Improving Near-Surface Retrievals of Surface Humidity Over the Global Open Oceans From Passive Microwave Observations.

Ocean evaporative fluxes are a critical component of the Earths energy and water cycle, but their estimation remains uncertain. Near-surface humidity is a required input to bulk flux algorithms that relate mean surface values to the turbulent fluxes. Several satellite-derived turbulent flux products have been developed over the last decade that utilize passive microwave imager observations to estimate the surface humidity. It is known, however, that these estimates tend to diverge from one another and from in situ observations. Analysis of current state-of-the-art satellite estimates provided herein reveals that regional-scale biases in these products remain significant. Investigations reveal a link between the spatial coherency of the observed biases to atmospheric dynamical controls of water vapor vertical stratification, cloud liquid water, and sea surface temperature. This information is used to develop a simple state-dependent bias correction that results in more consistent ocean surface humidity estimates. A principal conclusion is that further improvements to ocean near-surface humidity estimation using microwave radiometers requires incorporation of prior information on water vapor stratification and sea surface temperature. Plain Language Summary To reduce the uncertainties in ocean evaporation-which is currently of leading order in water and energy budget studies-progress is necessary to understand the sources of error in the surface measurements needed to model the ocean-atmosphere exchange of water. Recent studies have independently demonstrated systematic uncertainties in estimates of ocean evaporation and surface humidity related to large-scale dynamics and vertical water vapor stratification. The results presented herein directly establish the link between the large-scale patterns of uncertainties in ocean evaporation estimates and the sources of errors in estimation of surface humidity from remote sensing observations. These results are used to characterize the relationship of surface humidity errors to physical quantities-sea surface temperature, water vapor stratification, and cloudiness-that are used to develop simple corrections for these errors. While effective, these corrections indicate the need for future algorithm development that incorporates information on these conditions directly into the remote sensing inverse-modeling process.