SUMMARYStructure functions for the 3D-Var assimilation scheme of the European Centre for Medium-Range Weather Forecasts are evaluated from statistics of the differences between two forecasts valid at the same time. Results compare satisfactorily with those reported in the existing literature. Non-separability of the correlation functions is a pervasive feature. Accounting for non-separability in 3D-Var is necessary to reproduce geostrophic characteristics of the statistics, such as the increase of length-scale with height for the horizontal correlation of the mass variable, sharper vertical correlations for wind than for mass and shorter horizontal length-scales for temperature than for mass. In our non-separable 3D-Var, the vertical correlations vary with total wave-number and the horizontal correlation functions vary with vertical level.

Four-dimensional variational data assimilation (4D-Var) systems are ideally suited to obtain the best possible initial model state by utilizing information about the dynamical evolution of the atmospheric state from observations, such as satellite measurements, distributed over a certain period of time. In recent years, 4D-Var systems have been developed for several global and limited-area models. At the same time, spatially and temporally highly resolved satellite observations, as for example performed by the Spinning Enhanced Visible and InfraRed Imager (SEVIRI) on board the Meteosat Second Generation satellites, have become available.Here we demonstrate the benefit of a regional NWP model's analyses and forecasts gained by the assimilation of those radiances. The 4D-Var system of the HIgh Resolution Limited Area Model (HIRLAM) has been adjusted to utilize three of SEVIRI's infrared channels (located around 6.2 µm, 7.3 µm, and 13.4 µm, respectively) under clear-sky and low-level cloud conditions. Extended assimilation and forecast experiments show that the main direct impact of assimilated SEVIRI radiances on the atmospheric analysis were additional tropospheric humidity and wind increments. Forecast verification reveals a positive impact for almost all upper-air variables throughout the troposphere. Largest improvements are found for humidity and geopotential height in the middle troposphere. The observations in regions of low-level clouds provide especially beneficial information to the NWP system, which highlights the importance of satellite observations in cloudy areas for further improvements in the accuracy of weather forecasts.

SUMMARYStratospheric humidity analyses produced operationally by the European Centre for Medium-Range Weather Forecasts (ECMWF) are discussed for the period since late January 1996 when the practice of resetting the upperlevel specific humidity to a fixed value at each analysis time was abandoned. Near-tropopause analyses are in reasonable overall agreement with independent observations. Very low humidities occur in conjunction with deep convection and a particularly cold tropopause over the equatorial western Pacific during the northern winter. Drying occurs also in the cold core of the Antarctic polar-night vortex. The lower stratosphere is moistened in the outer tropics and subtropics in summer and autumn, predominantly in the northern hemisphere. Changes associated with the latest occurrence of El Niiio are illustrated.Analysed temperatures near the tropical tropopause are generally in good agreement with corresponding radiosonde measurements, with standard-level biases of the order of 0.5 degC or less. The past two years are the coldest by about 1 degC in a series of tropical mean 100 hPa analyses extending back to 1979. A cooling trend of about 0.6 degC per decade is seen in the global means of the 100 hPa analyses.Moisture is spread zonally and upward from the tropical tropopause as the data assimilation proceeds, but the rate of upward transfer is much faster than observed. Substantial lateral mixing can occur within the stratosphere over the course of a season. Moistening at middle and high latitudes due to mixing with more humid tropospheric air is confined, realistically, to a shallow layer at the base of the stratosphere.The rate of upward transfer of tropical stratospheric moisture is much more realistic in a multi-year simulation using a version of the model that has finer stratospheric resolution than the version used for the operational data assimilation. Temperatures at the tropical tropopause and in the Antarctic polar night are accurately simulated, apart from excessive persistence of cold south-polar temperatures in late winter and early spring. The latter is conducive to drying the model stratosphere; lack of a parametrization of moistening due to methane oxidation is an obvious deficiency in this regard.

Abstract. Sub-daily meteorological observations are needed for input to and assessment of high-resolution reanalysis products to improve understanding of weather and climate variability. While there are millions of such weather observations that have been collected by various organisations, many are yet to be transcribed into a useable format.Under the auspices of the Uncertainties in Ensembles of Regional ReAnalyses (UERRA) project, we describe the compilation and development of a digital dataset of 8.8 million meteorological observations of essential climate variables (ECVs) rescued across the European and southern Mediterranean region. By presenting the entire chain of data preparation, from the identification of regions lacking in digitised sub-daily data and the location of original sources, through the digitisation of the observations to the quality control procedures applied, we provide a rescued dataset that is as traceable as possible for use by the research community.Data from 127 stations and of 15 climate variables in the northern African and European sectors have been prepared for the period 1877 to 2012. Quality control of the data using a two-step semi-automatic statistical approach identified 3.5 % of observations that required correction or removal, on par with previous data rescue efforts.In addition to providing a new sub-daily meteorological dataset for the research community, our experience in the development of this sub-daily dataset gives us an opportunity to share some suggestions for future data rescue projects.All versions of the dataset, from the raw digitised data to data that have been quality controlled and converted to standard units, are available on PANGAEA: https://doi.org/10.1594/PANGAEA.886511 (Ashcroft et al., 2018).

An extended observation operator for the direct assimilation of cloud-affected infrared satellite radiances in the High Resolution Limited Area Model (HIRLAM) is examined. The operator includes a simplified moist-physics scheme, which enables the diagnosis of cloudiness in itself using background values of temperature, moisture and surface pressure. Subsequently, a radiative transfer model provides simulated cloud-affected radiances to be used as background equivalents to the satellite observations. The observation operator was evaluated by using infrared observations measured by the Spinning Enhanced Visible and Infrared Imager (SEVIRI). An observation-screening procedure, which incorporates SEVIRI cloudretrieval products, supports an improved selection of usable cloudy scenes, leading to good agreement between the observations and background equivalents. The tangent-linear observation operator was verified against finite differences from its nonlinear formulation. The increments revealed a near-linear behaviour for the selected channels for a large number of cases. The adjoint observation operator was used to derive brightness-temperature sensitivities with respect to temperature and moisture changes in the presence of radiance-affecting clouds. Differences from the clear-sky sensitivities were found in and below clouds. In a four-dimensional variational data assimilation experiment, cloud-affected SEVIRI observations were assimilated, resulting in additional increments in both moisture and wind fields. The corresponding analysis fields revealed a reduced deviation from the observations for the majority of all cloudy scenes and a reduced bias for wind and temperature in the upper troposphere against independent radiosonde observations. Overall, our results highlight the capability of this observation operator in the HIRLAM assimilation system and encourage its application for the extended usage of cloudy satellite observations in numerical weather prediction.