Hip — Historical data — model uncertainty

For all the bearing data used in calibration and validation of DK model HIP, calculated mean error (ME) extracted from DK model HIP (100 m and 500 m) and time series with simulated and observed pressure level extracted from DK model HIP (100 m) are available. Both ME and time series have been extracted for the calibration period 2000-2010, as well as the validation periods 1990-1999 and 2011-2019. The resulting mean error (ME) values are usable for assessing the areas for which the model has the largest deviations from observation data. This difference may partly be due to uncertainties on the model, but may also be due to the fact that the bearing observations are not representative in relation to the model’s description of the depth to ground water. For example, some bearing intakes may be in, below or just above a clay layer where there may be pressure level differences between observed bearings and model, e.g. in outflows (areas of leakage towards e.g. watercourses) and inflows (areas with leakages). Time series of the depth of the off-road groundwater calculated with the model (phreatic depth) supplied as a time series for the grid in question may, as a result of such conditions, differ from time series of extracts from layer-statistics, which otherwise interpolate from four nearest grids. For example, in clay-laying areas, pictorials from the top 10 m may represent deeper aquifers, instead of representing the upper water table. In other cases, the model can simulate a ‘temporary’ hanging water table that occurs during the winter period, but which may not be described by bearing drillings filtered into layers in contact with deeper aquifers.