Jumbo Content

ROSES Proposals

Improving IOP Measurement Uncertainties for PACE Ocean Color Remote Sensing Applications

PI: James Sullivan - Harbor Branch Oceanographic Institute of Florida Atlantic University
A goal of the PACE Science Team is to achieve consensus and develop community-endorsed paths for measurement suites required for the PACE mission. This proposed work will specifically address the inherent optical properties of absorption and backscattering, better quantifying uncertainties using current and emerging methods, and improving uncertainties in both reprocessed historical data and future data collections for the PACE mission. Accurate values of absorption and backscattering, and estimates of their uncertainties, are critical for remote sensing validation and development/refinement of retrieval algorithms. However, one aspect of each property stands out as an enduring source of uncertainty. For absorption measurements in particle fields, this aspect is the scattering error associated with reflective tube absorption technologies (as in the widely used WET Labs ac devices). Recent testing in our labs has shown that the different schemes used to correct this error in virtually all of the ac device data submitted to SeaBASS over the last 20 years has significant errors (10% or >). At this time, these errors are acknowledged (e.g. Leymarie et al. 2010; McKee et al. 2013), but there is no consensus on a recommended protocol for correcting scattering errors with the best accuracy possible. For backscattering, the largest area of uncertainty in clear ocean waters is the backscattering contribution from the pure seawater itself, which can comprise 80-90% of total backscattering for great swaths of the ocean. Recently, Zhang et al. (2009) revised the theory to describe pure seawater scattering as a function of the physical properties of water (i.e., temperature, salinity, pressure), but a single critical physical constant of pure water used in these calculations remains poorly known: the depolarization ratio, which is the ratio of horizontally polarized light to vertically polarized light in the scattered beam at 90°. The value the in situ optics community is currently using comes from a single study conducted nearly 40 years ago (Farinato & Rowell 1976). In that work, 3 experimental values were actually derived: 0.051, 0.045, and 0.039, each with different viewing optics. Currently, the lowest one measured, 0.039 is usually recommended because of the difficulties with stray light contamination possibly elevating their other experimental values. For decades, the remote sensing community has typically used the pure seawater scattering values of Morel (1974), which followed a different theoretical approach, with a recommended depolarization value of 0.09 (more than 100% higher than our current best guess), based on the state of knowledge at the time. Thus, one is left with virtually no confidence in the accuracy of this parameter, only a gut feeling that we are in the ballpark. We should note here that even relatively modest uncertainties in the depolarization ratio (e.g. 10%) could translate to large uncertainties in open ocean particulate backscattering retrievals (e.g. 40%), due to the fact that the subtracted pure seawater component can be 80-90% of the entire water-leaving signal. We propose to conduct comprehensive historical data analyses to fully characterize uncertainties in the scattering error for reflective tube absorption meters, and to determine and validate optimal scattering correction methods through minor lab experimentation for different deployment configurations and suites of ancillary measurements. The impacts of using different depolarization ratio values for the determination of pure water backscattering on uncertainties in ocean color retrieval algorithms will also be investigated. We further propose to use modern, specialized, bench top volume scattering function equipment and water purification methods in our lab to determine and validate with rigorous uncertainty the precise value of the depolarization ratio to improve the accuracy of retrievals for historical data and the future PACE mission.