Using NASA earth observations and Google Earth Engine to map winter cover crop conservation performance in the Chesapeake Bay watershed

Authors: THIEME, A. - Goddard Space Flight Center; YADAV, S. - Goddard Space Flight Center; ODDO, P.C. - Goddard Space Flight Center; FITZ, J.M. - Goddard Space Flight Center; MCCARTNEY, S. - Goddard Space Flight Center; KING, L.A. - US Department Of Agriculture (USDA); KEPPLER, J. - Maryland Department Of Agriculture; McCarty, Gregory; HIVELY, W.D. - Us Geological Survey (USGS)

Submitted to: Remote Sensing of Environment
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 5/5/2020
Publication Date: 7/15/2020
Citation: Thieme, A., Yadav, S., Oddo, P., Fitz, J., McCartney, S., King, L., Keppler, J., McCarty, G.W., Hively, W. 2020. Using NASA earth observations and Google Earth Engine to map winter cover crop conservation performance in the Chesapeake Bay watershed. Remote Sensing of Environment. 248:111943. https://doi.org/10.1016/j.rse.2020.111943.
DOI: https://doi.org/10.1016/j.rse.2020.111943

Interpretive Summary: The Maryland Agricultural Water Quality Cost-Share (MACS) program, managed by the Maryland Department of Agriculture (MDA), provides cost-share grants to farmers to help offset the costs associated with the implementation of certain Best Management Practices (BMPs) that address water quality concerns on Maryland farms. The Maryland Cover Crop Program, established in 1997 as a component of MACS, incentivizes farmers to grow winter cover crops to reduce nutrient and sediment loss from farmland. The U.S. Department of Agriculture - Agricultural Research Service (USDA–ARS) and the U.S. Geological Survey (USGS) have collaborated with the MDA since 2006, developing remote sensing techniques to assess winter cover crop performance in Maryland. Using remote sensing imagery to evaluate all fields enrolled in the program will allow the MDA to assess winter cover crop performance in both the winter and spring months, and to potentially replace spot checks with imagery-based verification of management dates and practices. If successfully implemented, a statewide measurement and monitoring program for cover crops would allow analysis of the effect of different agronomic management practices on environmental outcomes. This study developed Google Earth Engine (GEE) scripts to create composite seasonal satellite reflectance indices covering the State of Maryland, and to combine that information with MACS cost-share enrollment field boundary data to produce a winter and springtime evaluation of winter cover crop performance for all enrolled fields falling within three test counties on the Eastern Shore, and one test county in western Maryland. To achieve these objectives the team used Earth observations from the NASA Landsat and ESA Sentinel programs, processed in GEE. Implementation of timely, well-calibrated Earth observing satellite data analyses to facilitate calculation of winter cover crop effectiveness will expedite and enhance key conservation management practices at the MDA. The tool can be modified for different seasonal cutoffs, utilize new sensors to capture phenology in winter and spring, and scale to larger regions for use in adaptive management of winter cover crops planted for environmental benefit. It is expected that this tool will now be used operationally by the MDA in implementation of their ongoing winter cover crop program.

Technical Abstract: Planting cover crops provides a range of benefits to farmers including increased soil health, reduced soil erosion, and improved water quality. Winter cover crops such as barley, rye, and wheat help to improve soil structure by increasing porosity, aggregate stability, and organic matter while reducing the leaching of agricultural nutrients into waterways. They are an essential component of conservation management practices such as those promoted by the Maryland Department of Agriculture (MDA) and the Chesapeake Bay Program Partnership. The MDA oversees an agricultural cost-share program that provides grants to farmers to offset seed, labor, and equipment costs associated with planting winter cover crops. The effectiveness of mitigating soil and nutrient loss varies based on factors including cover crop species, planting date, planting method, nutrient inputs, temperature, and precipitation. The U.S. Geological Survey (USGS) and the U.S. Department of Agriculture – Agricultural Research Service (USDA-ARS) have worked in partnership with the MDA to develop satellite remote sensing techniques for measuring cover crop performance. The MDA has developed the capacity to digitize field boundaries for all fields enrolled in their cover crop programs (approximately 26,000 fields per year) and has requested assistance to apply remote sensing tools to analyze cover crop performance at statewide scale. A NASA DEVELOP team used the Google Earth Engine (GEE) cloud computing platform to automate the acquisition, compositing, and extraction of wintertime vegetation data from Landsat 5, 7, 8, and Sentinel-2 imagery. The team calibrated cover crop performance models using linear regression between satellite vegetation indices and USGS / USDA-ARS ground truth data collected on farm fields within four Maryland counties from 2006 to 2012 (1,296 samples). Performance measures that were correlated to satellite-derived Normalized Difference Vegetation Index (NDVI) included the natural logarithm of cover crop biomass (p<=0.01, R2=0.60) and percent vegetative ground cover (p<=0.01, R2=0.68). Using these models, the GEE scripts composited seasonal maximum NDVI values for each boundary polygon for all fields enrolled in four counties of the MDA cover crop cost-share program, and calculated field-specific performance estimates for the winter and spring seasons of three enrollment years (2014-15, 2015-16, and 2017-18). Example results from Winter 2014-15 demonstrated rye and barley fields had higher biomass than wheat fields, and that early planting, along with planting methods that increase seed-soil contact, increased performance. By combining the capabilities of GEE for large scale image processing with the MDA’s geospatial enrollment dataset, the team created a scalable cover crop performance analysis. It is expected to have multiple applications for the MDA including the rapid identification of underperforming cover crop fields, the ability to distinguish fields with high or low biomass, and information to evaluate the environmental outcomes of various agronomic management strategies. The tool can be modified for different seasonal cutoffs, utilize new sensors to capture phenology in winter and spring, and scale to larger regions for use in adaptive management of winter cover crops planted for environmental benefit.