Franz, Bryan: NASA Goddard Space Flight Center (Project Lead)
Project Funding:
2014 - 2017
NRA: 2013 NASA: Terra and Aqua: Algorithms--Existing Data Products
Funded by NASA
Abstract:
This proposal seeks to continue the efforts of the NASA Ocean Biology Processing Group (OBPG) to maintain the quality of several standard ocean color products currently produced from MODIS, including spectral water-leaving "remote sensing" reflectance (Rrs), chlorophyll concentration (Chl), and marine diffuse attenuation at 490nm (Kd490).
Rrs is the fundamental product derived from ocean color sensors. It is the product from which most bio-optical or bio-geochemical products are subsequently derived. Thus, maintenance of the Rrs product at the highest possible quality, and quantification of the uncertainties in that Rrs product, is paramount to continuation of meaningful ocean color science from MODIS. The retrieval of Rrs from spaceborne sensors, however, is highly sensitive to the instrument calibration and the retrieval algorithms. Maintenance of the instrument calibration for MODIS has historically required significant efforts to quantify and correct for sensor degradation issues that change radiometric performance over time, across scan, and between detectors and mirror sides. The OBPG has maintained the calibration of MODIS on Terra and Aqua, using vicarious and cross-sensor techniques, to achieve a level of stability that far exceeds what can be achieved through the on-board (solar, lunar) calibration alone. We propose to continue this effort, which grows more challenging as the sensors age and radiometric degradation becomes less predictable. We also seek to maintain and continue minor refinements to the Rrs algorithm. This algorithm is actually a complex series of algorithms, collectively referred to as atmospheric correction, that account for the radiative transfer of light from the sun through the atmosphere, into the euphotic zone of the upper ocean and back to the MODIS sensor. The algorithms account for absorbing atmospheric gases, scattering by air molecules, scattering and absorption by aerosols, effects of surface whitecaps and sun glint, transmittance through the air-sea interface, and the bidirectional structure of the sub-surface light field. While the algorithms currently employed for MODIS atmospheric correction are well established, this is an active area of international research and continuous refinements are anticipated, especially in the areas of aerosol modeling and atmospheric correction over turbid coastal waters. We propose to maintain the existing atmospheric correction algorithm and evaluate and incorporate minor refinements consistent with the state of the art.
The standard MODIS Chl and Kd490 algorithms are empirical band ratio algorithms derived by relating field measurements of spectral Rrs to field measurements of Chl and Kd490 using the NASA bio-Optical Marine Algorithm Dataset version 2 (NOMADv2). We propose to update NOMAD with a host of new data sources and data products to increase robustness of the empirical algorithm regressions, and provide updated coefficients for the standard Chl algorithm (OC3) and the standard Kd490 algorithm (KD2) of MODIS. In addition, we propose to incorporate and evaluate minor refinements to the form of these empirical algorithms based on methods already published.
The standard mechanism for assessing quality of ocean color products is through match- ups with field measurements. The OBPG currently maintains the SeaWiFS Bio-optical Archive and Storage System (SeaBASS). We will utilize SeaBASS to provide validation statistics for Rrs, Chl, Kd490, and all other MODIS standard ocean color products for which quality field data exists. In addition, we will perform global and regional time- series analyses and comparative analyses with other ocean color mission such as SeaWiFS and MERIS. These studies will provide a more complete, global and temporal picture of the uncertainties in the satellite record, far beyond what can be assessed through comparison with geographically and temporally sparse field measurements.
More details may be found in the following project profile(s):
