Linden Ashcroft, Digitising Sub-Daily Weather Observations in Europe to Improve Climate Reanalysis

An output from a regional reanalysis product built as part of the UERRA project. This image shows the differences between temperatures calculated using two different models over the European region. Source: UERRA/SMHI.

---

​Reanalyses are educated estimates of how the atmosphere behaved in the recent past. They are built by combining computer models of the atmosphere and ocean with weather observations to build an approximation of what the entire atmosphere was doing every six hours or so.
 
Climatologists and meteorologists all over the world use reanalyses to research our climate and weather. Most of the reanalysis products that exist cover the entire planet, but if you want to study storms over Paris for example, or wind changes in the Spanish Pyrenees, then it’s better to use a regional reanalysis, which has a higher spatial resolution.

​Uncertainties in Ensembles of Regional Reanalysis (UERRA) is a European Union-funded project that aims to build better regional reanalyses for Europe and the Mediterranean.
 
One part of UERRA is focused on recovering historical weather records that have a high temporal resolution (more than once a day) for regions that currently have poor data coverage. This is because one of the requirements for high-quality reanalysis is a lot of high-quality observations. The more observations you have, the better your estimation of the atmospheric conditions will be.
 
So far, UERRA and its partner national meteorological services have recovered over 8 million observations from 15 different countries, including Egypt, Slovenia, Germany and Romania. The observations range from snowfall totals on high mountaintops, to wind observations from the late 19th century in the Sahara, to relative humidity along the Mediterranean coastline. Our task now is to check the quality of these values before using them to build and verify the next generation of reanalyses for Europe.