WP1 - Data Rescue and development, gridded and obervational datasets [Months: 1-36]
Data rescue (DARE) is the all encompassing term used within climate science to include determining the location
of likely sources, organising plans for developing inventories and preserving (producing images of the material)
the data they contain and digitising these data into usable forms. Within UERRA, the emphasis is twofold: to infill
gaps over data-sparse European regions and the post-1950 period and to recover, digitise and develop longer
time-series since the beginning of the 20th century. The key aspect in any data rescue activity is to determine both
what is most needed (e.g. what will have the greatest effect in any new re-analysis) and what is possible given the
many NMHS constraints on accessing both the archives and more importantly their digital holdings.
Under UERRA a new battery of QC tests will be defined and the digitized synoptic-scale data will be assessed to
identify non-systematic errors at the sub-daily time scale. We will also explore the applicability of currently available state-of-the-art in homogenisation methods to adjust time-series at the hourly scale.
The E-OBS gridded observation dataset is being expanded across a number of regions where NMHSs are
supplying KNMI with more extensive versions of their daily station data (principally for Tx, Tn, precipitation, but
datasets of MSLP and snow cover are available for some parts of the continent). The differences in station
availability are now quite marked across some parts of Europe. UERRA will investigate whether this dramatic
difference in data density is adequately being catered for by the operational software. Specifically KNMI and UEA
will investigate different approaches to the gridding, such as transformations of precipitation data to improve the
interpolation of extreme values and to assess whether the resolution of the whole dataset could be improved in
some parts of the continent. This will involve developing a gridding tool as a specific deliverable. Another issue is
how susceptible the gridding and particularly the gridding of extremes might be susceptible to changes in station