Type Article
Date 2022-04
Language English
Author(s) Frouin Robert1, Tan Jing1, Compiègne Mathieu2, Ramon Didier2, Sutton Marshall3, Murakami Hiroshi4, Antoine David5, 6, Send Uwe1, Sevadjian Jeff1, Vellucci Vincenzo6
Affiliation(s) 1 : Scripps Institution of Oceanography, University of California San Diego, La Jolla, CA, United States
2 : HYGEOS, Euratechnologies, Lille, France
3 : NASA Goddard Space Flight Center, Greenbelt, MD, United States
4 : Earth Observation Research Center, Japan Aerospace Exploration Agency, Ibaraki, Japan
5 : Remote Sensing and Satellite Research Group, School of Earth and Planetary Sciences, Curtin University, Perth, WA, Australia
6 : Laboratoire d’Océanographie de Villefranche, CNRS, Sorbonne Université, Institut de la Mer de Villefranche, Villefranche-sur-Mer, France
Source Frontiers In Remote Sensing (2673-6187) (Frontiers Media SA), 2022-04 , Vol. 3 , P. 833340 (17p.)
DOI 10.3389/frsen.2022.833340
WOS© Times Cited 2
Keyword(s) photosynthetically available radiation, satellite remote sensing, Lagrange L1 orbit, EPIC sensor, DSCOVR mission, light absorption and scattering, ocean biogeochemistry

The EPIC/DSCOVR observations of the Earth’s surface lit by the Sun made from the first Lagrange point several times during the day in spectral bands centered on 443, 551, and 680 nm are used to estimate daily mean photosynthetically available radiation (PAR) at the ice-free ocean surface. The PAR algorithm uses a budget approach, in which the solar irradiance reaching the surface is obtained by subtracting from the irradiance arriving at the top of the atmosphere (known), the irradiance reflected to space (estimated from the EPIC Level 1b radiance data), taking account of atmospheric transmission and surface albedo (modeled). Clear and cloudy regions within a pixel do not need to be distinguished, which dismisses the need for often-arbitrary assumptions about cloudiness distribution within a pixel and is therefore adapted to the relatively large EPIC pixels. A daily mean PAR is estimated on the source grid for each EPIC instantaneous daytime observation, assuming no cloudiness changes during the day, and the individual estimates are remapped and weight-averaged using the cosine of the Sun zenith angle. In the computations, wind speed, surface pressure, and water vapor amount are extracted from NCEP Reanalysis 2 data, aerosol optical thickness and Angström coefficient from MERRA-2 data, and ozone amount from EPIC Level 2 data. Areas contaminated by Sun glint are excluded using a threshold on Sun glint reflectance calculated using wind data. Ice masking is based on NSIDC near-real-time ice fraction data. The product is evaluated against in situ measurements at various locations and compared with estimates from sensors in polar and geostationary orbits (MODIS, AHI). Unlike with MODIS, the EPIC PAR product does not exhibit gaps at low and middle latitudes. Accuracy is satisfactory for long-term studies of aquatic photosynthesis, especially given the much larger uncertainties on the fraction of PAR absorbed by live algae and the quantum yield of carbon fixation. The EPIC daily mean PAR product is generated operationally on a Plate Carrée (equal-angle) grid with 18.4 km resolution at the equator and on an 18.4 km equal-area grid, i.e., it is fully compatible with the NASA Greenbelt OBPG ocean-color products. Data are available since the beginning of the DSCOVR mission (i.e., June 2015) from the NASA Langley ASDC website.

Full Text
File Pages Size Access
Publisher's official version 17 4 MB Open access
Top of the page