Development of a surface wind retrieval by analysing sunglint geometry from specMACS radiance measurements

Near-surface winds are a key component for the coupling of the atmosphere- ocean system. Convergence and divergence patterns can be inferred from measurements of the surface wind vector, enabling the characterization of meso- and synoptic-scale atmospheric dynamics. Observations over remote areas, such as the Atlantic ocean, are mostly limited to satellites and buoys. Geostationary satellites derive wind data primarily from cloud tracking and thus do not measure surface winds. In contrast, polar-orbiting satellites can provide surface wind data predominantly using active remote sensing instruments, such as scatterometers, wind-lidars or synthetic aperture radars, providing better spatial but certainly lower temporal resolution. Finally, buoy measurements are point-observations and cannot be employed for large-scale wind field analyses. This work aims to explore an alternative approach for quantifying surface wind fields over the ocean by analysing high-spatial resolution imagery from airborne observations.Measurements of specularly reflected solar radiation (sunglint) by the hyperspectral and polarized imager specMACS aboard the German research aircraft HALO are employed for the development of a surface wind retrieval. Size and shape of the sunglint are predetermined by wind speed and direction: With ocean surface roughness directly corresponding to near-surface wind speed, the specMACS retrieval makes use of the relationship between ocean wave slope distribution and angular variation of sunglint radiance. SpecMACS measurements of spectral radiances are evaluated against simulated spectral radiances for different solar zenith and azimuth angles, as well as surface wind speeds and directions. The radiative transfer simulations are done with the Monte Carlo (MYSTIC) solver of the libRadtran radiative transfer package.The overarching goal of this work is the development of an operational surface wind retrieval after analysing selected cases as an initial step. The retrieval requires a view of the ocean surface from the aircraft. We aim to explore to what extent the wind retrieval can be employed for (a) partially cloud-covered scenes or (b) scenes with an optically thin cirrus layer above or below the aircraft. A first application will be the analysis of data acquired during the recent ’Persistent EarthCare underflight studies of the ITCZ and organized convection’ (PERCUSION) sub-campaign in the tropical Atlantic. Research concerned with the horizontal wind field structure of atmospheric phenomena, e.g. the doldrums in the deep tropics, will benefit from along flight-track surface wind observations. Continuous surface wind data also further supplement dropsonde point-measurements.