Spaceborne Hyperspectral EO-1 Hyperion Data Pre-Processing

Abstract

Increases in global populations and acceleration in land cover and land use change necessitate the rapid monitoring of such changes to address issues like global food security and global water security. Remote sensing has been instrumental in doing this over the past 50 years. There have also been many advances in remote sensing technology; although many of these advances have been limited to multispectral sensors, we have hyperspectral data from the recently decommissioned Hyperion sensor, and upcoming hyperspectral sensors such as Germany's EnMAP and the National Aeronautics Space Administration's (NASA's) HyspIRI. However, there is no standardized protocol for pre-processing hyperspectral data, such as the Hyperion imagery. We need to establish such protocols for hyperspectral data to facilitate the use of these large datasets efficiently to address ecological questions at global extents. In this chapter, we review methods available for pre-processing Hyperion data and suggest a workflow for Hyperion image pre-processing. Examples of these pre-processing steps are also provided for the Google Earth Engine (GEE) cloud-computing platform, which facilitates studies at a global extent by eliminating the need to store and process imagery on a personal computer. These hyperspectral datasets, once pre-processed, are useful for many applications including vegetation classification, biomass estimation, and crop water productivity estimation.