Introduction: High-resolution regional reanalysis is a research area of increasing scientific interest and high societal importance. There is, in addition, the need for the provision of improved and reliable climate services that are required by society to better face the impacts on the socio-ecosystems that climate change and natural climate variability poses worldwide. The EU-funded project entitled “Uncertainties in Ensembles of Regional Reanalyses” (UERRA, Grant Agreement 607193: http://www.uerra.eu/) is intended to (1) pioneer and create an ensemble system of regional reanalyses over Europe (ERRA) for weather and climate time scales, (2) assess uncertainties in the reanalysis and (3) develop friendly data services to contribute to future Copernicus’ climate change services.

Work Package 1 (WP1) of UERRA on Data Rescue and Development and Enhancement of Gridded and Observational Datasets is destined to recover, digitize and homogenize historical (last century) synoptic-scale observations over data-sparse regions of Europe, among other objectives. In this contribution, we assess (1) input data shortcomings, both in time and space at current ERRA (European Regional Reanalysis) products by exploring current data availability at the MARS Archive from the European Centre for Medium-Range Weather Forecast (ECMWF) and the International Surface Pressure Databank (ISPD), (2) we explore accessible data sources that provide scanned data from different holders, both available online and in physical archives, and (3) we provide some preliminary results on the current status of the digitisation task in WP1.

1. Exploring the basic input for European Regional Reanalysis

In WP1 Data Rescue (DARE) component, we explored, first, land surface climate data series availability at the ISPD (http://www.reanalysis.org /observations/international-surface-pressure-databank), which along with the International

Comprehensive Ocean-Atmosphere Data Set (ICOADS: http://icoads.noaa.gov/) for marine surface, constitute the basic input for global and regional Reanalysis. Second, we explored current land surface climate time-series availability at

the MARS Archive at ECMWF, which constitutes the basic input for European Regional Reanalysis (ERRAs) over the whole 20th century and integrate most of the main Essential Climate Variables at the finest time-steps (synoptical and

hourly observations).

Figure 1 provides snapshots of ISPDv3 stations availability at different years for the 20th century over Europe, while Figure 2 (3) gives SLP monthly coverage 1960-2010 available at the MARS Archive for the Mediterranean window (Eastern

Europe and Scandinavian countries), both showing predominance of stations with less than 20% of monthly coverage. The evaluation of both input data pointed to poor pre-1960 data availability, and the Mediterranean region, Eastern Europe

and Scandinavia as the European data-sparse sub-regions for the post-1960 period. In addition, a survey on the variables that can have a higher impact to improve ERRAs was carried out among UERRA Reanalysis people, pointing to the

need to recover hourly values for SLP, temperature, dew point, wind speed and direction and relative humidity, while the recovery of snow-depth and precipitation at the daily scale can help to downscale high resolution ERRAs.

2. Setting the targets for digitisation to enhance ERRAs

Building upon the climate data sources gathered under EU-EURO4M project and on additional on-line and physical archives searches, an initial UERRA’s DARE and digitisation plan was set

up. Table 1 shows selected data holders and sources from where synoptical observations are being digitised. The focus is set over the post-1960 period and over southern and Middle East

Mediterranean countries and over eastern Europe for all the mentioned synoptical observations, while some effort is also placed on recovering pre-1960 hourly observations over the

whole Europe. In addition, SD observations will be recovered over the Scandinavian countries.

3. Rending first results on the DARE effort carried out in URRA WP1

Digitisation of the identified and relevant hourly observation series started on early 2014. 10 URV students were contracted and they have been key entering the data from the gathered

sources , while a fraction of the sources are suitable to be digitised using OCR. Figure 3 (Figure 4) provides (a) details on the no. of observing times per stations being digitised and (b) on the

no. of variables observed at each station (current status of the already digitised data and under digitisation.

Further information at:

http://www.uerra.eu/work-

packages/wp1.html