Uncertainties in Ensembles of Regional ReAnalyses
UERRA is a European FP7 reanalysis project of meteorological observations. It includes recovery of historical (last century) data, estimating uncertainties in the reanalyses and user-friendly data services. It aims to prepare for and contribute to a future Copernicus climate change service. It is led by the Swedish Meteorological and Hydrological Institute (SMHI) and runs from 1 January 2014 to 31 December 2017.
UERRA is designed to deliver observations and atmospheric data sets of climate quality and to show the quality and uncertainty for climate research and applications in Europe.
There will be delivered several new European reanalysis datasets and the different uncertainty estimates are completely novel. Several methods will be developed and explored. The time period is much longer than in earlier reanalysis datasets and the horizontal resolution is much higher. There are novel ensemble-based data assimilation methods that will be developed and verified in UERRA.
The meteorological conditions in Europe over a large part of the preceding decade will be reanalysed from all available observations. Data rescue of observations from manual archives through digitization and quality processing will continue from the EURO4M FP7 project, but with emphasis on the sub-daily time scale for reanalysis use. The reanalyses will both be made directly from observations (gridding) or, to a large extent, by assimilating comprehensive sets of many types of observations into meteorological models.
One measure of uncertainty will be estimated from the quality of observations. Most of the model assimilations will be run in ensemble mode and the spread of such reanalyses will provide another measure of uncertainty. The UERRA reanalyses will be made at quite high resolution, from 40 km of ensembles and 20, 11 km or 5 km for the various model-based reanalyses. All data sets can then be compared with independent observations not used in the reanalysis,e.g. from space or from gridded climate observations, which provides another estimate of uncertainty. The reanalyses will also be used in hydrological off-line models and give another independent measure. Statistical modelling and investigations of time and space dependencies will give further insight.
Large data sets of atmospheric and surface variables will be produced and archived, and data services will be provided to facilitate a wide use of the products among both scientists and policy makers in the society.
