A model for groundwater rise
- Insert x and y coordinates (km) of the topographical map of the Netherlands. The model only gives results for x = 70 – 180, and y = 410 – 474, and only within the Holocene delta. The result applies to the centre of a 1 km2 cell. No results are available for the last 2200 cal yr BP, due to the absence of peat.
- Present natural groundwater level is presumed to be approximately equal to surface level of distant flood basins.
- Time (Age) is in cal yr BP.
- GW level (m) is the calculated groundwater level, based on model results.
- KRIGVAR or Variance (in arbitrary units) is a measure of the distance to data points. The smaller this value, the closer the prediction for that position (x,y and age) is to an observation used for the interpolation (one of the >300 radiocarbon dates).
- STDDEV (m) is the standard deviation in the model result.
- Error (in meters) is the accuracy of the estimate. It is based on the Variance produced by the interpolation. The minimum error is ~0.15 m for index points within a series, and ~0.30 m for isolated index points.
- Software used for geostatistical interpolation: GSTAT.
- Data used: see Berendsen & Stouthamer (2001, their Appendix 1, updated). Available from the downloads section.
- Cite these data as: Cohen (2005), through www.geo.uu.nl/fg/palaeogeography .
Since 2003, ~30 additional groundwater rise index points have become available, and ~40 new dates are being processed. A new interpolation is expected in 2007. Sampling and dating activities are now concentrated on older basal peat (below 10 m-NAP, before 7500 cal yr BP) from the western part of the delta.
Dating channel belts, using the groundwater model
The model is a powerful tool to obtain approximate dates for the end of activity of channel belts, in case no direct radiocarbon dates are available yet. Figure 1 shows an example of a local groundwater curve near Montfoort (coordinates 124-448). The channel belt sands and Pleistocene substratum have been plotted in various colors. The top of the channel belt sands (relative to NAP = Dutch Ordnance Datum) is based on borehole information. The moment peat formation starts on top of the Pleistocene substratum (~ 8000 cal yr BP) is derived from the intersection of the groundwater curve with the top of the Pleistocene substratum. The intersection of the top of the channel belt sands with the groundwater curve gives an approximation of the moment the channel belt sands may have become covered by peat (groundwater had risen above the level of the channel belt sandbody's top). This method generally is accurate to within a few hundred years.
- Figure 1 Local curve of groundwater rise near Montfoort, and top level of channel belt deposits (= top of colored bands).
The final phase of activity of the channel belts, present in this area, has been independently radiocarbon dated (residual channel dates) as follows (Berendsen & Stouthamer 2001, updated):
- Hollandse IJssel channel belt: 650 cal yr BP
- Stuivenberg channel belt: 3379 cal yr BP
- Blokland channel belt: 4683 cal yr BP
- Cabauw channel belt: 6098 cal yr BP
- Benschop channel belt: 6950 cal yr BP
Approximately the same dates can be derived from the diagram. Other deposits at compaction-free levels can be dated in a similar way. The most accurate dates are obtained in the steepest part of the groundwater curve.
The deposits of successive channel belts are generally thinner if they are younger; this is related to the decrease of accommodation space as determined by sealevel rise. The thickness of the colored bands in Figure 1 is not necessarily related to the thickness of the deposits of a channel belt; in this case only the depth of the top of the channel belt sandbody is important.
The topsoil (upper 40 cm) is due to human activities: clay excavated from ditches was spread out over the land.
Total compaction can be estimated by substracting the compaction-free peat (=groundwater) levels that the model produces by the actual peat level of the same age in a core. In the western part of the delta, compaction may be as much as 2 m. For peats, underlying natural levee deposits, compaction values as high as 4 m have been recorded. A Ph.D. study, quantifying compaction starts in 2006.
Using groundwater model results in archeology
The moment the natural levees became covered by peat can be determined by plotting the top of the natural levees in a similar diagram as Figure 1. As soon as groundwater levels rose higher than the natural levees, the levees drowned and became covered by peat. This is of interest for archeology, because beginning peat formation normally means that settlements will become abandoned. Natural levees were standing out in the paleo-landscape and therefore it took longer (compared to the top of the sandbody) before they became covered by peat. In addition, the method can be used to check suspect radiocarbon dates.
Interpretation of groundwater model results on delta scale
During the Holocene, groundwater has been rising because of downstream eustatic sea-level rise and local tectonic land subsidence. There are spatial differences in groundwater levels (inland slightly higher than in the centre of the lagoon) and there are temporal trends in the rate at which groundwater was rising. In the Middle Holocene, as the North Sea approached its present extent, the delta was drowning the fastest. The more westward, the closer the curves of groundwater rise resemble sea-level rise, particularly after 8500 cal yr BP. Before 8500 cal yr BP, groundwater levels were rising slowly, and were not yet controlled by eustatic sea-level rise. Tectonic land-subsidence (mainly of glacio-isostatic origin) was the main contributor to relative groundwater rise in the Early Holocene.