Vegetation Index

The PRF VI Program is based on the use of (NDVI) data that has been derived from scanners on satellites observing the changes in greenness of vegetation of the earth since 1989. The vegetation index does not explicitly predict individual forage production on a given producer's operation. Instead, the index is related to the amount of vegetation on earth and the changes in greenness over time. The index is correlated with forage production but does not directly predict actual forage production. Key components of the PRF-VI Pilot Program include:
  • The temperature constrained vegetation index is derived from two data sources. One is the NDVI data from NASA that is processed by EROS, and the second is gridded average daily temperature data produced by the NOAA Climate Prediction Center.
  • Losses and payments are based on a sub-county level. The data is collected for 1-km grids (0.6 miles) and then averaged across 8 km or 4.8 mile grids to produce the base value used in the vegetation index. The bi-weekly values are stored on computers at EROS from the maximum observed daily value observed during each 2-week period.
    • Using a ratio of infrared and visible light reflectance from the vegetation, an index of 0.0 to 1.0 is derived which essentially reflects the abundance of vegetation on the ground. The visible light values are sensitive to the amount of green chlorophyll in the plant which can absorb it, while the infrared light is sensitive to the amount of plant cells inside the leaf, which reflect it back to the satellite. Therefore, the more chlorophyll and the more plant cell bulk in the plant the higher the NDVI index recorded by the satellite's sensor.
    • Vegetation can remain green in very cold or very hot weather and yet not grow. There are also examples of pasture species with thick canopies that prevent light from fully penetrating and reflecting back from all leaf surfaces. For these reasons, a temperature correction has been added to correct the final index value. A mathematical function has been added that gives the NDVI more weight when temperatures are close to optimum for plant growth in a region and discounts the index when temperature deviates from this optimum value. Optimum plant growth temperatures are also adjusted for elevation differences between the NDVI grids in a given region.
    • Four intervals were selected for indexing in an annual cycle with each interval having 3-month duration. Interval I begins on April 1 and ends on June 30. After extensive discussions with the subject matter experts it was recognized that there were significant variations between regions of the United States. It was determined that the best way to capture these variations in conditions and allow producers the most flexibility to insure periods in the year that best reflect their local conditions was to divide the year into 4, 3-month intervals. Increasing intervals beyond 3-months reduced signal variability and responsiveness of the index in the higher producing, mixed forest-pastureland conditions of the United States. Choosing one or more 3-month interval allows producers to manage their risk based on their diversity of pasture and rangeland forage.
All of EROS's NDVI data and NOAA's temperature data are subjected to rigorous quality control processes to insure that the data is stable and as accurate as possible. Both datasets are widely used as part of the global monitoring system used by various government agencies, including national security and defense organizations. There are multiple satellite systems that have been placed in orbit that are calibrated to each other so that if one fails they can use the other system to cover the same area. In addition, NASA plans to launch a series of satellites using the same sensors and mathematics over the next decade to insure continuity of the program.

When the data is reported by EROS and NOAA, computers automatically log onto these sites, download the data and update the index value. The index values are maintained on two separate computer systems and are reported at the end of each 3-month interval. This value is then made available to insurance companies for computation of indemnity payments via RMA's web site.