Applying Lidar Observations for Decoupling Aerosol Contamination of Hyperspectral Radiance Retrievals during Data Assimilation

Hyperspectral radiance assimilation is performed at all major forecasting centers to improve the analysis temperature and moisture fields in operational weather models. These radiance retrievals and, by default, the assimilation of these radiances, are vulnerable to atmospheric dust aerosols, which absorb, scatter, and emit in the infrared (IR) spectrum. Previous studies have shown IR-absorbing dust induced biases on Atmospheric Infrared Sounder (AIRS) shortwave and thermal radiance retrievals—the magnitude of which was dependent on both altitude and aerosol optical depth (AOD). Furthermore, dust is often present in relatively high concentrations along the Eastern Atlantic tropical cyclone (TC) formation zone during hurricane season. Assimilation of these dust-biased radiances plausibly impact the retrieved temperature and moisture profiles, and thus, the forecasts of TC growth, intensity, and track based on these profiles. To mitigate these issues, dust contaminated observations must either not be assimilated, or the dust contribution must be decoupled from the observed radiances.

In this study, we conceptualize how lidar observations can be used to help decouple the dust contribution from observed radiance spectra during data assimilation. We show how the inclusion of lidar retrieved aerosol profiles during assimilation improves temperature and moisture retrievals using a modified one-dimensional variational assimilation system coupled with a radiative transfer model adequately equipped with aerosol optical properties. We then quantitatively describe the sensitivity of temperature and moisture retrievals to aerosol present by depicting bias in model increments with respect to plume altitude and optical depth. A method of correcting for aerosol bias on radiance assimilation operationally using CALIOP is also presented. The result is an introduction to the community of the effect that IR-absorbing aerosols can have on standard passive hyperspectral infrared retrievals and the role that lidars can play in providing atmospheric characterization and potential correction.