Temperature-Dependent Infrared Sea Surface Effective-Emissivity (IRSSE) Model: Theoretical Development and Validation

For satellite IR remote sensing applications, the surface emissivity/reflectance spectrum must be specified with a high degree of absolute accuracy; a 0.5% uncertainty can result in ≈0.3–0.4 K error in LWIR window channels. Over ocean surfaces, the IR (and solar) reflectance is characterized as quasi-specular, meaning that the sky reflectance originates from a range of downwelling angles surrounding the specular angle, leading to an observed systematic underestimation in surface-leaving radiance from conventional emissivity models. To account for this problem in a practical manner, an IR effective-emissivity (IRSSE) model (Nalli et al., 2008a,b) was developed for the Community Radiative Transfer Model (CRTM) in an effort to obtain improved agreement with surface based radiance observations (viz., MAERI spectra) over the usual range of satellite zenith angles, IR wavelengths, and surface wind speeds. However, although there was a known dependence on surface temperature, it was not until recent findings of Liu et al. (2017) that a significant systematic bias (as much as 1 K) was revealed to occur on a global scale in cold waters (i.e., the North Atlantic and Southern Oceans), which has brought attention back to this issue. The Joint Center for Satellite Data Assimilation (JCSDA) and the Joint Polar Satellite System (JPSS) are thus currently supporting work toward model upgrades to address this problem for both the CRTM and the SARTA models. This presentation will provide an overview the IRSSE model along with the upgrade status.