Home >> Methodology
CropWatch analyses of global production and environmental and agricultural trends are based on remote sensing and ground data and a combination of well-established and innovative methodologies. Over the years, the CropWatch system has been updated regularly as new data and methodologies became available or matured. The latest bulletins have used a new hierarchical approach of indicators and spatial scales and a new operational methodology, while also taking advantage of several new remote sensing data sources.

Hierarchical approach

The CropWatch hierarchical approach for crop production monitoring involves the use of specific environmental and agricultural indicators on different scales, using the combined information to assess global, regional, and national (as well as sub-national) crop condition, production, and agricultural trends. The analysis covers the following four levels:
  • •Global—Inputs include rainfall, temperature, photosynthetically active radiation, and potential biomass. Outputs on this level are assessments of environmental indices for sixty crop production system zones.
  • •Regional—Inputs include the vegetation health index, uncropped arable land, cropping intensity, and maximum values for the vegetation condition index during the growing season. Outputs are information on agricultural patterns, farming intensity, and biomass trends in six major production zones in the world.
  • •National—Inputs are NDVI, crop cultivation area, and time profile clustering. Outputs are information on cropland use intensity, crop condition, yield, and production, for each of thirty key countries and China, together representing at least 80 percent of global production of the major cereals and soybean.
  • •Sub-national—For large countries, analysis is performed on a sub-national level. In the case of China, additional information on crop planting structure is analyzed. Outputs on the sub-national level are similar to those on the national level, but with more detailed information
For the analysis, each level is covered with specific indicators adapted to its scale, while the next broader level provides the overall background. Using the outputs on the four levels, a synthesis of crop production and crop condition estimates is published quarterly in the CropWatch bulletin.

Data sources and additional information

CropWatch currently mostly relies on publicly available satellite remote sensing and in situ data. Detailed descriptions of CropWatch methodology and definition of spatial units are posted with each bulletin and on the bulletin page, with additional detail provided in the publications. Because of the relative newness of some of the methods and data, further adjustments of the methodologies are expected for future bulletins, to incorporate ongoing research on the data and experimental methodologies, as well as feedback on these initial approaches.