BGR Bundesanstalt für Geowissenschaften und Rohstoffe

Groundwater Background Values

Country / Region: Germany

Begin of project: April 1, 2005

End of project: December 31, 2014

Status of project: December 31, 2014


Background:
The achievement of a good chemical status of groundwater, which is one of the main objectives of the European Water Framework Directive (EU WFD), requires the knowledge of the hydrogeochemical groundwater background values.

To fulfill the requirements of the EU WFD, the Geological Surveys of the Federal States of Germany (SGD) founded the task force EU-WRRL (UAG EU-WRRL) as a coordinating committee. A main product of the activities of the UAG EU-WRRL is the Hydrogeological Map of Germany 1:200.000 (HÜK200).

A subdivision of the UAG EU-WRRL, the working group "groundwater background values" (PK HGW) has been engaged since 2005 in identifying and statistically evaluating groundwater background values for the upper aquifer.

Project aim and approach:
The main objective of the PK HGW was the nationwide compilation and visualization of aquifer based groundwater background values in Germany on the basis of over 50,000 groundwater samples collected between 1980 and 2005. In a first step, the hydrogeological units (about 1100) of the Hydrogeological Map of Germany (HÜK200) were aggregated into hydrogeochemical (HGC) units, which are defined as hydrogeological units with typical distributions of hydrogeochemical properties.

After every groundwater sample had been assigned to the appropriate HGC unit, the data were subdivided into 10 hydrogeological regions (Fig. 1) and statistically evaluated by the Geological Surveys of the Federal States (SGD) and the Federal Institute for Geosciences and Natural Resources (BGR). This resulted in a total of 186 HGC units, of which 112 units contained sufficient (at least 10) samples to perform the statistical analysis. The evaluated units cover about 97.5% of Germany.

The HGC units for the northern region 1 were generated from the area related hydrogeological sub-regions, hence covering 100% of the region. Due to the strongly depth related variations in groundwater quality in the northern unconsolidated aquifers, the evaluation for this region was reduced to groundwater samples from a depth < 50 m.

Since the first data collection in the year 2005 showed an insufficient territorial coverage of samples for a number of trace elements, a new data collection was performed in the year 2012 for relevant trace elements with available minor threshold values (see also LAWA (2004): arsenic, boron, barium, cadmium, cobalt, chromium, copper, fluoride, mercury, molybdenum, nickel, lead, antimony, selenium, thallium, uranium, vanadium, zinc).

Method:
The statistical evaluation of each parameter was performed semi-automatically using probability plots (WALTER et al., 2006). This method allows the identification of separate populations and anomalies compared to the background or normal population, as different populations can be distinguished through changes in inclination of the slope (Fig. 2). These anomalies (MARCZINEK et al., 2008) could be caused by anthropogenic influences, but also by local natural phenomena like ore mineralization, coastal and inland salinization or acidification of crystalline rocks regions.

Example of statistical evaluation for a parameter in the probability plotFig. 2: Example of statistical evaluation

From the trend of the normal population in the probability plot the statistical values (mean, standard deviation and percentiles) can be calculated. In this project the background value was defined as the 90th percentile.

For an efficient evaluation, an excel-tool was developed, which allows the determination of the range of data, which represent the normal population of a data set (Fig. 2). The results of the procedure were reviewed visually by a comparison of a histogram of the data to the calculated normal distribution. Goodness of fit was also tested by the correlation coefficient and the d’Agostino-Pearson-test for normality. These methods allow an easy verification, if the assumed distribution fits the actual distribution. It is also possible to determine if the data are best represented by a normal or lognormal distribution.

The statistical evaluations were controlled and, if necessary, modified by the geological surveys responsible for the respective HGC unit. The tables with the results were then included into the database of the HÜK200.

A more detailed description of the method can be found in WAGNER et al. (2014).

Results:
The results of the data analysis are available for download via the BGR Product Center or can be accessed via the internet through a web map service (WMS) in the form of maps (Fig. 3) and info-queries (German only), showing the statistical values within the areal extent of each hydrogeochemical unit in the upper aquifer. In addition the WMS "HUEK200 HGW" is visualised in the BGR Geoviewer.

Up to 39 parameters (see Tab. 1) can be displayed, though for some parameters, especially trace elements, and for some small HGC units not enough samples were available to perform a statistical analysis (minimum requirement for statistical analysis: 10 samples).

Values for nitrate are not shown, because essentially all higher values of nitrate (> 10 mg/l) are of anthropogenic origin and nitrate is a seasonally very variable parameter, so that in most cases the statistic evaluation would not lead to representative background values.

A detailed description of the contents displayed in the WMS and more information about the project are given in WAGNER et al. (2014).


Geoviewer:

Maps:

Data:

Literature:

Partner:

Contact 1:

    
Dr. Bernhard Wagner
Phone: +49-(0)9281-1800-4740

Contact 2:

    
Dr. Stefan Broda
Phone: +49-(0)30-36993-250
Fax: +49-(0)511-643-531250

This Page: