDVI) data for Lower Mississippi River Valley (LMRV) watershed region. A harmonized product combining Landsat-7, Landsat-8, and Sentinel-2 data will be created after applying cloud masking, different time-window filtering and smoothing algorithms. Phenology indices (such as the onset of greenup, peak NDVI time, and end of season etc) will be derived from the NDVI timeseries data. Phenology metrics will be evaluated against Moderate Resolution Imaging Spectroradiometer (MODIS) Land Cover Dynamics (MCD12Q2, Version 6.1) science data products. CPSRU research field data provide an additional layer of evaluation for the derived high resolution data products. Different maturity groups of soybeans have been assessed through strip trials experiments since 2021. Another set of strip trials involving corn have been evaluating N rates (100, 200, and 300 lb/ha) under both irrigated and rainfed condition. Historic yield monitor data from these fields, along with phenological metrics (including derived from UAV and canopy hyperspectra) will be used to develop yield estimation algorithms specific to the LMRV region. Combining soil, rainfall, temperature, and other climate and weather variables over the LMRV study area (2010 to 2023) will allow evaluating the impact of climate variables on yield and management decisions (eg. planting and harvesting date). Growing degree day (GDD) accumulation before the onset of greenness can estimate planting dates for corn and soybeans. Total days from planting to harvesting (senescence or dormancy) can identify the crop cultivar type (maturity group) planted within the region. Finally, considering different climate model projections, this study will explore optimum planting date, best performing cultivars (maturity group), and water and nutrient management strategies within the LMRV watershed.