Storm damage assessment support service in the U.S. corn belt using RapidEye satellite imagery

摘要

The systematic use of satellite images to support crop damage assessment has been hampered by two reasons: lack of satellite systems with high revisit frequency have made it difficult to image affected areas within the necessary time windows, and second, the high cost of satellite images with appropriate resolution have made their integration into existing processes too costly to be valuable. RapidEye has developed an operational service to produce damage assessment maps on insured fields in the U.S. corn belt that overcomes these problems. The service will be capable of delivering the maps in a suitable time frame by taking advantage of the RapidEye constellation. The production of these maps is based on analysis of after-storm images only, thus reducing the cost of the service. The processing is based on a six step process that includes: pre-processing of the images; estimation of agricultural land vegetation parameters from the imagery; separation of corn from soybean fields; estimation of the statistical descriptors for corn and soybean fields separately; inner-field classification into damage classes; and integration of the results in maps with road and land cover data. Results from the 2008 and 2009 crop seasons are presented.