This dataset is part of the regional hydrogeological mapping of the Calgary-Lethbridge Corridor (CLC) in southern Alberta. It represents the potentiometric surface of the Scollard / Willow Creek hydrostratigraphic unit (HSU) interpolated from wells in Alberta Environment and Parks' (AEP) Alberta Water Well Information Database (AWWID). The potentiometric surface was mapped at a regional scale using many years of data, thus providing a good indication of regional hydraulic head and flow direction rather than site specific values.
This dataset supplements Alberta Geological Survey Report 91, which includes details on the data processing and potentiometric surface interpolation in the CLC.
See DIG 2016-0022 and AGS Report 91 for details on the vertical accuracy of the input hydraulic head point data, which is affected by the locational accuracy of the wells.
The following statistics describe the accuracy of the interpolated grid in relation to the input hydraulic head point data:
Root mean square error = 3.8 m
Mean error = -0.4 m
Standard deviation = 3.7
Maximum error = +21.7 m/-33.3 m
Methodology to create this grid dataset is documented in AGS Report 91. AWWID dataset (from which hydraulic head values were sourced) was current up to April 2013.
Step 1 (Modelling the surface): An ordinary kriging algorithm was used in the ArcMap Geostatistical Analyst extension to interpolate the hydraulic head values and create a grid of the potentiometric surface of the Scollard / Willow Creek HSU. Wells in the AWWID are often only located to the middle of the quarter section (or sometimes section) and thus co-located data points exist. In these cases a mean of the hydraulic head was used for the interpolation. Initially, all wells in the AWWID with suitable hydraulic information (i.e. wells with only one well record, a single screened interval, a single static water level, and not under flowing artesian conditions) were used to interpolate the potentiometric surface. Wells with large deviations from the interpolated grid were examined and outliers removed if necessary. A final version of the grid was interpolated with remaining wells (wells found in DIG 2016-0022).
Method Report:
Method: CoKriging
Type: Ordinary
Output type: Prediction
Trend type: First
Trend removal: Local polynomial interpolation
Power: 1
Variogram: Semivariogram
Number of lags: 18
Lag size: 400
Nugget: 13.683
Measurement error percentage: 100
Model type: Stable
Range: 7200
Anisotropy: No
Partial sill: 707.176
Total points: 5870
Cell Size: 400 m
Root mean square error (considering only final clipped grid): 3.8 m