Abstract
Although based on the same fundamental approach to data assimilation as the global and regional model schemes, the mesoscale version of the U.K. Meteorological Office data assimilation system receives a more varied observational input than its larger-scale counterparts. Several surface data types are assimilated only at mesoscale resolution. Also, a moisture observation preprocessing system blends information from satellite imagery, radar, and surface cloud reports with a model forecast to produce humidity soundings for assimilation by the model. Particular problems connected with cloud-top height assignment from infrared imagery are explained, along with solutions devised to overcome them. Examples are given of the impact on analyses and forecasts from assimilating these various mesoscale data sources.
