Groundwater hydrology of boreal peatlands above a bedrock tunnel – Drainage impacts and surface water groundwater interactions

Groundwater hydrology of boreal peatlands above a bedrock tunnel – Drainage impacts and surface water groundwater interactions

Journal of Hydrology 403 (2011) 278–291 Contents lists available at ScienceDirect Journal of Hydrology journal homepage: www.elsevier.com/locate/jhy...

2MB Sizes 0 Downloads 3 Views

Recommend Documents

No documents
Journal of Hydrology 403 (2011) 278–291

Contents lists available at ScienceDirect

Journal of Hydrology journal homepage: www.elsevier.com/locate/jhydrol

Groundwater hydrology of boreal peatlands above a bedrock tunnel – Drainage impacts and surface water groundwater interactions Jens Kværner ⇑, Petter Snilsberg 1 Bioforsk – Norwegian Institute for Agricultural and Environmental Research, Soil and Environment Division, Frederik A. Dahls vei 20, N-1432 Ås, Norway

a r t i c l e

i n f o

Article history: Received 18 May 2010 Received in revised form 1 March 2011 Accepted 4 April 2011 Available online 12 April 2011 This manuscript was handled by P. Baveye, Editor-in-Chief Keywords: Groundwater Peatland Swamp Tunnel Drainage Hydrology

s u m m a r y Little attention has been given to the pattern of hydrological connectivity between peatlands and water bearing fracture zones in crystalline bedrock. The construction of the railway tunnel Romeriksporten provided an opportunity for studying impacts of bedrock tunnelling on peatland hydrology and the hydraulic connectivity between peat deposits and deeper layers in an area with fractured Precambrian gneisses, exposed bedrock, and surficial covers of thin till deposits and peatsoils. Above the tunnel the water level in peat wells fluctuated with maximum depths up to 3 m, and water that otherwise would have generated surface runoff infiltrated in the peat. Drawdowns of the groundwater table in peatlands were observed as far as 340 m away from the tunnel trace. The deep drainage base provided prolonged water table drawdowns in peatlands in dry periods, and differential settlements in drained peatsoils resulted in secondary changes in patterns of surface water storage and flow. The groundwater drawdowns were influenced by the balance between tunnel leakage and water supplies from catchments and the wetland position within the groundwater flow system. Deep and simultaneous lowering and fluctuations of hydraulic heads in wells in the peat, the subpeat sediments and the bedrock above the tunnel demonstrated hydraulic connectivity between the peat layers and the bedrock, and revealed vertical flow even through highly humified peat layers in the catotelm. This shows that in peatlands with a subsurface of a relatively thin till cover above fractured crystalline bedrock, favourable conditions for groundwater flow from peatsoils to rock are not always restricted to a few specific sites, and indicates that attention should be given to the influence of peat water–bedrock water connectivity on impacts of groundwater exploitation, droughts and climate changes in such areas. Ó 2011 Elsevier B.V. All rights reserved.

1. Introduction The increasing recognition of the importance of wetlands and the human exploitation of groundwater resources during the last decades have brought drainage impacts on wetlands and interactions between ecosystems and deeper aquifers into focus. In Europe, the new EU Water Frame Directive (European Commission, 2000) emphasizes the importance of considering these aspects in future groundwater management. Peatlands are the most widespread wetland type on earth, are found in all continents, and cover about 3 per cent of the earth’s terrestrial and freshwater surface (Joosten, 2004; Environment and Heritage Service Belfast, 2010). These areas are important for landscape and biological diversity (Korpela, 1998) and can also be important for regulations of catchment runoff (Kværner and Kløve, 2008). Peatlands are physically and ecologically adapted to stable water tables fluctuating near the surface. The water table levels in peatlands are crucial for the ⇑ Corresponding author. Tel.: +47 92491309; fax: +47 64948110. 1

E-mail address: [email protected] (J. Kværner). Present address: Asplan Viak AS, Raveien 2, N-1430 Ås, Norway.

0022-1694/$ - see front matter Ó 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.jhydrol.2011.04.006

ecological niches of plant species (Malmer, 1962; Økland, 1989; Rydin and Jeglum, 2006) and thus even for peat development (Ingram, 1983). Alterations in wetland hydrology and water table levels can have large effects on water quality, as wetlands can be transformers and sinks of different elements such as sulphate (Devito, 1995; Devito and Hill, 1997), nitrogen, phosphorous (Devito and Dillon, 1993) and methylmercury (Branfireun et al., 1996). In dry periods sulphides can be oxidised to sulphate, and stream acidification caused by episodic SO24 release from wetlands following droughts has been reported in several studies (Brække, 1981; Wieder, 1985; Bayley et al., 1986; van Dam, 1986). Moreover, alterations of water tables in peatlands might have important consequences on the release of greenhouse gases (Glaser et al., 2004; Strack et al., 2004; Strack and Waddington, 2007). The acrotelm–catotelm model of the structure of peatlands (Ingram, 1978, 1983) has gained wide acceptance and use. The model assumes that peatlands consist of two main layers, a surface layer (acrotelm) with thickness up to 0.5 m characterised by relatively undecomposed peat and high hydraulic conductivity, and the deeper peat (catotelm) with higher degree of peat decomposition and lower hydraulic conductivity. At present two

J. Kværner, P. Snilsberg / Journal of Hydrology 403 (2011) 278–291

alternative models of groundwater hydrology of peatlands (Reeve et al., 2000) implying different vulnerability to underground drainage exist. The shallow flow model is based on the acrotelm–catotelm concept and assumes that groundwater flow in peatlands is restricted to the acrotelm due to low permeability in deeper peat layers (Reeve et al., 2000; Siegel and Glaser, 2006). Since 1980s flow systems have been observed in deeper peat layers in several studies (Glaser et al., 2006), and an alternative model has been proposed, the groundwater flow model, which assumes that the extent of vertical groundwater flow in peatlands is primarily not controlled by the permeability of humified peat layers, but by the permeability of the underlying mineral soil (Reeve et al., 2000). Recent studies have shown that the type and extent of groundwater interaction can influence the amplitude and duration of water-level fluctuations in undisturbed peatlands, and that the hydrogeological settings, landscape position and climatic conditions may be important for the extent and role of groundwater–wetland interactions (Ferone and Devito, 2004; Vidon and Hill, 2004; Price et al., 2005). Although several peatland studies have examined relationships between groundwater and surface water in different hydrogeological settings (Vidon and Hill, 2004; Price et al., 2005), the research has mainly focused on interactions with surficial deposits and surrounding catchments. Little attention has been given to the pattern of hydrological connectivity between peatlands and water bearing fractures zones in crystalline bedrock. The issue of groundwater flow from soil to bedrock has been addressed only in a limited number of studies, and is still poorly understood (Rohde and Bockgård, 2006; Praamsma et al., 2009). Numerous tunnels and other underground constructions are being built in bedrock for different purposes such as roads, railways, water supply, sewage transport, mining, and hydropower development. Tunnel studies have usually focused on inflow and engineering aspects, whereas environmental impacts have received less attention. Groundwater decrease and inleakage of surface water have been described from several tunnels (Olofsson, 1993; Statens Offentliga Utredningar, 1998; Cesano et al., 2000; Mabee et al., 2002; Kim and Lee, 2003; Vincenzi et al., 2009). Although water contributions from peatland water storage to tunnels have been reported (Skjeseth, 1982; Olofsson, 1991), the spatial and temporal effects of tunnelling on peatland hydrology are seldom described. Studies of drainage impacts on peatland hydrology have traditionally been restricted to surficial drainage, as surface ditches for forestry, agricultural or peat mining purposes. Exploring the hydrological impacts of tunnel drainage on peatlands may improve the understanding of the hydrogeology and vulnerability of peatlands, and thus improve the basis for management of such ecosystems and adjacent deposits and bedrock. The construction of the railway tunnel Romeriksporten provided an opportunity for studying impacts of bedrock tunnelling on peatland hydrology and hydraulic connectivity between peat deposits and deeper layers. During the construction of the tunnel subsidence of the peat surface, slides, cracks, and dry holes in the peat deposits around Lake Northern Puttjern indicated that peatland hydrology was affected by tunnel leakage (Kværner and Snilsberg, 2008), and it was decided to monitor the groundwater tables in the peatlands. This paper presents the results from the geohydrological investigations carried out in the peatlands around Lake Northern Puttjern and adjacent areas. The objectives of the study were: (1) to reveal the spatial and temporal effects of tunnel leakage on water levels and hydrology of peatlands in a catchment with crystalline bedrock and thin till deposits, (2) to relate the hydrological effects on peatlands to catchment hydrology and wetland position, (3) to provide insight into leakage processes and hydrological connectivity between peatlands and bedrock.

279

2. Site description The study has been conducted in the Lake Northern Puttjern area above the Romeriksporten railway tunnel and nearby catchments in Østmarka, northeast of Oslo in Southern Norway (Fig. 1). The investigation comprises five peatland areas, the peatlands around Lake Northern Puttjern and Lake Southern Puttjern (70–350 m south-east of the tunnel trace), the mire Kjerringmyr north of Lake Northern Puttjern (100–250 m north-west of the tunnel trace), the peatlands 350–590 m south of Lake Southern Puttjern in the Lake Northern Puttjern valley (580–770 m south-east of the tunnel trace) and two reference peatlands situated west of Lake Lauvtjern and around Lake Rundtjern, respectively (Fig. 1b). The construction of the 13.7 km long tunnel between Oslo and Lillestrøm was commenced in 1994. Tunnelling under the Lake Northern Puttjern area started in autumn 1996 and excavation was finished September 4, 1997. The railway tunnel was officially opened August 21, 1999. A decline of the water level was discovered in Lake Northern Puttjern in February 1997 (Aars, 1998). During November 1997, 560 L min 1 of leakage was measured over a 600 m long section of the tunnel in this area (A–B Fig. 1b) (NSB Gardermobanen AS, 1998a). After the excavation was finished until February 1999, the tunnel was tightened by grouting. The grouting reduced the leakage in this section to 146 L min 1 in November 1999 (NSB Gardermobanen AS, 1999b). Between August 14, 1997 and September 8, 1998, water was periodically transferred to Lake Southern Puttjern from Lake Kroktjern in the neighbouring catchment to compensate for the leakage to the tunnel (Fig. 2f). In 1999 eighteen injection wells were drilled from the tunnel in the sections with most fractures (Kitterød et al., 2000). The wells were arranged in a fan-shape with the orientation ranging from almost parallel to the axis of the tunnel to an angle of 45°. During 1999 and 2000 water was injected in the bedrock from these wells in dry periods to counteract drawdown of groundwater and surface water tables (Fig. 2f). The climate in this area is suboceanic. Annual mean temperature is about 4.0 °C and annual mean precipitation is around 850 mm (Bendiksen et al., 2005). In 1997, 1998 and 1999 the annual precipitation was 587, 784 and 944 mm, respectively, at the nearby meteorological station at Blindern, whereas the annual mean precipitation at this station was 763 mm (The Norwegian Meteorological Institute, 2009). Average annual runoff for the normal period 1961–1990 is approximately 700 mm (Beldring et al., 2002). Precipitation normally falls as snow in December–March with snowmelt in April. The hydrological regime has a seasonal cycle with discharge minimums in summer and winter and maximums during spring due to snowmelt runoff and in autumn due to reduced evapotranspiration. The elevation of the study area around Lake Northern Puttjern is about 265 m above sea level, approximately 180 m above the Romeriksporten railway tunnel. The reference areas at Lake Lauvtjern and Lake Rundtjern are situated ca. 325 and 225 m above sea level, respectively. The catchments of Lake Northern Puttjern and the mire Kjerringmyr comprise 33.2 and 5.5 ha, respectively, whereas the catchments of the reference fields in the Lake Lauvtjern and the Lake Rundtjern areas cover 9.2 and 32.2 ha. The bedrock mainly consists of Precambrian gneisses (Berthelsen et al., 1996), belonging to the southeastern Norwegian basement, generated approximately 1.5–1.6  109 year BP (Graversen, 1984). The supracrustal gneiss is the oldest unit in the area, and is intruded by several generations of granitic and basic plutonic rocks. The intrusive events were separated by major deformation episodes. Four fold episodes of regional importance have been distinguished (Graversen, 1984). This has resulted in varying bedrock, with alternating rocks

280

J. Kværner, P. Snilsberg / Journal of Hydrology 403 (2011) 278–291

(a)

(b)

Fig. 1. The study area. (a) Rock types and weakness zones in the bedrock and (b) location of monitoring wells and peatlands. Rock types after Graversen (1984), weakness zones after Heimly (1994) and topography after Oppsal IF (1990, 1995).

like supracrustal gneiss and biotite gneiss (Fig. 1a). In addition dykes of pegmatite, diabase, etc. are common (Holmøy, 2008). During the Sweco-Norwegian orogenesis the bedrock was foliated in north– south direction. The area belongs to the bedrock zone outside the Oslo Graben, where Permian tension generated joints and faults appearing as north–south lineaments. The weakness zones in the bedrock extend mainly north–south and northwest–southeast and are distinguishable as depressions or valleys (Fig. 1). Many of these

are fracture zones parallel with the foliation of the rocks (Holmøy, 2008). Some weakness zones are faults. An evident fault strikes northeast–southwest and crosses the tunnel alignment in the Lake Northern Puttjern valley. Another fault crosses the mire Kjerringmyr and ends in the north-eastern part of Lake Lutvann (Holmøy, 2008). The study areas are located above the highest shoreline in the Oslo area, 221 m above sea level (Longva, 1991). Exposed bedrock and thin glacial till deposits (thickness <0.5 m) dominate the land

281

(a)

0.00 −1.00 −2.00 −3.00 2p 1p −4.00 2m −5.00 1/7−97 1/1−98 1/7−98 1/1−99 1/7−99 1/1−00 1/7−00 0.00

0.00

−2.00 −3.00

(b)

0.00

−3.00

Depth (m)

1.00

(c)

0.00

−1.00 −2.00 15p 15m

−3.00

−2.00

16p 16m

−3.00

Depth (m) Depth (m)

13p

1/7−97 1/1−98 1/7−98 1/1−99 1/7−99 1/1−00 1/7−00

(d)

0.00

(j)

−1.00

−1.00

−2.00

−2.00 17p 17m

−3.00

18p 18m

9p 9m

−3.00

10p 10m

−4.00 1/7−97 1/1−98 1/7−98 1/1−99 1/7−99 1/1−00 1/7−00

−4.00 1/7−97 1/1−98 1/7−98 1/1−99 1/7−99 1/1−00 1/7−00

(e)

0.00

(k)

−1.00

−1.00

−2.00

−2.00 19p 19d

−3.00

20p 20d

12p 14p 12m 14m −4.00 1/7−97 1/1−98 1/7−98 1/1−99 1/7−99 1/1−00 1/7−00 −3.00

400

(f)

0 Inleakage to tunnel Water transfer from Lake Kroktjern Water injection from the tunnel 1/7−97 1/1−98 1/7−98 1/1−99 1/7−99 1/1−00 1/7−00

Date

Sp.runoff (l s−1 km−2)

−4.00 1/7−97 1/1−98 1/7−98 1/1−99 1/7−99 1/1−00 1/7−00

−1

5p 5m

(i)

−1.00

−4.00 1/7−97 1/1−98 1/7−98 1/1−99 1/7−99 1/1−00 1/7−00

liter min

11p 11m

−4.00 1/7−97 1/1−98 1/7−98 1/1−99 1/7−99 1/1−00 1/7−00

−4.00 1/7−97 1/1−98 1/7−98 1/1−99 1/7−99 1/1−00 1/7−00

200

8p 8m

(h)

−2.00 3p 3m

−3.00

0.00

6p 7p

−1.00

−2.00

0.00

4p 4m

−4.00 1/7−97 1/1−98 1/7−98 1/1−99 1/7−99 1/1−00 1/7−00

−1.00

0.00

(g)

−1.00

250 200 150 100

0

(l)

20 Precipitation Discharge

40 60

50

80

0 100 1/7−97 1/1−98 1/7−98 1/1−99 1/7−99 1/1−00 1/7−00

Precipitation (mm)

Depth (m)

Depth (m)

J. Kværner, P. Snilsberg / Journal of Hydrology 403 (2011) 278–291

Date

Fig. 2. Timeseries of (a–e, g–k) depths to water tables in wells in peatlands, (f) tunnel leakage in the Lake Northern Puttjern zone after NSB Gardermobanen (1998b, 1999a,b), and periods with water transfer from Lake Kroktjern to Lake Southern Puttjern (after Færgestad, 1999) and injection of water into the bedrock from the tunnel, and (l) daily precipitation at Blindern (The Norwegian Meteorological Institute, 2009) and daily specific discharge at Gryta (The Norwegian Water and Energy Directorate, 2009).

surface of the uplands. The surficial deposits in depressions consist of peat and till. At the bottom of steep hill slopes rapid mass-movement deposits also occur. The peat thickness varies within the studied peatlands (Table 1). Maximum peat thickness in the peatlands west of Lake Lauvtjern is 5.0 m. The peat thickness ranges up to more than 6.0 m in the other peatlands studied. In most of the peatlands around Lake Northern Puttjern and Lake Southern Puttjern the peat is moderately decomposed (degree of humification H5–H6 according to the scale of von Post (1922)) in the upper meter. Below 2 m depth the degree of peat humification usually ranges from H5 to H8. The peat is most decomposed in the bottom layers. Along the southern side of Lake Northern Puttjern and the northern side of Lake Southern Puttjern, the peat is relatively undecomposed (degree of humification H2– H4) in the upper meter. In the northern parts of the mire Kjerring-

myr, the peat is relatively undecomposed (degree of humification H2–H4) in the uppermost 2–3 m and moderately decomposed (degree of humification H5–H6) between 3 and 6 m depths. In the other parts of the mire intermediately decomposed peat (degree of humification H5–H6) dominates the peat profile. In the peatlands west of Lake Lauvtjern and the northern parts of the peatlands at Lake Rundtjern, the peat is moderately decomposed (degree of humification H5–H6) in the upper 2 m and more decomposed (degree of humification H6–H8) below this layers. In the mire 350–590 m south of Lake Southern Puttjern and the peatlands around Lake Rundtjern the peat layers are less decomposed. The catchment uplands of the investigated peatlands are covered with Norway Spruce (Picea abies) and Scots Pine (Pinus sylvestris) forests. Whereas the lower-lying parts of the hill slopes are dominated of Bilberry (Vaccinium myrtillus) woodland and

282

J. Kværner, P. Snilsberg / Journal of Hydrology 403 (2011) 278–291

Table 1 Depth of well filters, peat humification according to von Posts classes (von Post, 1922) and peat thickness at different well sites in peatlands. Well

Filter

Degree of peat decomposition at different depths

Site

Depth (m)

0.3 m

0.5 m

1.0 m

1p 2p 2m 3p 3m 4p 4m 5p 5m 6p 7p 8p 8m 9p 9m 10p 10m 11p 11m 12p 12m 13p 14p 14m 15p 15m 16p 16m 17p 17m 18p 18m 19p 19d 20p 20d 21p

1.10–2.10 1.73–2.73 6.00–7.00 1.30–2.30 2.80–3.80 1.95–2.95 2.40–3.40 1.80–2.80 2.10–3.10 0.65–1.65 1.80–2.80 0.80–1.80 4.25–5.25 0.30–1.30 0.70–1.70 2.80–3.80 4.22–5.22 0.50–1.50 4.00–5.00 0.95–1.95 2.10–3.10 1.50–2.50 1.40–2.40 4.60–5.60 1.32–2.32 2.42–3.42 1.40–2.40 5.00–6.80 0.50–1.50 5.30–6.30 1.33–2.33 6.15–7.15 3.85–4.85 4.10–5.10 1.47–2.47 4.56–5.56 0.62–1.62

4 3 3 5 5 6 6

5 4 4 5 5 5 5

7 5 5 5 5 5 5

6 6 5 5 6 6 6 6 5 5 5 5 6 6 6 3 3 2 2 5 5 5 5 5 5 3 3

5 6 5 5 5 5 5 6 5 5 6 5 6 6 6 4 4 3 3 5 5 6 6 5 5 3 3

4 6 6 6 6 6 6 6 5 5 6 6 6 6 6 5 5 3 3 6 6 6 6 5 5 3 3

1.5 m

5 5

Peat 2.0 m

3.0 m

4.0 m

5.0 m

6 6 6 3 3 6 6

6 6 6 6 6

6 5 5

5 5

5 6 6 6

5

8

5 5

5 5

6 7

5 7 6 7 7 6 6 4 4 5 5 5 5 5 5 4 4

5

7

7 7 6

7

4 4 6 6 7 7 5 5 4 4

5 5 8 8 8 8 5 5 5 5

6.0 m

6 6 7

7 7

7 7 5 5 3 3

5 5 8 8

5 5 5 5

5 5 5 5

Depth (m) 5.5 6.0 6.0 2.6 2.6 3.0 3.0 >2.8 2.7 4.5 2.9 2.8 2.8 1.5 1.5 5.8 4.8 1.5 1.5 4.8 2.8 3.7 3.3 3.3 5.1 3.0 6.0 6.0 5.0 5.0 4.5 4.5 >6.0 >6.0 >6.0 >6.0 >6.0

small-fern woodland with Bilberry and Thelypteris phegopteris in the field layer, lichen (Cladonia) woodland and Cowberry (Vaccinium vitis-idaea) – Bilberry woodland dominate the upper parts of the uplands The peatlands studied are dominated by minerotrophic swamp- and mire (fen) elements, but at Kjerringmyr and around Lake Rundtjern ombrogenous mire (bog) elements also occur. Most of the peatlands around Lake Northern Puttjern and Lake Southern Puttjern are vegetated by forested swamps with Norway Spruce, Bilberry and Sphagnum girgensohnii. At the southern side of Lake Northern Puttjern and along the southern and northern sides of Lake Southern Puttjern, the vegetation is dominated by poor low fen carpets. Also along the western valley side between Lake Northern and Lake Southern Puttjern poor low carpets occur. At the mire Kjerringmyr the vegetation cover consists of swamp forests with Norway Spruce, Scots Pine, Downy Birch (Betula pubescens), and poor mire hummocks and carpets. Forested swamps with Norway Spruce and Bilberry dominate the vegetation at the monitoring site west of Lake Lauvtjern. The peatlands 350–590 m south of Lake Southern Puttjern and around Lake Rundtjern are characterized by poor mires and forested swamps with Norway Spruce.

established November 1997 in this area, at the mire Kjerringmyr, in the peatlands south of Lake Southern Puttjern, and in the reference peatlands at Lake Lauvtjern and Lake Rundtjern (Fig. 1b). Piezometers were installed both in peatsoils and subpeat tills at most monitoring sites. At some locations at Lake Rundtjern with particularly thick peat deposits, piezometers were installed at two depths in the peatsoils. The piezometers in peatsoils consisted of PEH (polyethylene high density) pipes with 6 cm in diameter above filters of slotted pipe 1 m in length, whereas the piezometers in tills were made of steel tubes with a diameter of 3.2 cm and 1 m long brass filters. The bottom of the piezometers was closed, and the top was covered against precipitation. Depths of well filters, peat humification and peat thickness at the different well sites are shown in Table 1. In November 1997 boreholes were drilled in the bedrock in the mire Kjerringmyr (B1) and the Lake Rundtjern area (B4) (Fig. 1b). Two additional wells were drilled in the bedrock north (B2) and north-east (B3) of Lake Southern Puttjern in November 1998 (Fig. 1b). The diameter of the boreholes was 135 mm, and the depths ranged from 60 to 98 m. The boreholes B1 and B3 were vertical, whereas B2 and B4 inclined with 60° dip.

3. Methods

3.2. Depths to groundwater tables and groundwater elevation levels in peatlands

3.1. Monitoring wells In July 1997 9 and 2 piezometers were installed in peatsoils and subpeat tills, respectively, in the peatlands between Lake Northern Puttjern and Lake Southern Puttjern. Additional piezometers were

The depths to the water tables in piezometers and boreholes in the bedrock were measured manually using a measuring tape with a dipper at the end. In the wells established in July 1997 water tables were measured at a few occasions until September 1997. Be-

J. Kværner, P. Snilsberg / Journal of Hydrology 403 (2011) 278–291

tween December 1997 and June 1998 depths to water tables were measured weekly in all wells. From July 1998 until April 2000 the water stages in the piezometers were registered twice or once a month. In some periods the water elevation levels in selected wells were automatically recorded at 6 h intervals by pressure transducers and stored in logger devices (SEBA MDS III). All water tables were measured relative to the top of the wells. In the Lake Northern Puttjern area, at the mire Kjerringmyr and partly in the Lake Rundtjern area the levels of the well tops were surveyed relative to benchmarks with known altitudes. 3.3. Peatland surface water The occurrence of stream water and the directions of surface flow in the Lake Puttjern area were observed, and the outlet thresholds of Lakes Northern and Southern Puttjern were elevated to uncover the relation between lake water tables and outflow of stream water. Also temporary dams with storage of surface water in depressions in the peatlands around Lake Northern Puttjern and at the mire Kjerringmyr were observed during the monitoring period. 3.4. Data analyses The analysis and comparison of groundwater table patterns in the different areas were based on relative water table levels and the distance between the terrain surface and water tables, whereas comparisons of water levels in the peatlands with lake water levels and groundwater levels in bedrock were based on absolute water elevation, i.e. the water table position above a common reference datum, e. g. the mean sea level. Different graphical presentations and statistical methods have been used to visualise and analyse the groundwater table data. Boxplots were used to show main differences in distribution patterns of water tables. Correlation coefficients were calculated for comparisons of the degree of association between depths to water tables in wells at different locations and depths. The graphical and statistical analyses were carried out with Matlab Version 7.1. 4. Results 4.1. Depths to groundwater tables Based on the means and amplitudes of the depths to the water table, the peatland wells cluster in two groups (Fig. 3). Around Lake Lauvtjern and Lake Rundtjern and in the peatlands 350–590 m south of Lake Southern Puttjern, the water tables were fairy stable with maximum depths in the range 0.1–0.6 m below surface (Figs. 2 and 3). The depths were clearly less than in the Lake Northern Puttjern area and at the mire Kjerringmyr, where water levels in peat wells fluctuated with maximum depths ranging up to 3 and 1.26 m, respectively (Fig. 2). In the Lake Northern Puttjern area the water table oscillations and maximum water drawdowns in peat wells were larger near the tunnel trace than around Lake Southern Puttjern. However, considerable fluctuations and groundwater drawdowns to depths of 1.39 and 1.59 m were observed in wells 14p and 14m, respectively, as far as 340 m southeast of the tunnel trace. In the Lake Northern Puttjern and Kjerringmyr areas the depths to the water tables also increased with the depths of the well filters (Fig. 4). The pattern and magnitudes of groundwater fluctuations in peatlands were similar throughout the different years of observation in the Lake Lauvtjern and the Lake Rundtjern areas (Fig. 2). In contrast, in the peatlands in the Lake Puttjern area the water table oscillations changed with different fluctuation pattern during

283

(i) 1st July 1997–1st July 1998, (ii) 1st July 1998–1st May 1999 and (iii) 1st May 1999–1st July 2000. The first period was characterized by deep groundwater levels or short periods with high groundwater levels alternating with rapid sinking to deep levels. During this period the water table tended to fluctuate more frequently in the wells around Lake Southern Puttjern than in the wells close to Lake Northern Puttjern. The second period was characterized by intermediate drawdowns of groundwater tables. The third period was characterised by small fluctuations and drawdown of groundwater tables. The correlation matrix for the water tables in different peat wells in 1998 revealed that the wells clustered in groups with different patterns of water table variation (Fig. 5). The water table fluctuations in the peat wells at Kjerringmyr (1p, 2p, 3p), at the southern side of Lake Northern Puttjern (4p, 5p, 6p, 7p) and along the eastern side of the peatlands between Lake Southern and Lake Northern Putttjern (10p, 12p) were correlated with correlation coefficients usually in the range 0.7–0.9. Also the water table variations in the wells 8p, 9p and 13p located north of the Lake Southern Puttjern were highly correlated. Similarly, the water table oscillations in the peat wells at Lake Lauvtjern, in the peat wells in the mire 350–590 m south of Lake Southern Puttjern, and most of the upper peat wells around Lake Rundtjern were correlated, with correlation coefficients of magnitude 0.7. With the exception of well 9m, the water table fluctuations in the wells in subpeat mineral deposits at Kjerringmyr and in the Lake Puttjern area were characterized by high correlation coefficients for water tables, often of magnitude 0.7–0.9. The water table fluctuations in the wells in the subpeat mineral deposits at the Lake Lauvtjern, in the mire 350–590 m south of Lake Southern Puttjern and at Lake Rundtjern were characterized by low correlation coefficients with other wells. At most well sites in the Lake Northern Puttjern and Kjerringmyr areas water table changes in adjacent wells in peat and subpeat sediments were characterized by correlation coefficients in the range 0.87–0.98, but at the well localities 10, 11 and 12 the coefficients were smaller (Fig. 5). In episodes with periodical injection of water in the bedrock from the tunnel, the water tables in the wells 3m, 8m, 11m and 12 m in subpeat mineral deposits in the peatlands at Kjerringmyr and between Lake Northern Puttjern and Lake Southern Puttjern fluctuated in time with the injection periods (Fig. 6). The amplitudes of the fluctuations were larger in the wells 8m and 12m than in 3m and 11m. Similar oscillations were not observed in the peat wells 3p, 8p, 11p and 12p. However, during the period with water injections in the bedrock in the middle of August 1999, the depths to the water table in the peat wells 3p, 8p and 11p to some extent followed the main pattern of water injection and the average water table in the bedrock above the injection wells, contrasting the recession pattern of the water table in the reference well 19p in the Lake Rundtjern area (Fig. 6). The groundwater tables rose rapidly at precipitation events, and fell in dry periods with little precipitation and water deliveries from the surrounding catchments (Fig. 2). The groundwater table declines in dry periods in peat and subpeat till layers were clearly faster in the Lake Northern Puttjern area than around Lake Lauvtjern and Lake Rundtjern. For example from May 22 to May 29 the average daily decline in water tables in the peat wells between the Lakes Northern and Southern Puttjern area ranged from 3.3 to 8.7 cm whereas the daily fall in water levels in the peat wells at Lake Lauvtjern and Lake Rundtjern ranged from 0 to 1.1 cm. 4.2. Groundwater elevation levels Between July 1997 and August 2000, the water table levels in the peat wells between Lake Northern Puttjern and Lake Southern Puttjern fluctuated between the lake surfaces (Figs. 7 and 8). The

284

J. Kværner, P. Snilsberg / Journal of Hydrology 403 (2011) 278–291

200 100

(a)

0

Depth (cm)

−100 −200 −300 −400 −500 −600

1p 2p 3p 4p 5p 6p 7p 8p 94p 10p 11p 12p 13p 14p 15p 16p 17p 18p 19p 20p 21p

Observation wells 200 100

(b)

0

Depth (cm)

−100 −200 −300

Higher inner fence th 75 percentile Median th 25 percentile Lower inner fence

−400 −500

Value beyond fences −600

2m 3m 4m 5m

8m 9m 10m11m12m

14m15m16m17m18m 19d 20d

Observation wells Fig. 3. Boxplot of depths to water tables for (a) wells in the uppermost 4 m of peat layers and (b) wells in subpeat mineral soil layers and deeper peat layers (wells 19, 20) in 1998.

water tables in the wells in the mineral deposits below the peat layers were also usually situated between the surface levels of the two lakes. In March 1999, however, the water levels in some wells were below the water tables in both lakes. In the wells south of Lake Northern Puttjern the elevation of the water tables usually increased with increasing distance from the lake. After January 1999, the water level in well 4p and 4m at the southern side of Lake Northern Puttjern closely followed the water table in Lake Northern Puttjern. Between January 1997 and May 1998 the water levels in the wells in peat and subpeat sediments at the southern side of Lake Southern Puttjern (14p, 14m) were periodically lower than the lake surface. Later the water table in these wells was above the lake surface. Whereas the hydraulic head was greater in the bedrock than in the peat layers at Lake Rundtjern, the opposite situation often occurred at Kjerringmyr, particularly in the first part of the observation period (Fig. 9).

tween Lake Southern Puttjern and the depression. Except during spring, stream water was never observed between the depression north of well 8p and Lake Northern Puttjern. From 1997 until October 29 1998, runoff did not occur from Lake Northern Puttjern valley. Afterwards periods with and without surface runoff alternated. The outlet stream from Lake Northern Puttjern normally flows towards north. During 1997 and summer 1998, no outflow occurred (Fig. 7) and the incline and flow direction of the outlet stream were reversed over a distance of 90 m. Locally altered terrain slopes and surface flow directions also occurred around peatland depressions that came into being on the northern parts of Kjerringmyr and along the eastern and western boundaries of the peatlands between Lake Southern Puttjern and Lake Northern Puttjern (Fig. 1b). During wet periods temporary storage of surface water was observed in these depressions.

4.3. Peatland surface water

5. Discussion

The pattern of surface flow in the Lake Northern Puttjern area was altered during the study period. The inlet stream south of Lake Southern Puttjern was dry in summer 1997, but water bearing at later inspections. Between June 1997 and November 1997 no runoff occurred from Lake Southern Puttjern (Fig. 7). Later, except for minor periods, surface water streamed from the northern end of the lake, towards a local depression in the peatlands 40 m north of well site 8, where the water infiltrated in the peat. In some periods all stream water infiltrated into the peatsoils be-

5.1. Groundwater table depths and hydrological impacts of tunnel drainage on peatlands The similar pattern of water table changes in peatland and bedrock wells in the reference areas at Lake Lauvtjern and Lake Rundtjern reflects that the water table fluctuations in the different subsurface layers are determined by the same hydrologic processes. The larger vertical amplitudes in bedrock reflect the low porosity of rocks, as total and effective porosities of magni-

285

J. Kværner, P. Snilsberg / Journal of Hydrology 403 (2011) 278–291

Depth to water table (m)

(a) 2.00 1.80 1.60 1.40 1.20 1.00 0.80 0.60 0.40 0.20 0.00 800

16

700

600

5

11 15 2

500

1

14

400

300

9 12 8

200

100

0

7

6

13

10 4

3

2

1

0

Distance from tunnel trace (m)

3

5

4

6

7

8

Depth of well filter (m)

Depth to water table (m)

(b) 2.00 1.80 1.60 1.40 1.20 1.00 0.80 0.60 0.40 0.20 0.00 800

11

16 15

700

2

14

600

500

3 10 8

12

400

300

200

100

0

Distance from tunnel trace (m)

4

5

9

0

1

2

3

4

5

6

7

8

Depth of well filter (m)

Fig. 4. Mean depth to water tables in 1998 for wells with different filter depths and distance from the tunnel trace. (a) Wells in the uppermost 4 m of peat layers, (b) wells in subpeat mineral soil layers and deeper peat layers (wells 19, 20).

tudes 0–10% and 0.00005–0.01%, respectively, have been reported from fractured crystalline rocks (Domenico and Schwartz, 1997), contrasting porosities of size 81–97% and 30–65% found in respectively peatsoils and mineral soils (Paavilainen and Päivänen, 1995). The observed depths to the water table in previous studies of pristine peatlands are generally small, despite climatic and seasonal variations. At the Åkhult mire in Sweden the depths to the water table usually ranged between 0.15 and 0.20 m (Malmer, 1962). In the mire Northern Kisselbergmosen in Norway a maximum average depth to water table of 0.37 m was observed (Økland, 1989). Maximum depths to the water table in the range 0.35–0.50 m have been reported from several Scandinavian mires (Malmer, 1962; Reinikainen, 1984; Kværner and Kløve, 2006) and swamps (Økland et al., 2001). Bragazza and Gerdol (1996) measured a maximal depth to the water table of 0.72 m in an Italian mire, whereas Devito et al. (1996) reported maximum water table depths of 0.14 and 0.625 m in two Canadian swamps. During a multiyear drought, however, water table drawdowns of over 1 m were observed on a raised bog in Northern Minnesota (Glaser et al., 1997). Thus, the observed depths to water tables in the peat wells in the Lake Lauvtjern area, in the Lake Rundtjern area, and in the mire 350–590 m south of Lake Southern Puttjern ranged within the depths usually observed in pristine peatlands. The water tables fluctuations in the peat wells in the Lake Puttjern area and at the mire Kjerringmyr with depths ranging up to 3 and 1.26 m, respectively, were clearly larger than in the reference areas. Although the filter depths will influence hydraulic heads in

piezometers in areas with vertical flow and thus complicate detailed interpretation and comparisons of differences in water table depths, the large depths to the water table in the peatlands in the Lake Northern Puttjern and Kjerringmyr areas compared to the shallow depths observed in the reference areas and in other studies in pristine peatlands, demonstrate that the tunnel leakage caused considerable water table drawdowns and increased seasonal fluctuations of the water tables in the peatlands in the Lake Northern Puttjern and Kjerringmyr areas. The depths to the groundwater table in the peatlands in the Lake Northern Puttjern and Kjerringmyr areas were clearly larger than the depths of magnitude 0.79 m observed by Hillman (1992) in peatlands with surficial drainage. This can be explained by the deep drainage base that provides prolonged drainage and water table drawdowns in peatlands during drought periods, contrasting surficial ditches, which primarily removes moisture surplus in wet periods by surface and surface near flow. The tendency to decreasing water table level drawdowns with increasing distance south from Lake Northern Puttjern reflects greater drainage impacts nearest the tunnel trace. The water table fluctuations and the large water table depths in the wells 14p and 14m at the southern side of Lake Southern Puttjern in February 1998 demonstrate that the tunnel affected peatland water tables 340 m away from the tunnel trace. This coincides with other studies in fractured bedrock demonstrating impacts far away from the tunnel, sometimes extending 1–2 km from the tunnel trace (Ishizaki, 1979 cited in Olofsson, 1993; Skjeseth, 1982).

286

J. Kværner, P. Snilsberg / Journal of Hydrology 403 (2011) 278–291

Observation wells

(a) 1p 2b 3p 4p 5p 6p 7p 8p 9p 10p 11p 12p 13p 14p 15p 16p 17p 18p 19p 20p 21p 1p 2b 3p 4p 5p 6p 7p 8p 9p 10p 11p 12p 13p 14p 15p 16p 17p 18p 19p 20p 21p

Observation wells

Observation wells

(b) 1

2m 3m 4m 5m 8m 9m 10m 11m 12m 14m 15m 16m 17m 18m 19d 20d

0.8

0.6

0.4

0.2

0

2m 3m 4m 5m 8m 9m 10m11m12m14m15m16m17m18m 19d 20d

Observation wells

(c) 2

3

4

5

8

9

10

11

12

14

15

16

17

18

18

20

Observation wells Fig. 5. Correlation between depths to water tables in 1998 for (a) wells in the uppermost 4 m of peat layers, (b) wells in subpeat mineral soils and deeper peat layers and (c) closely located wells with filters at different depths. The figure shows correlation coefficient between pairs of observation wells.

The rapid drawdowns of water tables in the Lake Puttjern area in dry periods indicate short transit times of groundwater in peat layers and subpeat till deposits. The faster drawdowns in monitoring wells compared to the water table changes in Lake Northern Puttjern indicate lower transit times in peatlands than in the lakes. The slow water table drawdowns in the undisturbed reference peatlands around Lake Rundtjern and Lake Lauvtjern indicate relatively small evapotranspiration in May 1998. Assuming evapotranspiration was of minor importance for alterations of water tables in this period, the velocities of water table drawdowns in the peat and peat subsurface wells between Lake Northern Puttjern and Lake Southern Puttjern in the range 4–8 cm day 1 correspond to water transit times of magnitudes 12–25 days in 1 m thick layers of peat or subpeat till deposits in these areas. The lowering of water tables in dry periods and the decreasing hydraulic heads with increasing filter depths in adjacent wells demonstrate large recharge at the Lake Northern Puttjern and Kjerringmyr areas compared to the reference area at Lake Rundtjern, and may indicate that the tunnel leakage has lead to an increase of the recharge areas. The long periods without surface

runoff from the Lake Northern Puttjern valley, and the dry outlet stream from Lake Southern Puttjern in large parts of the summer and autumn 1997 despite water deliveries from uplands, illustrate that the tunnel leakage resulted in water that would otherwise appear on peatland surface as stream water, was infiltrating in the subsurface. The new depressions with temporary surface water storage in the peatlands at Lake Northern Puttjern and Kjerringmyr and the reversed incline of the outlet stream from Lake Northern Puttjern probably reflects differential settlements in drained peatsoils (Kværner and Snilsberg, 2008) and show that groundwater lowering can lead to secondary hydrological changes in peatlands. Some of the reason for the temporary storage of surface water in the subsidence depressions, despite the tunnel drainage, is that water that leaks to underground constructions, opposed to surficial drainage, has to infiltrate and percolate through the subsurface. Peak water tables 40 cm above terrain level in peat wells situated in subsidence depressions (2p, 10p) demonstrate that groundwater drainage may result in considerable temporary elevations of water tables compared to undisturbed conditions in some local peatland areas.

287

−1.00

Depth (m)

−1.00 1/8−99

Depth (m)

10/8−99

20/8−99

−2.00

−3.00 30/8−99

0.00

0.00

−0.50

−1.00 8p 8m B2 19p

−1.00

−1.50 1/8−99

10/8−99

20/8−99

−2.00

−3.00 30/8−99

0.00

0.00

−0.50

−1.00 11p 11m B2 19p

−1.00

−1.50 1/8−99

Depth (m)

3p 3m B2 19p

−0.50

10/8−99

20/8−99

−2.00

−3.00 30/8−99

0.00

0.00

−0.50

−0.50 12p 12m B3 20p

−1.00

−1.50 1/7−00

10/7−00

20/7−00

−1.00

−1.50 30/7−00

Depth in bedrocks (m)

0.00

Depth in bedrocks (m)

0.00

Depth in bedrocks (m)

0.50

Depth in bedrocks (m)

Depth (m)

J. Kværner, P. Snilsberg / Journal of Hydrology 403 (2011) 278–291

Fig. 6. Depths to water tables in peatlands during events with periodical injection of water into the bedrock from the tunnel.

5.2. The influence of water balance, catchment water deliveries, and wetland position on groundwater tables The geohydrological main pattern in the Lake Northern Puttjern and Kjerringmyr areas, with groundwater water levels at a minimum during summer 1997, when the tunnel leakage was at maximum and water deliveries from catchments were low, and increasing water table levels at later episodic precipitation – runoff events, demonstrates how the balance between tunnel leakage and water supplies from surrounding catchments is decisive for groundwater drawdowns and extensions of the areas impacted by drainage. The fluctuations of groundwater levels in 1998 with drawdowns in dry periods and rises in wet periods demonstrate how the temporal fluctuation patterns of water tables in peatlands were determined by runoff events and water deliveries from surrounding catchments. The more frequent temporary rises of peat water levels after precipitation – runoff events in the peat wells close to the northern border of Lake Southern Puttjern compared to the wells nearest Lake Northern Puttjern reflect the influence of nearness to water deliveries from the uplands. The groundwater declines in dry periods stress the importance of the water deliveries from catchments for the consequences of water leakage on peatlands. Groundwater table falls of magnitude more than 1 m in peat wells during periods as short as two weeks, illustrate the importance of even short drought periods for impacts of groundwater leakage on peatlands.

Water tables in the peat wells between Lakes Northern and Southern Puttjern ranging between the elevations of the lake surfaces reflect larger drawdowns with decreasing distance to the tunnel trace, but may also illustrate the importance of lake water deliveries for aquifers and connected peatlands in catchment with small supplies of groundwater in dry periods. The drainable porosity of decomposed peat is small compared to lakes (Paavilainen and Päivänen, 1995). In areas with leakage to underground constructions and small supplies of groundwater in dry periods, e.g. small catchments with crystalline rocks and exposed bedrock and thin till deposits in uplands, the contribution of water from lakes may be an important source for recharge of groundwater in bedrock in drought periods and thus counteract water drawdowns in hydrologically connected peatlands. The significant fluctuations of the water tables in the wells located at the southern side of Lake Southern Puttjern in the first part of 1998, despite the high water table in the lake, and the considerable drawdowns of groundwater in the wells between Lake Southern Puttjern and Lake Northern Puttjern during summer 1999, however, demonstrate that even with high water tables in lakes, groundwater drawdowns might occur in adjacent peatlands. The limited drawdowns of peatland water tables between June 1998 and 1999 in the Kjerringmyr area close to the tunnel trace compared to the considerable water table drawdowns in the Lake Northern Puttjern area can be explained by the larger drawdown necessary to lower the potensiometric head below the altitude of

Meter above sea level

267 266 265 264 263 262 261 260 259 258 1/7−97

Meter above sea level

267 266 265 264 263 262 261 260 259 258 1/7−97

Meter above sea level

J. Kværner, P. Snilsberg / Journal of Hydrology 403 (2011) 278–291

267 266 265 264 263 262 261 260 259 258 1/7−97

Meter above sea level

288

267 266 265 264 263 262 261 260 259 258 1/7−97

14p 8p 11p 4p Lake Southern Puttjern Lake Northern Puttjern Thresholds lake outlets 1/1−98

1/7−98

1/1−99

1/7−99

1/1−00

1/7−00

14m 8m 11m 4m Lake Southern Puttjern Lake Northern Puttjern Thresholds lake outlets 1/1−98

1/7−98

1/1−99

1/7−99

1/1−00

1/7−00

12p 9p 10p Lake Southern Puttjern Lake Northern Puttjern Thresholds lake outlets 1/1−98

1/7−98

1/1−99

1/7−99

1/1−00

1/7−00

12m 9m 10m Lake Southern Puttjern Lake Northern Puttjern Thresholds lake outlets 1/1−98

1/7−98

1/1−99

1/7−99

1/1−00

1/7−00

Date Fig. 7. Timeseries of water elevation levels in wells and lakes in the Lake Northern Puttjern area. Lake water levels from Aars (1998) and The Norwegian Water and Energy Directorate (2009).

the terrain and the natural groundwater table at Kjerringmyr. This illustrates the importance of the altitude and position of wetlands within regional groundwater systems for vulnerability to disturbance impacts of groundwater lowering and leakage to underground constructions. 5.3. Leakage processes and hydrological connectivity between peatlands and bedrock The clear responses of the hydraulic heads in peat, subpeat tills, and bedrock after precipitation events indicate hydraulic connectivity between the terrain surface and the subsurface. The temporary dry outlet streams and periods with considerable water table drawdowns in Lakes Northern Puttjern and Southern Puttjern (Fig. 7), the disappearance of stream water north of Lake Southern Puttjern, and the emptying of temporary surface water storages in peatland depressions demonstrate infiltration of surface water in the peatsoils. Low water tables and decreasing hydraulic heads with increasing layer depths, i.e. in peat, subpeat till and bedrock, demonstrate downward flow of water from the peatlands around Lake Northern Puttjern and at Kjerringmyr through peat layers, subpeat till and bedrock fractures. The hydraulic connectivity is also illustrated by the widespread occurrence of drainage impacts on peatland surface in these areas, as surface subsidence (Kværner and Snilsberg, 2008). Since most of the monitoring wells were situated in the

vicinity of depressions in the peatland subsurface in assumed fracture zones in bedrock, the results demonstrate hydraulic connectivity in large parts of these zones. The periodically larger water table drawdowns in the wells 8m, 10m, 11m and 12m than in the wells 4m and 5m situated closer to the tunnel trace may reflect neighbouring important drainage pathways, but may for the wells 8m, 10m, 11m partly also be an effect of large filter depths. The low water tables in the peat wells in the Lake Northern Puttjern area and at the mire Kjerringmyr reflect drainage of water from peat layers to subpeat layers. Low water tables in the peat wells often closely follow water table patterns in adjacent subpeat wells, as is indicated by the high correlation coefficients, revealing hydraulic connectivity and vertical flow through the deeper peat layers in most of these peatlands. That water tables in the peat wells 3p, 8p, 11p and 12p did not follow the oscillations in groundwater in subpeat sediments and bedrock in periods with injection of water into bedrock from the tunnel may demonstrate dampening of the oscillations in the peat layers or a less direct connection to the water bearing fractures in bedrock than from the subpeat till. Two different conceptualisations of the groundwater hydrology in peatlands currently exist (Reeve et al., 2000). Whereas the shallow flow model assumes that the deeper, more decomposed peat (catotelm) is relatively impermeable, isolating the uppermost peat (acrotelm) from the mineral soil or bedrock, the groundwater flow model assumes that the extent of vertical groundwater flow in

289

Meter above sea level

268 267 266 265 264 263 262 261 260 259 258

Meter above sea level

268 267 266 265 264 263 262 261 260 259 258

Meter above sea level

J. Kværner, P. Snilsberg / Journal of Hydrology 403 (2011) 278–291

268 267 266 265 264 263 262 261 260 259 258

(a)

13 4

6

7

8

11 Lake S. Puttjern

Lake N. Puttjern

0

100

200

300

400

500

14

13 4

6

7

8

600

11

700

(b)

Lake S. Puttjern

Lake N. Puttjern

0

100

200

Well Water table Lake surface Peat bottom

300

400

500

14

13 4

6

7

8

600

11

700

(c)

Lake S. Puttjern

Lake N. Puttjern

0

100

200

300

400

500

600

700

Distance (m) Fig. 8. Longitudinal sections of the peatlands between Lakes Northern and Southern Puttjern with wells and groundwater levels. (a) August 13 1997, (b) Mars 26 1998, (c) June 11 1998. Lake water levels from Aars (1998) and The Norwegian Water and Energy Directorate (2009).

peatlands is primarily not controlled by the permeability of humified peat layers, but by the mineral soil permeability. In the peatlands above the tunnel the infiltration of surface water and the large and often parallel lowering of water tables in dry periods in the wells in peatsoils and subpeat mineral soils indicate that the catotelm did not prevent vertical groundwater flow and show that the groundwater flow model hypothesis is the most appropriate for these peatlands. Macroporosity has been observed in peatlands in several recent studies (Holden, 2006) and has also been found to increase following drought (Holden and Burt, 2002). Thus, the water table responses in the peatlands and the vertical flow even through highly humified peat layers in the catotelm may be explained by macropore flow. The slower water drawdowns in peat wells than in adjacent wells in mineral soils at some occasions, e. g. in the wells 8p and 8m in summer 1998, indicate that at least in parts of peatlands the catotelm in some periods may reduce the flow of water from the acrotelm. According to Olofsson (1993) groundwater flow from soil to rock only occurs at specific sites where a combination of favourable geological conditions exists; infiltration requires a permeable soil, or permeable horizons or macropores in a low permeability soil, as well as conductive fractures in the bedrock. Olofsson (1993) points out that in areas with unsorted soils, minor depressions in the bedrock surface usually formed around fractures in the bedrock are of particular importance for water flow from soil to rock. The deep and simultaneous lowering and fluctuations of hydraulic heads in the wells in the subpeat sediments in the Lake Northern Puttjern area demonstrate that hydraulic connectivity

between the subpeat soils and the water bearing fractures in the bedrock was not restricted to a few specific sites. The hydraulic connectivity is also illustrated by the water tables in the subpeat wells 3m, 8m, 11m and 12m fluctuating in time with the injections of water into bedrock from the tunnel. The results from the Lake Puttjern area indicate that in peatlands with a subsurface consisting of a relatively thin till cover above crystalline fractured rocks, favourable conditions for groundwater flow from soil to rock are not always restricted to a few specific sites. Reeve et al. (2000) points out that before a conceptual model of peatland hydrology can be applied to a system, it is essential to understand the hydrogeological setting. Several previous studies have shown the importance of hydrogeology of mineral soils for peatland hydrology (Devito et al., 1996; Reeve et al., 2000; Vidon and Hill, 2004). This study illustrates that for peatlands with a subsurface of a thin till cover above fractured rocks peat water– bedrock water connectivity should be considerate when conceptualizing peatland hydrology and assessing hydrological impacts of groundwater exploitation, drought periods and climate changes. The results may also indicate that water–bedrock water connectivity should receive attention when assessing impacts of drought periods and climate changes on release of greenhouse gases and ecology on such areas. 6. Conclusions Groundwater leakage to underground constructions in fractured crystalline bedrock may significantly alter the hydrology of

290

J. Kværner, P. Snilsberg / Journal of Hydrology 403 (2011) 278–291

Depth (m)

2.00 0.00 −2.00 −4.00

2p 2m B1

0.00 −20.00 1/1−98

1/7−98

1/1−99

1/7−99

1/1−00

Depth (m)

2.00 0.00 −2.00 −4.00

11p 11m B2

−6.00 −8.00 1/1−98

1/7−98

1/1−99

1/7−99

1/1−00

Depth (m)

2.00 0.00 −2.00 −4.00

12p 12m B3

−6.00 −8.00 1/1−98

1/7−98

1/1−99

1/7−99

1/1−00

Depth (m)

2.00 0.00 −2.00 −4.00

19p 19d B4

−6.00 −8.00 1/1−98

1/7−98

1/1−99

1/7−99

1/1−00

Fig. 9. Water levels in adjacent wells in peatlands and bedrock.

overlying wetlands. In peatlands in catchments with surficial covers of thin till deposits and peatsoils above fractured gneisses, tunnel leakage can cause substantial drawdowns and fluctuations of the groundwater table, increase the recharge areas and reduce the runoff. The drainage effects on peatlands of groundwater leakage to underground constructions differ from surficial drainage. The deep drainage base can provide prolonged and deep water table drawdowns in dry periods, and differential settlements in drained peatsoils may result in secondary changes in patterns of surface water storage and flow. The impacts of tunnel leakage on peatland hydrology are influenced by the balance between tunnel leakage and water supplies from surrounding catchments, and the wetland altitude and position within the groundwater flow system. In areas with leakage to underground constructions, the pattern of water deliveries from surrounding catchments is important for the temporal fluctuations of groundwater in peatlands. Even drought periods as short as two weeks may result in substantial groundwater lowering. Deep and simultaneous lowering and fluctuations of hydraulic heads in wells in the peat, the subpeat sediments and the bedrock above the tunnel demonstrated hydraulic connectivity between the peat layers and the bedrock, and revealed vertical flow even through highly humified peat layers in the catotelm. The study shows that in peatlands with a subsurface of a relatively thin till cover above fractured crystalline bedrock, favourable conditions for groundwater flow from peatsoils to rock are not always restricted to a few specific sites, and indicates that attention should be given to the influence of peat water–bedrock water connectivity

on impacts of groundwater exploitation, droughts and climate changes in such areas.

Acknowledgements This work has been financially supported by NSB Gardermobanen AS and Bioforsk – Norwegian Institute for Agricultural and Environmental Research, Soil and Environment Division. Furthermore, the authors wish to thank Øistein Johansen and Tom Helgesen for their work in the field.

References Aars, Ø., 1998. The Romerike gate. A railway turning into a local watergate. In: Kajander, J. (Ed.), XX Nordic Hydrological Conference, Nordic Associations for Hydrology, Helsinki, Finland, 10–13 August 1998. Nordic Hydrological Programme. NHP Report No. 44, pp. 234–241. Bayley, S.E., Behr, R.S., Kelly, C.A., 1986. Retention and release of S from a freshwater wetland. Water, Air & Soil Pollution 31, 101–114. Beldring, S., Roald, L.A., Voksø, A., 2002. Runoff map for Norway. Average annual runoff 1961–1990. The Norwegian Water Resources and Energy Directorate. Dokument 2 2002, 49 pp (in Norwegian). Bendiksen, E., Bakkestuen, V., Erikstad, L., Stabbetorp, O.E., Eilertsen, O., Wilman, B., 2005. Miljøundersøkelser i Østmarka. Overvåking av vegetasjonen i influensområdet til Romeriksporten 1998–2003. (Environmental investigations in Østmarka. Monitoring of vegetation in the influence area of Romeriksporten 1998–2003), 43 pp. Berthelsen, A., Olerud, S., Sigmond, E.M.O., 1996. Geologic map over Norway, Bedrock map Oslo 1:250 000, Geological Survey of Norway. Brække, F.H., 1981. Hydrochemistry in low-pH soils of south Norway. 1. Peat and soil water quality. Reports of the Norwegian Forest Research Institute 36, 1–32.

J. Kværner, P. Snilsberg / Journal of Hydrology 403 (2011) 278–291 Bragazza, L., Gerdol, R., 1996. Response surfaces of plant species along water-table depth and pH-gradients in poor mire on the southern Alps (Italy). Annales Botanici Fennici 33, 11–20. Branfireun, B.A., Heyes, A., Roulet, N.T., 1996. The hydrology and methylmercury dynamics of a Precambrian Shield headwater peatland. Water Resources Research 32, 1785–1794. Cesano, D., Olofsson, B., Bagtzoglou, A., 2000. Parameters regulating groundwater inflows into hard rock tunnels – a statistical study of the Bolmen tunnel in southern Sweden. Tunnelling and Underground Space Technology 15 (2), 153– 165. Devito, K.J., 1995. Sulphate mass balances of the Precambrian Shield wetlands: the influence of catchment hydrogeology. Canadian Journal of Fisheries and Aquatic Science 52, 1750–1760. Devito, K.J., Dillon, P., 1993. The influence of hydrologic condition and peat oxia on the phosphorous and nitrogen dynamics of a conifer swamp. Water Resources Research 29, 2675–2685. Devito, K.J., Hill, A.R., 1997. Sulphate dynamics in relation to groundwater–surface water interactions in headwater wetlands of the southern Canadian Shield. Hydrological Processes 11, 485–500. Devito, K.J., Hill, A.R., Roulet, N., 1996. Groundwater–surface interactions in headwater forested wetlands of the Canadian Shield. Journal of Hydrology 181, 127–147. Domenico, P.A., Schwartz, F.W., 1997. Physical and Chemical Hydrogeology, second ed. John Wiley & Sons Inc, 506 pp. Environment and Heritage Service Belfast, 2010. . European Commission, 2000. EU Water Framework Directive (Directive 2000/60/ EC). European Parliament and Commission, Official Journal (OJ L 327) on 22 December. Ferone, J.M., Devito, K.J., 2004. Shallow groundwater–surface water interactions in pond–peatland complexes along a Boreal Plains topographic gradient. Journal of Hydrology 292, 75–95. Færgestad, A., 1999. Personal communication. Glaser, P.H., Siegel, D.I., Romanowicz, E.A., Shen, Y.P., 1997. Regional linkages between raised bogs and the climate, groundwater, and landscape features of northwestern Minnesota. Journal of Ecology 85, 3–16. Glaser, P.H., Chanton, J.P., Morin, P., Rosenberry, D.O., Siegel, D.I., Ruud, O., Chasar, L.I., Reeve, A.S., 2004. Surface deformations as indicators of deep ebullition fluxes in large northern peatland. Global Biogeochemical Cycles 18, GB1003. doi:10.1029/2003GB002069. Glaser, P.H., Siegel, D.I., Reeve, A.S., Chanton, J.P., 2006. In: Martini, I.P., Martínez Cortizas, A., Chesworth, W. (Eds.), Hydrogeology of Major Peat Basins in North America. Evolution and Records of Environmental and Climate Changes, Peatlands, pp. 347–376. Graversen, O., 1984. Geology and structural evolution of the Precambrian rocks of the Oslofjorden–Øyeren area, southeast Norway. Geological Survey of Norway Bulletin 398, 1–50. Heimly, P., 1994. Ingeniørgeologisk kart, Tunnel Etterstad Stalsberg, NSB Gardermobanen A/S 10.02.1994. Hillman, G.R., 1992. Some hydrological effects of peatland drainage in Alberta’s boreal forest. Canadian Journal of Forest Research 22 (1), 1586–1596. Holden, J., 2006. Hydrology. In: Martini, I.P., Martinez Coritizas, A., Chesworth, W. (Eds.), Evolution and Records of Environmental and Climate Changes. Developments in Earth Surface Processes, vol. 9. Peatlands, pp. 319–346. Holden, J., Burt, T.P., 2002. Laboratory experiments on drought and runoff in blanket peat. Soil Science 53, 1–15. Holmøy, K.H., 2008. Significance of geological parameters for predicting water leakage in hard rock tunnels. Doctoral Thesis at NTNU 2008:29. Norwegian University of Science and Technology, 224 pp. Ingram, H.A.P., 1978. Soil layers in mires: function and terminology. European Journal of Soil Science 29, 224–227. Ingram, H.A.P., 1983. Hydrology. In: Gore, A.J.P. (Ed.), Ecosystems of the World, vol. 4A. Elsevier, Amsterdam, pp. 67–158. Ishizaki, A., 1979. Examples and research aspect for dearth of water caused by railway tunnelling in Japan. In: Kitamura, I. (Ed.), Tunneling under Difficult Conditions. Proceedings of the International Tunnel Symposium, 1978. Pergamon Press, Tokyo, pp. 241–246. Joosten, H., 2004. The IMCG Global Peatland Database. www.imcg.net/gpd/gpd.htm. Kim, Y.-Y., Lee, K.-K., 2003. Disturbance of groundwater table by subway construction in the Seoul area, Korea. Geosciences Journal 7 (1), 37–46. Kitterød, N.-O., Colleuille, H., Wong, W.K., Pedersen, T.S., 2000. Simulation of groundwater drainage into tunnel construction and numerical analysis of leakage remediation. Hydrogeology Journal 8, 480–493. Korpela, L., 1998. Effect of drainage on the diversity of boreal forested mire margin vegetation in Finland. In: Sapo, R. (Ed.), Proceedings of the International Peat Symposium the Spirit of Peatlands. International Peat Society, pp. 145–147. Kværner, J., Kløve, B., 2006. Tracing sources of summer streamflow in boreal headwaters using isotopic signatures and water geochemical components. Journal of Hydrology 331, 186–204. Kværner, J., Kløve, B., 2008. Generation and regulation of summer runoff in a boreal flat fen. Journal of Hydrology 360, 15–30.

291

Kværner, J., Snilsberg, P., 2008. The romeriksporten railway tunnel – Drainage effects on peatlands in the Lake Northern Puttjern area. Engineering Geology 101, 75–88. Longva, O., 1991. FET 1814 I, Quaternary geologic map 1:50 0000. Norwegian Geological Survey, Trondheim. Mabee, S.B., Curry, P.J., Hardcastle, K.C., 2002. Correlation of lineaments to ground water inflows in a bedrock tunnel. Ground Water 40 (1), 37–43. Malmer, N., 1962. Studies on mire vegetation in the Archaean area of southwestern Götaland (south Sweden). I. Vegetation and habitat conditions on the Akhult mire. Opera Botanica 7 (1), 1–322. NSB Gardermobanen AS, 1998a. Personal communication. NSB Gardermobanen AS, 1998b. Romeriksporten. Miljøovervåking og tiltak i Østmarka. Status oktober 1998 (Environmental Monitoring and Mitigation Efforts in Østmarka. Status October 1998), Oslo. NSB Gardermobanen AS, 1999a. Romeriksporten. Miljøovervåking og tiltak i Østmarka oktober 1998–juni 1999 (Environmental Monitoring and Mitigation Efforts in Østmarka October 1998–June 1999), Oslo. NSB Gardermobanen AS, 1999b. Miljøvervåking og tiltak i Østmarka, juli–desember 1999 (Environmental Monitoring and Mitigation Efforts in Østmarka, July– December 1999), Oslo. Økland, R.H., 1989. A phytoecological study of the mire Northern Kisselbergmosen, SE Norway. I. Introduction, flora, vegetation and ecological conditions. Sommerfeltia 8, 1–172. Økland, R.H., Økland, T., Rydgren, K., 2001. Vegetation-environment relationships of boreal spruce swamp forests in Østmarka Nature Reserve, SE Norway. Sommerfeltia 29, 1–190. Olofsson, B., 1991. Impact on groundwater conditions by tunnelling in hard crystalline rocks. Dr. Thesis. KTH Royal Institute of Technology, Stockholm. Olofsson, B., 1993. Flow of groundwater from soil to crystalline rock – a review. In: Banks, S., Banks, D. (Eds.), Hydrogeology of Hard Rock. Memoires of the XXIVth Congress International Association of Hydrogeologists 28th June 2nd July 1993, Ås (Oslo), Norway, pp. 915–931. Oppsal IF, 1990. Puttåsen, Oppsal’s orienteering map nr. 32. Oppsal IF, 1995. Lutvann, Oppsal’s orienteering map nr. 25. Paavilainen, E., Päivänen, J., 1995. Peatland Forestry. Ecology and Principles, Springer Verlag, 248 pp. Praamsma, T., Novakomski, K., Kyser, K., Hall, K., 2009. Using stable isotopes and hydraulic head data to investigate groundwater recharge and discharge in a fractured rock aquifer. Journal of Hydrology 366, 35–45. Price, J.S., Branfireun, B.A., Waddington, J.M., Devito, K.J., 2005. Advances in Canadian wetland hydrology 1999–2003. Hydrological Processes 19 (1), 201– 214. Reeve, A.S., Siegel, D.I., Glaser, P.H., 2000. Simulating vertical flow in large peatlands. Journal of Hydrology 227, 207–217. Reinikainen, A., 1984. Ecological variation of mire site types in the small kettle-hole mire Heinisuo, Southern Finland. Annales Botanici Fennici 21 (1), 79–101. Rohde, A., Bockgård, N., 2006. Groundwater recharge in a hard rock aquifer: a conceptual model including surface-loading effects. Journal of Hydrology 330, 389–401. Rydin, H., Jeglum, J., 2006. The Biology of Peatlands. Oxford University Press, 300 pp. Siegel, D.I., Glaser, P., 2006. The hydrology of peatlands. In: Wieder, R.K., Vitt, D.H. (Eds.), Boreal Peatland Ecosystems. Ecological Studies, vol. 188, pp. 189–311. Skjeseth, S., 1982. Vassdragregulering, konsekvenser for grunnvann og miljø. Eksempler fra Tokke-utbyggingen (Hydropower regulation, impacts on groundwater and environment. Examples from the Tokke project). In: Helzen, A.M., Nilsen B., Mowinckel, A. (Eds.), Fjellsprengningsteknikk Bergmekanikk Geoteknikk, 1981. Tapir, 8 pp. Statens Offentliga Utredningar, 1998. Miljö i grund och botten – erfarenheter från Hallandsåsen. Slutrapport av Tunnelkommissionen. SOU 1998:137, Stockholm, 124 pp. Strack, M., Waddington, J.M., 2007. Response of peatland carbon dioxide and methane fluxes to a water table drawdown experiment. Global Biogeochemical Cycles 21, GB1007. doi:10.1029/2006GB002715. Strack, M., Waddington, J.M., Tuittila, E.-S., 2004. The effect of water table drawdown on northern peatland methane dynamics: implications for climate change. Global Biogeochemical Cycles 18, GB4003. doi:10.1029/2003GB002209. The Norwegian Meteorological Institute, 2009. . The Norwegian Water Resources and Energy Directorate, 2009. Data from the Hydra database. van Dam, H., 1986. Acidification of three moorland pools in the Netherlands by acid precipitation and extreme drought periods over seven decades. Freshwater Biology 20, 157–176. Vidon, P.G.F., Hill, A.R., 2004. Landscape controls on the hydrology of stream riparian zones. Journal of Hydrology 292, 210–228. Vincenzi, V., Gargini, A., Goldscheider, N., 2009. Using tracer tests and hydrological observations to evaluate effects of tunnel drainage on groundwater and surface waters in the Northern Apennines (Italy). Hydrogeology Journal 17, 135–150. von Post, L., 1922. Sveriges geologiska undersøknings torvinventering og nogra av dess hittils vunna resultat. Svenska Mosskulturföreningens Tidskrift 1, 1–27. Wieder, R.K., 1985. Peat and water chemistry at Big Run Bog, a peatland in the Appalachian mountains of West Virginia, USA. Biogeochemistry 1, 277–302.