Extended Abstract (944K)8.1
Validation of satellite-based estimates of whitecap coverage: Approaches and initial results
Magdalena D. Anguelova, NRL, Washington, DC ; and J. P. Bobak, W. E. Asher, D. J. Dowgiallo, B. I. Moat, R. W. Pascal, and M. J. Yelland
Within the framework of the WindSat mission at Naval Research Laboratory (NRL), we have a dedicated project to address the high variability of whitecap coverage (W) and the sea salt source function. A major task of this project is retrieving W on a global scale from satellite-based measurements. This will allow us to build a database representing W under a wide range of meteorological and environmental conditions. The algorithm for estimating W from satellite measurements relies on changes of ocean surface emissivity at microwave frequencies (6 to 37 GHz) due to presence of sea foam on a rough sea. After proving the feasibility (Anguelova and Webster, 2006), NRL now has the capability to obtain W globally using WindSat data and an improved version of the algorithm.
Direct validation and assessment of the quality of satellite-based whitecap coverage on a global scale is not straightforward due to a lack of ground truth values under sufficiently varying conditions. Plans for new whitecap measurements are sporadic and non-systematic, making the creation of a database of in situ W values co-located in time and space with satellite-based estimates difficult. To compensate for insufficient ground truth data, we use different approaches for validation.
1) Validation with the historical in situ database: Used to demonstrate consistency of the satellite-based W values with previously measured whitecap coverage.
2) Validation with existing wind-speed-dependent parameterizations of whitecap coverage: Used to illustrate similarities due to wind-speed dependence and differences due presumably to other meteorological and oceanographic factors.
3) Validation with air borne data: Conducted in August 2007, field campaign Radiometry and Sea Surface Imagery (RASSI) involves measurements of breaking waves with radiometric system (from NRL) and high-resolution video camera (from University of Washington). Data were collected over the Atlantic coast during low wind speed conditions on several days and over the Gulf of Mexico while heading towards Hurricane Dean on 21 August. Initial comparisons between satellite and aerial observations are good (Bobak et al. 2008).
4) Validation with ship borne data: Two recent field campaigns provide in situ data for W matched in time and space with WindSat observations, namely High Wind Air-Sea Exchanges (HiWASE) experiment (Yelland et al., 2007) and Sea spray, gas flux, and whitecap (SEASAW) study (Brooks et al., 2007). The HiWASE instrumentation, mounted on the Ocean Weather Ship Polarfront occupying Station Mike (at 66N 2E), will be in continuous operation until September 2009. Initial results, using data collected at station Mike on the Polarfront ship from September 2006 to September 2007, are quite encouraging.
While each approach adds more data at more validation points, the diversity of measuring techniques used on different scales (in situ, aerial, and satellite) poses different challenges.
Validation with in situ data consists of comparison of spatially averaged satellite values and temporally averaged in situ values. For meaningful inter-comparison between satellite and in situ data, finding the optimal parameters for the spatial and temporal averaging is a major task.
For inter-comparison between satellite and airborne data it is necessary to understand discrepancies due to the different resolutions of the satellite radiometer, airborne radiometer, and airborne video camera.
Finally, differences between radiometric and photographic measurements of W are expected because sea foam is seen differently in the optical and microwave regions of the electromagnetic spectrum. These need to be understood so photographic and radiometric observations are properly compared.
Results from different validation approaches will be shown and the issues raised above will be discussed.
Session 8, Sea-Surface Physics, Including Waves, Whitecaps, and Aerosol Generation: Part 1
Wednesday, 14 January 2009, 8:30 AM-9:45 AM, Room 128A
Next paper