Investigating the impacts of assimilating All-Sky Window-Channel Infrared Radiances from GOES-ABI using MPAS-JEDI

The assimilation of cloud and precipitation affected microwave radiances has allowed significant improvements in analyses and forecasts in several operational numerical weather prediction (NWP) centers. Currently, efforts are being made to assimilate water vapor bands from infrared sensors, but this is still under research in various centers. Infrared window channels have been less explored due to the many challenges associated with it, such as observations error model, cloudy radiance observations operator, variational bias correction, and cloud representation in the models. In this study, we investigate the assimilation of infrared window channel 13 from GOES-ABI over the ocean, using the Model for Prediction Across Scales – Atmosphere (MPAS-A) coupled with the Joint Effort for Data assimilation Integration (JEDI). The Community Radiative Transfer Model (CRTM) interfaced in JEDI-UFO is used as a cloudy radiance observations operator for the channel 13's all-sky assimilation and forecast evaluation. We analyze the sensitivity to GOES-ABI channel 13 by using various observation error models and variational bias correction using cloud-dependent predictors. Numerical experiments are performed using the hybrid-3DEnVar method of MPAS-JEDI with a global quasi-uniform 30km grid. Analyses and cloud layering forecasts are analyzed through brightness temperature and hydrometeors verification. Results indicate that variational bias correction is essential to alleviate observation-minus-background bias due to cloud effects as well as the need for an appropriate observations model error.