OR/14/051 Quantifying ground surface change

Whitbread, K. 2014. The geomorphic impact of road construction: a case study of the A9 in Scotland. British Geological Survey Internal Report, OR/14/051.

It is hypothesised that ground surface change associated with road construction will be characterised by ground lowering in areas of excavation marked by cuttings (worked ground) and raising in areas of deposition marked by embankments (made ground). This translates to the prediction that values of H_diff calculated using only boreholes drilled in the 1970s (prior to road construction) will be negative in areas mapped as worked ground (H_B>H_PT), and positive in areas mapped as made ground (H_B<H_PT).

The distribution of worked ground (WGR) and made ground (MGR) in the test areas are shown in figures in Appendix 1 - Maps of the distribution of made and worked ground. In the northern test area (Daviot-Moy) WGR covers approximately 47% of the road length and MGR covers approximately 36%. In the southern test area (Dunkeld) WGR covers 31% and MGR 25% of the road length. Borehole records were attributed according to whether they occur in areas mapped as made ground or worked ground. Boreholes lying outside areas of made and worked ground were not attributed with an artificial ground type.

Surface elevation change

Mean values of H_diff for the boreholes classified according to location in areas of mapped made or worked ground are shown for the different test areas in Table 4. For boreholes located in areas of made ground, the mean H_diff is significantly greater than zero for both test areas. The results indicate that where embankments have been constructed, the modern ground surface is on average 1.5 m higher than the pre-construction level in the Daviot area and 3.4 m higher in the Dunkeld area.

In areas of worked ground, the reverse is found; values of H_diff are significantly less than zero, and indicate that average reductions in the surface level in cuttings of 4.2 m and 4.7 m has occurred as a result of the road construction. H_diff values for each borehole relative to the extent of areas of made and worked ground along the road are shown for the northern test area in Figure 3 and the southern test area in Figure 4.

Table 4    Comparison of H_diff values for made and worked ground areas in the test sites

	Northern area (Daviot-Moy)			Southern area (Dunkeld)
Hdiff	MGR	WGR^	Unknown	MGR	WGR	Unknown
Mean	1.50	-4.29	-2.11	3.37	-4.22	-0.21
Std Dev	1.94	2.80	5.05	3.45	4.24	2.46
SE Mean	0.54	0.50	1.91	0.39	0.56	0.28
CI -	0.63	-5.30	-6.78	2.73	-5.15	-0.77
CI +	2.54	-3.28	2.56	4.01	-3.29	0.34
n	13	32	7	80	58	78
P for 1 sample t test*	0.006	<0.001	0.31	<0.001	<0.001	0.45

^  The values are for the sample with one outlier H_diff value of -19.29 removed (see Figure 3 for location of point).
Inspection of this borehole and its surrounding location suggests the borehole start height recorded in the log may be wrong.
*  For MGR and WGR one-tailed one sample t tests were used to assess whether the mean H_diff was greater than or
less than zero respectively. For unclassified boreholes a two-tailed test for difference of the mean to zero was used.

Figure 3    Plot of H_diff values for made and worked ground groups with distance along the A9 (measured from the northern edge of the test area) for the northern test area (Daviot-Moy). The distribution of made and worked ground along the section, taken at the centre of the road, is shown by the red and green lines at zero H_diff. (Note that most boreholes are not along the centre of the road and hence, in some cases, the borehole attribution does not appear to match the distribution of artificial ground).

Figure 4    Plot of H_diff values for made and worked ground groups with distance along the A9 (measured from the southern edge of the test area) for the southern test area (Dunkeld). The distribution of made and worked ground along the section, taken at the centre of the road, is shown by the red and green lines at zero H_diff. (Note that most boreholes are not along the centre of the road and hence, in some cases, the borehole attribution does not appear to match the distribution of artificial ground).

MGR in borehole records

Based on the comparison of ground surface elevations, the surface change in areas of made ground is estimated to be an increase in elevation of 1.6 to 2.7 m. This estimate is compared with actual values of made ground thickness in boreholes drilled along the road route after initial road construction (after c.1980). MGR thickness estimates are taken directly from the recorded logs and are therefore based on the driller’s interpretation. In some cases deposits not recorded as MGR were identified in logs, however, in most records it was not possible to verify the base of MGR from the descriptions given, hence the drillers interpretation was used.

The A9 route is largely through the rural upland landscape of the Grampian Highlands and pre- existing made ground is not considered likely to have been present along the route prior to road construction. This is confirmed by the borehole logs recorded in 1970–1977 site investigations; thin developments of made ground are only recorded in boreholes drilled where pre-existing railway lines, tracks or roads intersect the A9 route.

The thickness of MGR associated with the A9 road construction was estimated from boreholes drilled in 2005 in the Moy section of the northern test area. Pre-1980 boreholes in the same area have no made ground recorded, however all of the logs for boreholes and trial pits excavated in the vicinity of the road in 2005 record MGR deposits including Tarmac, concrete and sand and gravel fill.

Figure 5 shows a histogram of the thickness of MGR in 2005 Moy boreholes, indicating that there is a skewed distribution with most bores only proving MGR of less than 0.6 in thickness. By contrast the mean difference in surface elevation recorded for eleven 1970 boreholes that occur within the area covered by the 2005 site investigation is 1.19 m (SE = 0.43). The apparent thickness of MGR in the boreholes is considerably lower than the average thickness estimated from the change in ground level, only three of the 2005 boreholes record actual MGR thicknesses within this range.

It should be noted that many of the borehole records contain the drillers interpretation only and boreholes and trial pits are located close but peripheral to the existing road, or in central reservation areas, where the full depth of made ground may not be encountered. It is also significant that in most of these boreholes a thin layer recorded as MGR is underlain by deposits of sand, gravel and cobbles which are recorded in several boreholes as containing fragments of concrete, timber and/or plastic but not classified as MGR (Figure 6). This layer of sand, gravel and cobbles, occurs consistently in boreholes along the road route, even in areas where geological maps indicated that the road is underlain by rock or glacial till (stiff to very stiff clay with gravel, cobbles and boulders), and it is thought likely to be a sub-base layer of fill (i.e. MGR). The sand, gravel and cobble layer ranges between 1–1.5 m in thickness, consistent with the estimated MGR thickness derived from the ground level change (1.19 ± 0.42).

Figure 5    Histogram of depths of MGR recorded in boreholes in the Moy area drilled in 2005.

Figure 6    Log of trial pit for borehole record NH73SE39 showing 0.3 m of recorded MGR underlain by sand with gravel, cobbles and boulders containing fragments of plastic to a depth of 1.5 m below the surface.