Ensemble-based evaluation of the forecast impact expected from EPS-Sterna. Part II: Adding the 325 GHz channels

The proposed EUMETSAT Polar System (EPS)-Sterna constellation of small satellites comprises passive microwave (MW) sounding instruments with an additional set of novel 325 GHz channels, providing further humidity-sounding capabilities with higher ice-cloud sensitivity. These sub-millimetre channels have only recently become available from cross-track scanners on a space-borne platform. On EPS-Sterna they complement more traditional temperature and humidity-sounding channels around 50–60 GHz and 183 GHz. A strategy is developed to simulate and assimilate the 325 GHz channels in an all-sky framework. This includes development of an observation error model based on a new cloud indicator exploiting the cloud signal extracted from the lowest peaking 325 GHz channel to assign larger observation errors in cloud-affected regions. The new model leads to a more Gaussian distribution of background departures normalised by the assigned observation error, compared to using the same cloud indicator as for the 183 GHz channels. After developing an assimilation strategy, the Ensemble of Data Assimilations (EDA) method is used to evaluate the expected forecast benefit of these new channels. Results show that the 325 GHz channels produce a similar positive impact as the 183 GHz channels when each channel set is added separately to the temperature-sounding channels at 50 GHz. When combined with 50 GHz and 183 GHz sounding channels, the impact of the 325 GHz channels is mostly neutral, with benefits for mid-tropospheric relative humidity. Inflating the observation errors assigned to the 325 GHz channels is found necessary to achieve this impact. Potential avenues to increase the impact through data assimilation enhancements are discussed (e.g. from better utilising the ice cloud information).