Evaluation of the NASA OBPG MERIS ocean surface PAR product in clear sky conditions.

Jing Tan,Dominique Jolivet,Didier Ramon,Mathieu Compiègne,Robert Frouin

doi:10.1364/oe.396066

Abstract

The operational MEdium Resolution Imaging Spectrometer (MERIS) daily mean photosynthetically available radiation (PAR) product generated by the NASA Ocean Biology Processing Group (OBPG) was evaluated in clear sky conditions against in-situ measurements at various sites in the northwestern Mediterranean Sea (BOUSSOLE buoy), the northwestern Pacific (CCE-1 and -2 moorings), and the northeastern Atlantic (COVE platform). The measurements were first checked and corrected for calibration errors and uncertainties in data processing by comparing daily means for clear days (i.e., no clouds from sunrise to sunset and low aerosol abundance) with theoretical values from an accurate Monte Carlo radiative transfer code. The OBPG algorithm performed well when sky was completely cloudless during daytime, with a bias of 0.26 E/m2/d (0.6%) and a RMS difference of 1.7 E/m2/d (4.0%). Using satellite-derived aerosol optical thickness (AOT) and Angström coefficient instead of climatology slightly degraded the results, which was likely due to uncertainties in the aerosol retrievals. A sensitivity study to aerosol properties indicated that climatology may not work in some situations (e.g., episodic dust, pollution, or biomass burning events), suggesting that it is best to use actual aerosol estimates in clear sky conditions. The analysis also revealed that specifying aerosol properties, therefore atmospheric transmittance, from AOT and Angström coefficient, even retrieved from the satellite imagery, may not be sufficient in the presence of absorbing aerosols, especially when loadings are important. Performance was degraded when including situations of clear sky at the time of the MERIS observation but cloudy sky before and/or after overpass, resulting in a bias (overestimation) of 2.8 E/m2/d (7.3%) and a RMS difference of 6.0 E/m2/d (15.8%). The relatively large overestimation was due to the inability of the OBPG PAR algorithm to detect cloudiness at times other than the time of satellite overpass. The key to improving the daily mean PAR estimates in such situations does not reside so much in improving the radiative transfer treatment or specifying more accurately aerosol properties, but rather in accounting properly for the diurnal variability of cloudiness. To this end, a methodology that utilized Modern Era Retrospective Reanalysis for Research and Applications, Version 2 (MERRA-2) hourly cloud data (fractional coverage, optical thickness) was proposed and tested, reducing the bias to 1.6 E/m2/d (4.2%). Improvement was not sufficient in some situations, due to the coarse resolution and uncertainties of the MERRA-2 products, which could not describe properly the cloud properties at the local scale (MERIS pixel). The treatment is applicable to any cloud situation and should be considered in a future version of the of OBPG PAR algorithm. This would require, however, refreshing the standard OBPG PAR products generated as part of the ocean-color processing line according to MERRA-2 data availability.

Highlights

Available radiation (PAR) at the ocean surface, i.e., the solar energy flux reaching the surface in the spectral range 400-700 nm, controls the growth of phytoplankton and, the composition and evolution of marine ecosystems (e.g., [1])
While in-situ measurements of Photosynthetically available radiation (PAR) are quite scarce over the world oceans, especially dailyaveraged values at a given location, satellite remote sensing observations provide the opportunity to obtain those values at synoptic spatial scales
The Frouin et al [10] approach has been adopted by the NASA Biology Processing Group (OBPG) and Japan Aerospace Exploration Agency (JAXA) Earth Observation Research Center (EORC) to generate routinely Level 2 and Level 3 daily mean PAR products from the Sea-viewing Wide Field-of-view Sensor (SeaWiFS), the MODerate resolution Imaging Spectroradiometer (MODIS), the Visible Infrared Imaging Radiometer Suite (VIIRS), Medium Resolution Imaging Spectrometer (MERIS), Global Imager (GLI), Second-Generation Global Imager (SGLI), Ocean and Land Colour Instrument (OLCI), and Advanced Hiwamari Imager (AHI) data

Summary

Introduction

Available radiation (PAR) at the ocean surface, i.e., the solar energy flux reaching the surface in the spectral range 400-700 nm, controls the growth of phytoplankton and, the composition and evolution of marine ecosystems (e.g., [1]). Sunlight in the PAR spectral range warms the upper ocean layers and distributes heat in the vertical. This heat distribution is modulated by water composition, creating density differences that generate currents, with implications on atmospheric circulation and climate, local and remote [3,4,5,6,7]. Studies using the OBPG PAR products have addressed a variety of topics, including biosphere productivity during an El Niño transition [18], carbon-based ocean productivity modeling [19], climate-driven trends in ocean productivity [20,21], inter-comparison of ocean productivity algorithms [22], and relation between primary productivity, vertical mixing, and atmospheric inputs in the Yellow Sea [23]

Results

Discussion

Conclusion