Location: Hydrology and Remote Sensing Laboratory
Title: Advanced multispectral sensor requirements for remote sensing of agriculture and land cover Authors
Submitted to: Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: December 18, 2010
Publication Date: March 29, 2012
Citation: Serbin, G., Hunt Jr, E.R., Daughtry, C.S., Anderson, M.C. 2012. Advanced multispectral sensor requirements for remote sensing of agriculture and land cover [abstract]. Joint Agency Commercial Imagery Evaluation Workshop. http://calval.cr.usgs.gov/JACIE_files/JACIE11/Presentations/ThurPM/130_Serbin_JACIE_11.120.html Technical Abstract: Modern agricultural and land cover monitoring programs require frequent data acquisitions and increased spectral resolution to acquire a greater number of parameters in a more accurate manner. Whereas hyperspectral sensors could provide the required information, agriculture's biggest need is for frequent revisit intervals over large areas, which cannot be met by pointable sensors. Future meteorological satellites will provide frequent coverage over large areas; however, these sensors will not have the spectral bands required, and the pixel sizes are too large for agricultural monitoring. Vegetation generally has the same spectral features, so an advanced multispectral system containing 13 spectral bands would be ideal for such purposes. Important biophysical parameters that would be monitored include live vegetation cover, chlorophyll content, vegetation water content, non-photosynthetic vegetation cover (NPV, e.g., crop residues and dry grasses), evapotranspiration, and vegetation stress. Remote sensing of NPV requires three shortwave infrared bands between 2.025 – 2.230 'm that allow for calculation of the Cellulose Absorption Index. While current and planned multispectral sensors can acquire most of these parameters, none are suited for remote sensing of NPV, which provides unique and important information about tillage practices, rangeland health, brush fire hazards, soil organic carbon dynamics, soil erosion, and water quality. We also include bands for atmospheric correction. Other requirements for an agricultural monitoring system are 12-bit or better quantization, signal-to-noise ratio > 250, temporal resolution < 7 days to ensure cloud-free scenes and to capture critical crop growth stages, and pixel size < 60 m. This concept is based on discussions about satellite data requirements for agricultural monitoring and does not represent USDA policy.