Unified estimation of spectral coastal land and water reflectances with uncertainties
Abstract
Coastal ecosystem studies are well-suited for remote visible/infrared spectroscopy. These investigations typically invert an atmospheric model to estimate water-leaving reflectance or water-leaving radiance spectra. This inversion is challenging because the combined effects of turbid backscatter, atmospheric aerosols, and sun glint can make it difficult to disambiguate surface and atmosphere effects. Simultaneous estimation of the surface and atmosphere can resolve the ambiguity while providing rigorous uncertainty quantification. We demonstrate simultaneous retrievals adapting the Optimal Estimation (OE) model inversion formalism of Rodgers (2000) to coastal environments including both land and water. We compare two surface parameterizations: (i) a parametric bio-optical model and (ii) a vector representation of surface reflectance in each instrument channel. The latter is suited to both land and water reflectance, unifying atmospheric correction across terrestrial and aquatic domains. We test these models with both vector and scalar Radiative Transfer Models (RTMs). We report field experiments by NASA's Portable Remote Imaging SpectroMeter (PRISM) over Santa Monica, California, and NASA's Next Generation Airborne Visible Infrared Imaging Spectrometer (AVIRIS-NG) over the Wax Lake Delta and lower Atchafalaya River, Louisiana. In situ validation measurements match remote water-leaving reflectance estimates to high accuracy in both locations. In addition to reducing error, posterior error predictions provide a closed account of uncertainties consistent with measured residuals.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2019
- Bibcode:
- 2019AGUFMGC51E1125T
- Keywords:
-
- 1632 Land cover change;
- GLOBAL CHANGE;
- 1640 Remote sensing;
- GLOBAL CHANGE;
- 4333 Disaster risk analysis and assessment;
- NATURAL HAZARDS;
- 4217 Coastal processes;
- OCEANOGRAPHY: GENERAL