Identification of crop type and areal extent is a challenge, made difficult by the variety of cropping systems, including crop types, management practices, and field sizes. The goal of this project is to evaluate the integrated use of Landsat and Sentinel 2 data in quantifying cultivated area by major commodity crop type. The first evaluation objective is correct identification of crop type. MODIS data, due to its high image cadence, are appropriate for and have been extensively used for mapping crop. Using MODIS as a high temporal reference, an assessment of combined Landsat and Sentinel 2 observations in identifying crop type will be performed. For any given crop type, its extent is required in estimating production. RapidEye data represent a high temporal, high spatial resolution imaging capability over limited areas. RapidEye data will be used to evaluate area estimation of selected crop types using combined Landsat and Sentinel 2 data. Results will inform users of the potential value of Landsat and Sentinel 2 data to identify and map the extent of key commodity crops for a variety of landscapes, including wheat, corn and soybean. The project’s ultimate goal is to demonstrate the viability of accurate large area crop type mapping using medium spatial resolution Landsat and/or Sentinel 2 data.
A multi-resolution approach to national-scale cultivated area estimation of soybean- In Review
Presenting the results from our generic multi-resolution approach to sample-based crop type area estimation at the national level using soybean as an example crop type. Historical MODIS (MODerate resolution Imaging Spectroradiometer) data were used to map percent soybean cover for stratifying USA and Argentina into strata of low, medium and high soybean cover for 20km x 20km blocks. Twenty-five samples per stratum were selected and current year Landsat data used to map soybean area within each sample block. National-scale soybean cultivated area was estimated from the Landsat mapped blocks using current year percent soybean from MODIS in a regression estimator procedure. A brief description and results of this work are available for download from the links below.
Song, X.P., Potapov, P.V., Krylov, A., King, L., Di Bella, C.M., Hudson, A., Khan, A., Adusei, B., Stehman, S.V., Hansen, M.C. (2017) National-scale soybean mapping and area estimation in the United States using medium resolution satellite imagery and field survey. Remote Sensing of Environment, 190, 383-395.