Providing initial conditions for limited-area forecasting models with small grid-lengths (typically 1-10km).
In high resolution limited-area models, the focus is on local and surface conditions. Data assimilation development for the 4km and 1.5km grid-length UK and the 12km grid-length North Atlantic and European (NAE) model is aimed at improving forecasts of key weather elements such as rain, cloud, visibility, 2-metre temperature and 10-metre wind, out to two days ahead.
The NAE model uses the same 4-dimensional variational data assimilation scheme (4D-Var) as the global model. 4D-Var is able to extract useful information from a sequence of observations distributed in time, whereas 3D-Var analyses in 3 spatial dimensions only. The UK models currently use 3D-Var, as 4D-Var is too computationally expensive to run operationally at such high resolution with our present computers.
Both the UK and NAE models assimilate some observations not used by the global model. Most of the extra data assimilated provides information on the moisture distribution. Cloud top temperature data from infrared satellite imagery are combined with surface cloud reports to produce a 3-dimensional cloud analysis in a Moisture Observation Pre-processing System (MOPS). These MOPS data have been found beneficial in the analysis and prediction of stratocumulus and in thunderstorm situations. During a stratocumulus episode, for example, assimilation of cloud data can lead to significant improvements in the accuracy of 2-metre temperature. Since 1996, rainfall data from the UK Weather Radar Network and nearby European countries have been assimilated by a technique known as latent heat nudging, giving improvements in precipitation forecasts. In future we hope to replace latent heat nudging with a 4D-Var scheme for rainfall assimilation. Visibility is related to the relative humidity and aerosol content of the atmosphere in a very nonlinear way, which poses a challenge for variational assimilation systems. Nevertheless a scheme to assimilate visibility observations into regional models has been developed and run operationally since 1999.
A number of studies have shown that there can be strong coupling between soil moisture and atmospheric variables. The Met Office uses the Unified Model soil moisture nudging scheme to analyse soil moisture. The nudging scheme operates on the principle that model errors in soil moisture lead to model errors in near surface temperature and humidity. Therefore, model errors in screen-level temperature and humidity can be used to diagnose model errors in soil moisture. Where the model boundary layer is too cool and moist, it is likely that the model soil is too moist. Where the model boundary layer is too warm and dry, it is likely that the model soil is too dry. Additional criteria are used so that the nudging scheme is only applied where there is a strong likelihood that model errors in screen temperature and humidity are mainly due to model errors in soil moisture. Recently the scheme has been supplemented by a method to assimilate ASCAT satellite-derived soil moisture data.
To develop data assimilation schemes for the 4km and 1.5km UK models and the 12km NAEmodel, aiming to improve forecasts of key weather elements such as rain, cloud, visibility, screen temperature and 10-metre wind.
To construct datasets which specify the surface conditions for the Unified Model. The datasets contain details of the land sea mask, orography, vegetation type, soil type, sea surface temperature, deep soil temperatures, sea ice, soil moisture and snow amount.
Develop an hourly data assimilation cycle for the 1.5km UK model.
Develop a scheme for 4D-Var assimilation of precipitation rate in the NAE model.
Develop a limited-area assimilation system within a European re-analysis project, capable of generating high-quality surface and hydrological fields suitable for climate monitoring.
Develop a land surface data assimilation scheme which can assimilate a mixture of satellite and conventional observations related to soil moisture.