Estimating Achievable Accuracy for Global Imaging Spectroscopy Measurement of Non-Photosynthetic Vegetation Cover
Abstract
Terrestrial vegetation is dynamic, expressing seasonal, annual, and long-term changes in response to climate and disturbance. Phenology and disturbance (e.g. drought, insect attack, and wildfire) can result in a transition from photosynthesizing "green" vegetation to non-photosynthetic vegetation (NPV). NPV cover can include dead and senescent vegetation, plant litter, agricultural residues, and non-photosynthesizing stem tissue. NPV cover is poorly captured by conventional remote sensing vegetation indices, but it is readily separable from substrate cover based on spectral absorption features in the shortwave infrared. We will present past research motivating the need for global NPV measurements, establishing that mapping seasonal NPV cover is critical for improving our understanding of ecosystem function and carbon dynamics. We will also present new research that helps determine a best achievable accuracy for NPV cover estimation. To test the sensitivity of different NPV cover estimation methods, we simulated satellite imaging spectrometer data using field spectra collected over mixtures of NPV, green vegetation, and soil substrate. We incorporated atmospheric transmittance and modeled sensor noise to create simulated spectra with spectral resolutions ranging from 10 to 30 nm. We applied multiple methods of NPV estimation to the simulated spectra, including spectral indices, spectral feature analysis, multiple endmember spectral mixture analysis, and partial least squares regression, and compared the accuracy and bias of each method. These results prescribe sensor characteristics for an imaging spectrometer mission with NPV measurement capabilities, as well as a "Quantified Earth Science Objective" for global measurement of NPV cover. Copyright 2016, all rights reserved.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2016
- Bibcode:
- 2016AGUFMGC52A..05D
- Keywords:
-
- 0410 Biodiversity;
- BIOGEOSCIENCESDE: 1615 Biogeochemical cycles;
- processes;
- and modeling;
- GLOBAL CHANGEDE: 1632 Land cover change;
- GLOBAL CHANGEDE: 1635 Oceans;
- GLOBAL CHANGE