Assessment of Precipitation Trends over Europe by Comparing ERA-20C with a New Homogenized Observational GPCC Dataset

Abstract:

Reliable data is essential for robust climate analysis. The ERA-20C reanalysis was developed during the projects ERA-CLIM and ERA-CLIM2. These projects focus on multi-decadal reanalyses of the global climate system. To ensure data quality and provide end users with information about uncertainties in these products, the 4^th work package of ERA_CLIM2 deals with the quality assessment of the products including quality control and error estimation.

In doing so, the monthly totals of the ERA-20C reanalysis are compared to two corresponding Global Precipitation Climatology Centre (GPCC) products; the Full Data Reanalysis Version 7 and the new HOMogenized PRecipitation Analysis of European in-situ data (HOMPRA Europe).

ERA-20C reanalysis was produced based on ECMWFs IFS version Cy38r1 with a spatial resolution of about 125 km. It covers the time period 1900 to 2010. Only surface observations are assimilated namely marine winds and pressure. This allows the comparison with independent, not assimilated data.

The GPCC Full Data Reanalysis Version 7 comprises monthly land-surface precipitation from approximately 75,000 rain-gauges covering the time period 1901-2013. For this paper, the version with 1° resolution is utilized.

For trend analysis, a monthly European subset of the ERA-20C reanalysis is investigated spanning the years 1951-2005. The European subset will be compared to a new homogenized GPCC data set HOMPRA Europe. The latter is based on a collective of 5373 homogenized monthly rain gauge time series, carefully chosen from the GPCC archive of precipitation data.

For the spatial and temporal evaluation of ERA-20C, global scores on monthly, seasonal and annual time scales are calculated. These include contingency table scores, correlation, along with spatial scores such as the fractional skill score. Unsurprisingly regions with strongest deviations are those of data scarcity, mountainous regions with their luv and lee effects, and monsoon regions. They all exhibit strong biases throughout their series, and severe shifts in the means.

The new HOMPRA Europe data set is useful in particular for trend analysis. Therefore it is compared to a monthly European subset of the ERA-20C reanalysis for the same period, i.e. the years 1951-2005, to study the ERA-20C capability in reproducing observed trends across Europe.