ABSTRACT:The Met Office has developed a 4D-Var data assimilation system, which was implemented in the global forecast system on 5 October 2004. This followed a development path based on the previous 3D-Var configuration, with many aspects kept in common. A 4D-Var capability was provided by the introduction of a linear perturbation forecast model based on the Unified Model, the non-hydrostatic grid-point model producing our operational forecasts.There were clear advantages in verification of forecasts compared to the equivalent 3D-Var configuration, with an improvement of 2.6% in a composite skill score verified against observations during pre-operational trials. The largest differences in model evolution occur in storm-track regions in the extratropics. Overall, improvements in verification scores as measured against observations were larger than those measured against analyses, particularly at upper levels. There is an improvement in verification of surface parameters (10 m wind, 1.5 m temperature and relative humidity) against analyses. The strongest impact on fitting observations is seen for satellite radiances with weighting functions peaking in the stratosphere and upper troposphere. The largest changes to assimilation increments occurred in the top model levels, particularly wind increments which became much larger near the model top.Similarities were found in the signal of 4D-Var versus 3D-Var for models at two different resolutions, from which we infer that low-resolution trials remain valid for exploring some aspects of 4D-Var before confirmation in full-scale tests.  Crown

A particular problem encountered during development was significantly poorer tropical verification scores when measured against own analyses. In contrast, verification against independent (ECMWF) analyses gave scores that were much more consistent with those against observations.

The Met. Office has developed a variational assimilation for its Unified Model forecast system, which contains a grid‐point mode) that is run operationally in global, mesoscale, and stratospheric configuration. Key characteristics of the design are: a development path from three‐dimensional to four‐dimensional variational assimilation; global and limited‐area configurations; variational analysis of perturbations; and a carefully designed, well conditioned background term. The background term is implemented using a sequence of variable transforms to independent balanced and unbalanced variables, to vertical modes, and to spectral coefficients. The coefficients used are based on statistics from differences of one‐ and two‐day forecasts valid at the same time. The covariance model represents many of the features seen in the covariances of forecast differences. The three‐dimensional variational data assimilation (3D‐Var) system was implemented in the operational global forecast system on 29 March 1999. In parallel trials, the 3D‐Var system gave a 2.7% improvement in a composite skill score (verified against observations and weighted according to the importance of each field).

This paper describes the development and evaluation of the UK's new high resolution global coupled model, HiGEM, which is based on the latest climate configuration of the Met Office Unified Model, HadGEM1. In HiGEM, the horizontal resolution has been increased to 1.25 • x 0.83 • in longitude and latitude for the atmosphere, and 1/3 • x 1/3 • globally for the ocean. Multi-decadal integrations of HiGEM, and the lower resolution HadGEM, are used to explore the impact of resolution on the fidelity of climate simulations.Generally SST errors are reduced in HiGEM. Cold SST errors associated with the path of the North Atlantic drift improve, and warm SST errors are reduced in upwelling stratocumulus regions where the simulation of low level cloud is better at higher resolution. The ocean model in HiGEM allows ocean eddies to be partially resolved, which dramatically improves the representation of sea surface height variability. In the Southern Ocean, most of the heat transports in HiGEM is achieved by resolved eddy motions which replaces the parametrised eddy heat transport in the lower resolution model. HiGEM is also able to more realistically simulate small-scale features in the windstress curl around islands and oceanic SST fronts, which may have implications for oceanic upwelling and ocean biology.Higher resolution in both the atmosphere and the ocean allows coupling to occur on small spatial scales. In particular the small scale interaction recently seen in satellite imagery between the atmosphere and Tropical instability waves in the Tropical Pacific ocean is realistically captured in HiGEM. Tropical instability waves play a role in improving the simulation of the mean state of the Tropical Pacific which has important implications for climate variability.In particular all aspects of the simulation of ENSO (spatial patterns, the timescales at which ENSO occurs, and global teleconnections) are much improved in HiGEM.2

The Met Office has developed an ensemble-variational data assimilation method (hybrid-4DEnVar) as a potential replacement for the hybrid four-dimensional variational data assimilation (hybrid-4DVar), which is the current operational method for global NWP. Both are four-dimensional variational methods, using a hybrid combination of a fixed climatological model of background error covariances with localized covariances from an ensemble of current forecasts designed to describe the structure of ''errors of the day.'' The fundamental difference between the methods is their modeling of the time evolution of errors within each data assimilation window: 4DVar uses a linear model and its adjoint and 4DEnVar uses a localized linear combination of nonlinear forecasts. Both hybrid-4DVar and hybrid-4DEnVar beat their three-dimensional versions, which are equivalent, in NWP trials. With settings based on the current operational system, hybrid-4DVar performs better than hybrid-4DEnVar. Idealized experiments designed to compare the time evolution of covariances in the methods are described: the basic 4DEnVar represents the evolution of ensemble errors as well as 4DVar. However, 4DVar also represents the evolution of errors from the climatological covariances, whereas 4DEnVar does not. This difference is the main cause of the superiority of hybrid-4DVar. Another difference is that the authors' 4DVar explicitly penalizes rapid variations in the analysis increment trajectory, while the authors' 4DEnVar contains no dynamical constaints on imbalance. The authors describe a four-dimensional incremental analysis update (4DIAU) method that filters out the high-frequency oscillations introduced by the poorly balanced 4DEnVar increments. Possible methods for improving hybrid-4DEnVar are discussed.