CDB description

The Climate Database (CDB) contains meteorological time series of daily temperature (maximum, minimum and mean) and daily total precipitation for more than 250 station sites in Trentino – South Tyrol region. The data were collected from the regional meteorological networks of Bolzano and Trento Provinces and include open access records from several close sites in Austria. The spanned period is 1950 – 2019 for temperature and 1946 – 2019 for precipitation series. For each site, the climatologies (30-year averages) over the 1981 – 2010 period of mean, minimum and maximum monthly temperature and monthly total precipitation are also provided. Climatological values are available for several additional Swiss and Austrian sites used within the processing phase. Note that mean temperature is defined by averaging minimum and maximum values in all cases.

The CDB was built in the framework of the Use-Case number 8 of the DPS4ESLAB project.

Data sources

The data comes from four sources:

  1. The WISKI database for station data of the province of Bolzano, which has an agreement with Eurac to freely use the meteo data. Any Eurac employee can request the data (and more like for example 10 minute series and other parameters) using an ICT ticket for Meteodata province. More information can be found in the knowledge area on the cockpit.
  2. The Meteotrentino website has all meteo data from the province of Trento freely available.
  3. The Swiss data is free for research use after registering at IDAWEB from MeteoSwiss.
  4. Some Austrian stations were used from the HISTALP project from ZAMG, which are free for research use at the HISTALP website

Data Processing

The collected series of daily temperature (maximum and minimum) and daily precipitation underwent a processing phase composed by four main activities:

  1. Merging: Meteorological records over consecutive periods collected by close sites were unified in a single series in order to improve the time coverage of observations for those locations.
  2. Quality check: All daily temperature and precipitation series were checked for quality and consistency in order to remove gross errors, outliers and suspicious entries. In addition, the monthly records for each station were simulated on the basis of surrounding series and possible inhomogeneous or malfunctioning periods were identified by the comparison with observations. The series which show low-quality data were flagged accordingly in the metadata.
  3. Homogenization: Homogeneity controls based on the Craddock test (Craddock, 1979) were applied to detect breaks and signals in the daily series due to non-climatic features, e.g. station relocations. The most evident inhomogeneities were corrected at daily scale by the estimation of specific correction factors. In the homogenization of minimum and maximum temperature the internal consistency of the series after the correction was preserved. In addition, the homogeneity test was applied only to the records with at least 15 years of data. Homogeneous series and homogenized records were labeled in the metadata accordingly. In some cases, the homogenization was not applied due to the difficulty in defining proper correction factors and to the lack of information about the station history which help a more accurate identification of artificial signals in the series due to changes of site and its surrounding. A specific label in the metadata was assigned to the series with possible inhomogeneities.
  4. Gap Filling: A statistical daily gap-filling procedure was performed on the quality-checked and homogenized records to maximize the data availability over the period spanned by each series. The reconstruction of the missing values for a series was applied only if a sufficient number of valid data was available in a window centered on the gap. Each missing value was computed by an anomaly-based procedure starting from the records of the station with the highest correlation and the greatest number of data in common with the series under reconstruction. As for the homogenization, the filling of minimum and maximum temperature series was checked in order to preserve the data consistency.

After the data processing, daily mean temperature series were computed as average of minimum and maximum temperature data.

The complete metadata of the data processing for each series including the details on removed values, homogenized periods and filled entries, are available on request from us (see contact above).

1981 – 2010 monthly climatologies

For the series with more than 10 years of available data, the monthly climatologies of maximum, minimum and mean temperature and total precipitation were computed for the reference period 1981 – 2010. The climatologies are mean values of reference computed over a 30-year interval and they represent the mean climatic conditions at a certain location.

In order to prevent the climatological values from being biased by the uneven data availability over the 30-year period, the missing monthly values in 1981 – 2010 interval for each station were reconstructed by means of the same method used for the series simulation in the quality check.

Further notes

The CDB does not provide real-time data but yearly updates are planned since 2020 in order to integrate to the current archive the most recent and validated station records.

Any information or request can be asked to: