GEOLOGICAL SURVEY OF CANADA
OPEN FILE 5624
The Vars–Winchester esker aquifer, South Nation River watershed, Ontario: CANQUA Fieldtrip, June 6, 2007
Cummings, D. I., Russell, H. A. J.
2007 GEOLOGICAL SURVEY OF CANADA
OPEN FILE 5624
The Vars–Winchester esker aquifer, South Nation River watershed, Ontario: CANQUA Fieldtrip, June 6, 2007
Compiled by
Don I. Cummings and Hazen A. J. Russell
With contributions from
Sam Alpay, Anne-Marie Chapman, Coralie Charland, George Gorrell, Marc J. Hinton, Tessa Di Iorio, André Pugin, Susan Pullan, and David R. Sharpe
2007
©Her Majesty the Queen in Right of Canada 2007 Available from Geological Survey of Canada 601 Booth Street Ottawa, Ontario K1A 0E8
Cummings, D. I. and Russell, H. A. J. 2007: The Vars–Winchester esker aquifer, South Nation River watershed, Ontario, CANQUA Fieldtrip Guidebook, June 6th 2007; Geological Survey of Canada, Open File 5624, 68 p.
Open files are products that have not gone through the GSC formal publication process.
Figure 1. Field-trip route and key geographic features. Mapped location of eskers includes known (exposed) and interpreted (buried) sections (from Gorrell, 1991). iii
Contributors
Sam Alpay Tessa Di Iorio Geological Survey of Canada South Nation Conservation Authority Natural Resources Canada P. O. Box 69 601 Booth Street, Ottawa, ON 15 Union Street Canada K1A 0E8 Berwick, ON K0C 1G0 Telephone: (613) 996-2373 Telephone: 613-984-2948 ext. 295 E-mail: [email protected] Toll Free: 877-984-2948 [email protected]
Anne-Marie Chapman André Pugin South Nation Conservation Authority Geological Survey of Canada P. O. Box 69 Natural Resources Canada 15 Union Street 601 Booth Street, Ottawa, ON Berwick, ON K0C 1G0 Canada K1A 0E8 Telephone: 613-984-2948 ext. 250 Telephone: (613) 943-6513 Toll Free: 877-984-2948 E-mail: [email protected] E-mail: [email protected]
Coralie Charland Susan E. Pullan Department of Earth Sciences Geological Survey of Canada University of Ottawa, Natural Resources Canada Ottawa, Ontario, 601 Booth Street, Ottawa, ON Canada, K1N 6N5 Canada K1A 0E8 E-mail: [email protected] Telephone: (613) 992-3483 E-mail: [email protected]
Don I. Cummings Hazen A.J. Russell Geological Survey of Canada Geological Survey of Canada Natural Resources Canada Natural Resources Canada 601 Booth Street, Ottawa, ON 601 Booth Street, Ottawa, ON Canada K1A 0E8 Canada K1A 0E8 Telephone: (613) 992- Telephone: (613) 992-3059 E-mail : [email protected] E-mail : [email protected]
George Gorrell David R. Sharpe Gorrell Resource Investigations Geological Survey of Canada RRR 1 Oxford Mills, Natural Resources Canada Ontario K0G, 1S0 601 Booth Street, Ottawa, ON E-mail: [email protected] Canada K1A 0E8 Telephone: (613) 992-3059 E-mail: [email protected]
Marc J. Hinton Geological Survey of Canada Natural Resources Canada 601 Booth Street, Ottawa, ON Canada K1A 0E8 Telephone: (613) 992-3059 E-mail: [email protected] iv
Table of Contents
Contributors...... iii
Table of Contents ...... iv
Acknowledgements ...... v
Field Trip Itinerary ...... vi
1. 0 Introduction...... 1
1.1 Trip objectives and outline ...... 2
2.0 South Nation Source Water Protection Region...... 2
2.1 Source water protection In Ontario ...... 2
2.2 Land use in South Nation...... 3
2.3 Water management in South Nation...... 4
3.0 Environmental Setting...... 6
3.1 Temperature and precipitation ...... 6
3.2 Evapotranspiration ...... 6
3.3 Hydrology ...... 7
3.4 Hydrogeology ...... 8
3.5 Inferences from regional groundwater geochemistry...... 9
4.0 Geological Setting ...... 11
4.1 Landscape and bedrock geology ...... 11
4.2 Stratigraphy...... 14
5.0 Hydrogeological implications ...... 38
6.0 Fieldtrip Stops...... 41
Stop 1. Seismic section French Hill Road East...... 34
Stop 2. Regimbald Road pit ...... 36
Stop 3. Watson Road pit ...... 38
Stop 4. Route 300 seismic profiles and cores...... 41
Stop 5 Kemptville/Loughlin Ridge esker – subaqueous fan ...... 46 v
7.0 References ...... 58
8.0 Appendix: Methods ...... 65
A.1 Basin analysis approach...... 65
A.2. Database development...... 65
A.3 Outcrop and core studies...... 66
A.4 Shallow seismic reflection methods...... 66
A.5 Simple, inexpensive methods for hydraulic head measurements and groundwater flux estimations ...... 72
Acknowledgements
The work presented in this guidebook was made possible thanks to the generous support of the South Nation and Raisin river conservation authorities, who funded the project using source-water protection money from the Ministry of Environment and the Ontario Geological Survey. Funds from the Earth Science Sector Groundwater Program of Natural Resources Canada allowed additional data to be collected. Data support by the Ontario Ministry of Natural Resources and the Ottawa–Carleton Regional Municipality is acknowledged, as is the permission of Tackaberry Sand and Gravel and Laurent Leblanc Inc. for access to their pits for field-trip stops. Seismic data was collected by R. Good, R. Burns, T. Patterson and M. Douma (Geological Survey of Canada). Assistance by R Lacroix on a number of graphics is greatly appreciated. The authors appreciate the GIS and database assistance of C. Logan and L Robertson. D. Ponomarenko calculated paleodischarge for the Ottawa River incised valley and analyzed air photos. André Martel of the Canadian Museum of Nature graciously identified several shell fragments collected from the esker. A critical review by R. Knight helped clarify and streamline the document. This publication is a contribution of the Geological Survey of Canada Groundwater Program and is part of a project entitled “Groundwater Inventory: Aquifer Systems in Canada Project Activities in Southern Ontario”. vi
Field Trip Itinerary
8:15 Carleton University
Proceed via Hwy 417 east toward Orleans changing to Route 17. Travel along the Ottawa River shoreline to Cumberland. Proceed south on rue Cameron, Market and Dunning. Proceed south on Dunning for ~5 km. Turn east on French Hill Road. Stop on flat field just east of 3233 French Hill Road. (~ 40 km)
Stop 1: Field trip overview, then discussion of French Hill Road buried valley with Seismic Landstreamer Demonstration (tentative)
From Stop 1, turn back and go west on French Hill Road. Turn left (south) on Sarsfield Road. Turn left on Regimbald Road. Continue until pit gate. Either stop and walk, or open gate and drive to pit. (< 5 km from Stop 1)
Stop 2: Rebimbald Road pit – esker gravel
From Stop 2, return to Sarsfield Road and proceed south for ~5 km, passing through the town of Sarsfield. Turn right on Watson Road. The Watson Road pit is located on a hill (the esker) on the north side of the road, approximately 5 km from the Watson Rd–Sarsfield Rd intersection.
Stop 3: Watson Road pit – esker gravel and sand, plus Champlain Sea mud
LUNCH
From Stop 3, proceed west on Watson Road. Turn left onto Dunning Road. Continue south until end of Dunning, then turn right (west) on Russland Road. Turn left (south) on Rockdale Road. Cross Highway 417 and continue south on St Guillaume Road. to Route 300 (~ 7 km south of Hwy 417). Turn east on Route 300 for 0.5 km. Stop just east of creek. (21 km from Stop 3)
Stop 4: Route 300 seismic section and core
Return to St Guillaume Road and head south to Embrun and Castor Road. Turn east on Route 3 and for ~200 m to St Andre Road. Proceed south on St Andre Road to Marionville Road (~7 km). Turn west on Marionville Road and follow for 2.75 km to Russell Road. Turn south on Russell Road to Morewood and continue toward Chesterville. Turn west on Route 43 toward Winchester and onward toward Kemptville. At the intersection of Highway 43 and Reids-Mills Road turn north on Reids-Mills Road (right) for 1.6 km to Loughlin Ridge Road. Turn west on Louglin Ridge Road for < 2 km. (~40 km from Stop 4)
Stop 5: Kemptville Tackaberry pit - subaqueous fan sedimentology
Return to Ottawa: Return to Route 43 and follow west to Highway 416 (~13 km). Proceed north on Highway 416 for Ottawa and Carleton University (~ 57 km) 1. 0 Introduction
Groundwater is the source of potable water for >65% of the 250,000 residents in the South Nation River and Raison River watersheds. Of particular importance are two subparallel eskers, the Vars– Winchester and Crysler–Finch eskers, which provide >80% of the groundwater used by municipal water systems (Figs 1, 2). The esker aquifers are prolific, with municipal wells yielding > 500 US gallons per minute (>31 L/s) compared to fractured-bedrock wells that typically yield < 10 US gallons per minute (<0.6 L/s) (Charron, 1978). The development of hydrogeological models to estimate groundwater flow in the esker aquifers has been hampered by a lack of knowledge of the size, shape and internal heterogeneity of the eskers. This presents significant constraints to advancing source water protection plans under the Ontario Clean Water Act (2006).
Water-supply management in Ontario has been significantly refocused through the Clean Water Act (2006). A notable part of this act is the “source-water protection” element that mandates municipalities to develop science based approaches to protect surface water and groundwater from which municipal drinking water is extracted. In the South Nation River watershed, the Vars–Winchester esker aquifer has been targeted as the first in a series of groundwater-focused source-water protection studies; the Crysler–Finch esker aquifer will be studied subsequently in 2007–2008. To facilitate this work, a collaborative partnership between the South Nation Conservation Authority, the Geological Survey of Canada (GSC), and the University of Ottawa has been developed. The GSC is delineating and characterizing the eskers using a basin-analysis approach that employs shallow reflection seismic geophysics, boreholes with continuous core collection and outcrop sedimentology. The University of Ottawa is concentrating on shallow hydrogeological measurements and numeric flow modeling of the esker aquifers.
Figure 2. A) Location map of the South Nation River watershed, Eastern Ontario. B) Simplified geology of the northern portion of the South Nation River watershed (modified from OGS). Esker locations from Gorrell (1991). C) General location of the Vars–Winchester esker study area. The Vars–Winchester esker is indicated by i, and Crysler–Finch esker indicated by ii. 2
Exploitation of eskers for groundwater is relatively common in Ontario, and in glaciated terrain in general (e.g. Caswell, 1988a; Artimo et al. 2003; Bolduc et al. 2006; Riverin, 2006). Despite this, most studies have concentrated on understanding eskers from a sedimentological perspective related to ice sheet processes (e.g. Brennand, 2000), aggregate resources (Gorrell, 1991; Spooner and Dalrymple, 1994) and mineral exploration (Levasseur and Prichonnet, 1995), with comparatively little effort invested in understanding eskers as aquifers. Partially buried or buried eskers in glacial basins can form prolific aquifers (Caswell, 1989) that present numerous challenges to groundwater studies due to the local variability of coarse sediment, elusive high hydraulic-conductivity pathways, narrow width, and long longitudinal extent. The basin analysis approach employed by the GSC to study the Vars– Winchester esker (Appendix A-1) is ideally suited to address these challenges, and will provide a scientific basis for improved source water protection planning.
1.1 Trip objectives and outline
The field trip focuses on the GSC contribution to the Vars–Winchester esker aquifer study. Particular objectives are:
x to observe the esker and interpret how it formed
x to observe the overlying Champlain Sea mud and interpret how it was deposited
x to consider how the stratigraphy affects groundwater flow
x to illustrate how detailed geological studies lead to effective source-water protection strategies
2.0 South Nation Source Water Protection Region Tessa Di Iorio and Anne-Marie Chapman, South Nation Conservation Authority
2.1 Source water protection in Ontario
The contamination of a municipal groundwater supply for the town of Walkerton in 2000 resulted in a major overhaul of drinking-water legislation and practices in Ontario. Following the Walkerton tragedy, where seven people died because of poor water management and testing practices, the provincial government commissioned the Walkerton Inquiry, which tabled two reports in 2002. Part of the government’s commitment to implement the recommendations of the Walkerton Inquiry is the development of the Clean Water Act (Bill 43), which received royal assent on October 19, 2006.
Under the Clean Water Act, local communities are expected to define existing and potential threats to their water and set out and implement the actions necessary to reduce or eliminate significant threats. Communities are empowered to take action to prevent threats from becoming significant, and to engage public participation in the source water protection plan. Source-water protection plans are implemented at a watershed scale, with collaboration between community groups, municipalities, conservation authorities, and provincial government agencies (http://www.ene.gov.on. ca/ en/water/clean_water_act/index.php).
In order to facilitate continuity and coordination of source water protection, the 36 conservation authorities across Ontario have been grouped into 19 watershed regions (Fig. 3). The Raisin – South Nation Source Water Protection Region (SWPR) consists of watersheds managed by the Raisin Region Conservation Authority and South Nation Conservation Authority. 3
Figure 3. Source water protection areas in Southern Ontario (Ontario Ministry of Environment; modified from http://www.ene.gov.on.ca/envision/water/cwadocs/5989e.pdf)
2.2 Land use in South Nation
The geology of a watershed and the nature of land use influence the approach needed for source water protection. For example, are the aquifers in the watershed vulnerable to disruption of recharge, to contamination, or to over-exploitation? This section briefly discusses land use and water management practices in the SWPR.
The region has a population density that ranges from 20 to 1,107 people per square kilometre, with an average of 39.4 people/km2. This predominantly rural population is dispersed amongst 126 settlements of which only 43 have full urban services (i.e. municipal sanitary sewer and water). The population is growing rapidly, with a 21% increase between 1981 and 2001.
The SWPR has a rural based economy. Agriculture is the main land use (54% of the area). The trend is toward fewer but larger farms. In the past 20 years the number of farms has decreased by 15%; however, the amount of arable land has increased by 4%. Soil types range from light, acid sand to clay and clay loam, the later of which are able to sustain high intensity agriculture. Low topographic gradients and generally impermeable soil (40% of area) require extensive artificial drainage (80% of cropland). The dominant crops are corn, hay, soybean and alfalfa. The dominant livestock operations are dairy, beef, poultry, pig, and sheep farms (Fig 4). Resource-based economic activities, such as licensed pits and quarries, are also widespread. 4
The SWPR has a mixed second-growth forest (34% of area) that has remained relatively unchanged over the past twenty years. In contrast to the stability of forested area, wetlands have been reduced from an average extent of 25% in the 1800’s to < 4%, and are seriously threatened.
Figure 4. A) General landuse in the SWPR, B) Principal crops grown in the SWPR, C) Livestock farms as a percentage of total livestock farms, D) Animals on livestock farms as percentage of total animals.
Figure 5. Distribution of the permits to take water (PTTW) by (A) source of water, (B) volume of water takings, and (C) by water use per sector for the SWPR.
2.3 Water management in South Nation
Water use in Ontario is managed by the Ontario Ministry of Environment (MOE). The MOE requires water takings of > 50,000 liters per day to be permitted and entered in the Permit To Take Water (PTTW) database. It has been estimated that the PTTW database only reflects 20 to 25% of actual withdrawals. For example, residential and agricultural use is rarely recorded in the database. Consequently, any estimate of water takings in the area is likely significantly compromised by an inadequate database.
In the SWPR there are 205 permits for water extraction, for a total of 2.3 billion m3/yr (Fig. 5). The number of permitted takings is divided nearly equally between surface and groundwater withdrawals; however, surface water accounts for 95.5% of the water volume withdrawn. Sixty percent of this surface water, or 1.4 billion m3/yr, is permitted to one source from the St. Lawrence River, and is used for industrial cooling. Water use per sector is divided between industrial, municipal, commercial, and agricultural (Fig. 5). Industry is the dominant user of both surface and ground water. For example, 95% of permitted groundwater extraction is for aggregate washing. Municipalities draw water from both rivers (88%) and groundwater aquifers (12%). There are 13 municipal ground-water systems with twenty-one (21) active wells, which extract a total of 9 million m3/yr, or <2% of the total permitted takings. The Vars–Winchester esker aquifer supplies water to 5 municipal systems (14 active wells) that withdraw 78% of the groundwater used by municipalities in the SWPR. The Crysler–Finch esker aquifer supplies an additional 2 municipal systems that account for 7% of municipal groundwater extraction. Sixty-five percent of the population in the SWPR, or ~150,000 residents, rely on private groundwater supply from ~30,000 wells. Estimated residential groundwater use would add an additional 59% to the permitted groundwater withdrawals. 5
2.4 Importance of the esker study in South Nation
Within the SWPR, 43 settlements rely on municipal water supplies. Fifteen of these use groundwater sources, of which 9 draw water from esker aquifers. The Vars–Winchester esker provides groundwater to 7 communities (total population ~15,250) and the Crysler–Finch esker supplies water to 2 communities (total population ~1,050). A better understanding of the size and extent of the eskers and occurrence and movement of groundwater within the eskers is required to assist in the protection of groundwater reserves within the esker complexes. The sustainability and vulnerability of this valuable and sensitive groundwater resource is a primary concern for local communities and SWP initiatives. 6
3.0 Environmental Setting Tessa Di Iorio, South Nation Conservation Authority
This section provides information on the climate, basin hydrology, and the groundwater system in the vicinity of the Vars–Winchester esker aquifer. Currently, there is an inadequate understanding of key elements of the hydrological system that pertain to esker-aquifer recharge, groundwater flow paths, and sustainable extraction rates from the aquifers. Results from this aspect of the study were not available at the time of writing. As such, this section presents a synthesis of data from consultants’ reports, government data sites, and earlier studies.
3.1 Temperature and precipitation
The region has warm summers and cold snowy winters (Fig. 6). The mean annual temperature is approximately 5 to 6°C, and average monthly temperatures range from -9°C in February to 20°C in July. Annual precipitation varies between 900 mm and 1050 mm per year. Monthly precipitation ranges from 60.3 mm in February to 97 mm in September.
Figure 6. Average temparature and precipitation across the SWPR (based on 1971–2000 Climate Normals, Environment Canada, http://climate.weatheroffice.ec.gc.ca/Welcome_e.html).
3.2 Evapotranspiration
Evapotranspiration (ET), the sum of evaporation and plant transpiration, is difficult to calculate in comparison to quantities such as precipitation and runoff because of the complex interaction between vegetation and the atmosphere, and the difficulty of regionalizing site specific measurements (e.g. Telmer and Veizer, 2000). Estimated actual ET for various land cover types ranges from 150 mm/year in urban areas to 640 mm/year on open water (Fig. 7A). Average ET for the SWPR is estimated to be 425 mm/year. A recent water-budget analysis of the SWPR calculated that approximately 43 % of precipitation is lost to ET.
Using the Penman and Thornthwaite methods, monthly potential ET was calculated as having strong seasonal variability with no ET in winter and peak ET in July (Fig. 7B). Annual estimates of potential ET for the region are similar, with values of 664 and 579 mm for the Penman method and Thornthwaite method, respectively. It is significant to note that potential ET exceeds precipitation during the summer months of June through August. Consequently, water courses in the area are sustained by groundwater flow during this period. 7
Figure 7. A) Calculated annual actual evapotranspiration by landuse (CH2MHill, 2001). B) Derived monthly potential evapotranspiration values for the SWPR using the Penman and Thornthwaithe methods (National Land and Water Information Service http://www.agr.gc.ca/nlwis-snite/index_e.cfm, Agriculture Canada, revised 1997).
3.3 Hydrology
The Vars–Winchester esker lies within the South Nation River watershed, which covers approximately 3822 km2, or 59% of the SWPR. The watershed is relatively flat and poorly drained. The South Nation River is about 177 km in length with an elevation change of about 80 m from the headwaters in the south to the confluence with the Ottawa River. The SWPR stream network has a cumulative length of ~10,252 km. Eighty eight percent of the streams are permanent and 12% are intermittent. Using fish species as an index, 93% of the permanent streams are classified as warm-water streams. The geographic distribution of the 7% of cold-water streams could potentially be used as a direct indicator of groundwater discharge.
Figure 8. Average daily water discharge in the South Nation River measured between 1915 and 2003 at a gauging station just upstream of the confluence with the Ottawa River. Data from Environment Canada (http://www.wsc.ec.gc.ca/products/hydat/main_e.cfm?cname=archive _e.cfm) 8
The Water Survey of Canada branch of Environment Canada operates stream-gauging stations on the inland rivers and streams, which monitor discharge in 4,423 km² of the 6,389 km2 drainage area. The highest discharges occur in March and April because of springtime snow-melt runoff (Fig. 8), with 70% of all measured annual discharge occurring by the end of April. Calculations of baseflow across the SWPR (using an automated hydrograph separation technique) indicate that baseflow contributes an average of 56% of streamflow. Between June to August, evaporation in the SWPR tends to exceed precipitation, and most streams are likely sustained entirely by baseflow. However, baseflow may not always be sufficient: during an extreme year, stream gauge measurements from Plantagenet Springs indicate that the South Nation River dried out completely between August 1930 and March 15, 1931 (Chin et al. 1980), likely due to a drought-induced low groundwater-level, or possibly pumping from the river.
3.4 Hydrogeology
With respect to the groundwater system, the stratigraphy in the SWPR can be reduced to five layers. From bottom to top these are i) fractured bedrock and stratified sediment overlying the bedrock surface (contact-zone aquifer), ii) till (aquitard), iii) esker (aquifer), iv) Champlain Sea mud (aquitard), and v) littoral–fluvial sand (shallow aquifer).
Approximately 88% of wells draw water from the contact-zone aquifer (Fig. 9). Traditionally, this aquifer has been interpreted to consist of two elements, the uppermost several metres of fractured bedrock, and sorted sediment between bedrock and till (sub-till sediment). No sub-till sediment was intersected in the 2007 drilling campaign (Section 4). A thin (<2 m), highly permeable gravel sheet over till was intersected locally (Sharpe and Pugin, 2007), however, and may represent a previously- unrecognized component of the contact-zone aquifer (see Sections 4 and 5). Regionally, the contact- zone aquifer is considered semi-confined; however, where it is overlain by dense fine-grained till and/or Champlain Sea clays it may be locally confined.
Twelve percent of wells extract water from the two unconsolidated-sediment aquifers (eskers and littoral–fluvial sand), with most water being drawn from eskers. Esker aquifers are prolific, with production rates of up to 500 gallons per minute (31 L/s).
In general, shallow groundwater appears to converge toward the surface water network (Fig. 9). Groundwater elevations in the South Nation River watershed are highest to the west along the boundary with the Rideau River watershed, and to the south and southeast along the boundary with the Raisin River watershed. The general direction of groundwater flow in the bedrock is likely similar to that of the overburden.
A preliminary water budget for the SWPR identified that ~28% of precipitation (or 51% of the water surplus, which is the water available once evapotranspiration is removed from precipitation) recharges the groundwater system. Most of the water that recharges the unconfined littoral–fluvial aquifer stays within the unconfined aquifer and discharges locally into surface water bodies or is intercepted by tile drains and directed to nearby watercourses. The esker aquifer is recharged from precipitation that falls directly on the exposed esker and possibly from streams that intercept the esker (see Section 5). Little or no water appears to move through the overlying mud and into the esker. Only ~2% of the recharge may reach the contact zone aquifer. This stresses the importance of quantifying the consumptive use and sustainability of water resources within the contact zone aquifer. It is inferred that a sizable percentage of recharge to the contact zone aquifer within the South Nation River watershed occurs through the topographically higher terrain within the Rideau River watershed where bedrock outcrops. 9
Figure 9. Piezometric surface from water-well data for the contact-zone aquifer in Eastern Ontario (courtesy of Charles Logan, Geological Survey of Canada). Contours are in metres above sea level. Insufficient data are available to construct similar regional maps for esker- and littoral–fluvial sand aquifers.
3.5 Inferences from regional groundwater geochemistry Sam Alpay and Marc J. Hinton, Geological Survey of Canada
Existing knowledge of the groundwater geochemistry within the Vars–Winchester esker includes a regional study of well water quality for the purpose of delineating regional groundwater flow (Charron, 1978). In the study, Charron (1978) points out that most residential groundwater wells penetrate the bedrock, which he considered to be effectively impermeable, except in the fractured zone at the contact with unconsolidated sediment. He also assumes that the overlying sediment was derived from the bedrock and that the groundwater chemistry from the sediment would reflect that of the bedrock, which may not necessarily be the case for allochtonous esker materials.
Charron (1978) used the distributions of chemical species in groundwater to infer regional groundwater flow directions assuming Chebotarev’s (1955) theoretical chemical sequence. Chebotarev classified hydrochemical facies (low salinity, transition, high salinity) by their relative position along a groundwater flowpath and residence time of encounter with different geologic media. For example, the ------theoretical sequence from recharge to discharge zones is: HCO3 ĺ HCO3 - Cl ĺ Cl – HCO3 ĺ Cl – 2- 2- - 2- - SO4 or SO4 –Cl ĺ SO4 ĺ Cl . Recharge areas, such as near Sarsfield, were identified as dilute - waters with low concentrations of total dissolved solids (TDS) and high HCO3 ; discharge areas were characterized by higher TDS and Cl- (e.g., Alfred Bog). The flow directions inferred by Charron (1978) occur in two patterns: along the esker system from Winchester to the north and from Sarsfield eastward toward the Alfred Bog, interpreted as the major regional discharge zone. Comparison of these flowpaths with more detailed mapped piezometric surfaces (CH2MHill, 2001) indicates that 10
Charron (1978) has identified the major groundwater discharge area but, due to the regional scale of his study, has not considered that several smaller flow systems occur within his regional flow pathways.
Charron’s (1978) regional study did not explicitly recognize the Vars–Winchester esker aquifer. A Piper plot of well data in its vicinity demonstrates a high diversity of groundwater geochemistry with Ca2+– - 2+ 2- + - + - HCO3 waters evolving to Ca –SO4 , Na –HCO3 , and Na –Cl waters (Fig. 10). When considered with the piezometric surface (CH2MHill, 2001) and the bedrock geology map (Charron, 1978), these results appear to suggest both more rapid geochemical evolution along shorter flowpaths and the influence of different bedrock lithologies.
Current plans for a groundwater geochemistry study in the Vars–Winchester aquifer aim to characterize and understand groundwater flowpaths and interactions between the esker and the bedrock using a combination of traditional geochemical methods and specific chemical tracers. In addition to studies of major ions, several analytical methods (e.g., ICP-MS, ICP-ES, dissolved organic carbon, 18GO and GD) will be used to identify natural tracers of groundwater flowpaths specific to the region and to quantify mixing of different source waters. Sampling at the regional scale will be mostly at residential well sites from representative geologic terrains within and surrounding the Vars– Winchester esker. Geochemical tracers will be used to establish a better understanding of the flow dynamics and mixing of groundwater both on the regional scale of the esker and on the scale of local flow systems. Geochemical signatures will also decipher between natural and anthropogenic impacts on water quality. The results will be used for understanding the local domestic, municipal, and agricultural water supplies in the Vars–Winchester aquifer and potentially for other esker aquifer systems in Eastern Ontario.
Figure 10. Piper plot of 35 samples selected in the vicinity of the Vars–Winchester esker (data from Charron, 1978). 11
4.0 Geological Setting Don I. Cummings, Geological Survey of Canada & South Nation Conservation Authority
4.1 Landscape and bedrock geology
Regional-scale landscape elements in the vicinity of the Vars–Winchester esker are tectonic in origin, but have been modified substantially by glacial and post-glacial erosion and deposition (Fig. 11). Uplands to the north (Laurentian Highlands), south (Adirondack Mountains) and west (Frontenac Arch) are cored by igneous and metamorphic rocks of the Precambrian Shield and have a relatively thin, discontinuous sediment cover (mostly till). The St. Lawrence Lowlands, by contrast, have a relatively thick cover of surficial sediment1 that is underlain by downfaulted, nearly horizontal Paleozoic carbonate and siliciclastic rocks2 (Kay, 1942; Williams, 1991; Dix et al. 1997). The structural fabric is predominantly east-west. At a local scale, several distinct landforms are discernable, including partially buried, north-south oriented esker ridges (e.g., Gorrell, 1991), north-south oriented drumlins (e.g., Sharpe, 1979), flat, near-horizontal Champlain Sea mud plains (e.g., Gadd, 1986), and the (huge) Ottawa River incised valley (Fig. 12).
Figure 11. Regional landscape elements in the vicinity of the Vars–Winchester esker-aquifer. Uplands to the north (Laurentian Highlands), south (Adirondacks) and west (Frontenac Arch) are underlain by igneous and metamorphic rocks of the Precambrian Shield. Sediment cover in uplands is generally thin, and consists primarily of subglacial till. The St. Lawrence Lowlands are underlain by normal-faulted, nearly horizontally stratified Paleozoic carbonate and siliciclastic rocks. Sediment cover in the lowlands is generally thicker (up to 170 m locally), due primarily to the enormous supply of mud to the post-glacial Champlain Sea.
1 Surficial sediments in the basin are typically 5–40 metres thick, but may reach 90–170 metres in bedrock depressions (valleys?) near the Ottawa River (e.g., MacPherson, 1968; Aylsworth et al. 2000; Aylsworth et al. 2003; Ross et al. 2006).
2 All Mesozoic rocks in the region, with the exception of the Monteregian instrusions, were likely eroded and shed through the paleo–St. Lawrence River to the Atlantic Ocean during the breakup of Pangea (Cummings and Arnott, 2005). 12
Figure 12. Data collected from the Vars–Winchester esker-aquifer during the 2006–2007 field season by the Geological Survey of Canada. Field-trip stops and major geomorphic elements are also shown. 13
Figure 13. Stratigraphy in the vicinity of the Vars–Winchester esker-aquifer. Vertical exaggeration is ~5. Cartoons are based primarily on seismic and continuous cores collected from Route 300, with additional observations from other parts of the study area (e.g., Sharpe and Pugin, 2007). (A) Key seismic reflections, lap-out relationships and seismic facies. (B) Physical, biological and chemical properties of the Quaternary strata, as ascertained by cores drilled to groundtruth seismic transects. North of Watson Road pit, the sandy carapace is absent. Porewater salinity in mud succession typically decreases upward, and may also be depressed near base locally. (C) Estimated porosity (ij) and permeability (hydraulic conductivity, K) of various units, as well as the characteristics of water-wells completed in the fractured bedrock and esker aquifers. Thin (<1 m), highly permeable sand beds exist locally in the stratified mud unit (not shown). (D) Sequence stratigraphic interpretation. As with non-glacial sedimentary systems, the main stratigraphic pattern is interpreted to reflect gradual changes in sediment supply and accommodation space (associated with horizontal translation of the ice front, not the shoreline); however, unlike non-glacial sedimentary systems, rapid meltwater events disrupt this pattern. This punctuation of gradual ice-mediated processes by rapid meltwater events is considered to be a hallmark of glacial sedimentary systems. 14
4.2 Stratigraphy
Between July 2006 and March 2007, the Geological Survey of Canada collected ~20 line- kilometres of seismic data and drilled 18 wells, six of which were continuously cored (Fig. 12). Although data analysis is still at a preliminary stage, the stratigraphy appears to consist of four lithologic units (bottom to top): 1) bedrock, 2) diamicton (till), 3) sand and gravel (esker), and 4) mud with minor sand near base and/or top (Champlain Sea deposits) (Fig. 13). Previous authors describe a similar stratigraphic succession (e.g., Johnston, 1917; Gadd, 1961, 1986; Douma and Nixon, 1993; Shilts, 1994; Pair and Rodrigues, 1993; Aylsworth et al., 2003; Ross et al., 2006).
4.2.1 Bedrock
With the exception of the single well drilled to bedrock along Route 300, which intersected shale, non-fissile carbonate-mudstone was recovered at the base of newly drilled wells (Fig. 14)3. Similar rock is mined in nearby quarries. In cores collected south of the Watson Road pit, the carbonate mudstone is typically massive and devoid of fossils or terrigenous material (Fig. 14A). North of the Watson Road pit, it commonly contains contain skeletal fragments and wispy shale- layers (0.1–2 cm) (Fig. 14B).
Figure 14. Bedrock lithologies intersected in new wells. Most wells intersected carbonate mudstone, the only exception being one well along Route 300 that intersected shale. (A) South of the Watson Road pit, the carbonate mudstone tends to lack skeletal fragments and terrigenous mudstone layers (photo from Ball Road well, 16.5 m below land surface), whereas north of the Watson Road pit (B), shale layers and skeletal fragments are present (photo from Watson Road well 1, 6.4 m below land surface). (C) Fissile, horizontally- stratified shale (no HCl reaction) was intersected at the base of the single well drilled to bedrock along Route 300 (photo from GSC-BH-07-02, 37.80 m below land surface).
From a hydrogeological perspective, it is important to note that the bedrock is typically fractured. For example, several of the cores fell apart along vertical fractures when extruded from the core barrel. Also, metre-spaced vertical fractures are ubiquitous in nearby quarries, and invariably extend beneath the quarry floor (Fig. 15). Visible weathering (e.g., pervasive rusty colour) is typically absent below the bedrock surface. However, hydrogeological data suggest that the upper several metres of bedrock has an order-of-magnitude higher permeability than the bedrock
3 The sediment–bedrock interface was invariably cored in wells (even “non-cored” wells) to insure that piezometers were placed at the sediment–bedrock interface. 15 below (George Gorrell, personal communication, 2007), which may reflect solution enhancement of the fractures by meteoric waters.
Figure 15. Carbonate-mudstone bedrock in outcrop (Canaan Quarry, 3.5 km east of Sarsfield). Note prominent vertical fractures (arrows) and near-horizontal stratification. The other vertical features are holes that were drilled for blasting.
4.2.2 Bedrock surface
The bedrock surface was not examined during the study, but has been investigated by previous workers (e.g. Sharpe, 1979; Ross et al., 2006). Smooth, unweathered bedrock surfaces are commonly striated (Fig. 16). Striae are spaced millimetres apart, are decimetres to metres in length, and are less than one millimetre in depth. Most workers believe that they form by differential movement of asperities (hard clasts) in basal ice over bedrock, and that different populations of striae record different ice-flow directions.
In the immediate vicinity of the Vars–Winchester esker, bedrock striae are oriented nearly N-S (Fig. 17; Sharpe, 1979). Near Montreal and on the north face of the Adirondacks, a second population of striae that trend NE-SW is observed. Most authors argue that N-S striae are related to regional ice-flow during the last glacial-maximum, whereas younger striae record topographically-steered flow after the onset of ice-sheet thinning (e.g., Ross et al., 2006). A clear reconstruction of the ice-flow event-sequence based on striae data, however, is muddled by inconsistent cross-cutting relationships; for example, NE-SW striae commonly cross-cut N-S striae, but also locally appear to be cross-cut by them.
Where esker material immediately overlies bedrock, sculpted forms such as potholes, flutes, cavettos, sichelwannen and muchelbruchen are commonly present (Fig. 18; Henderson, 1988; Sharpe and Shaw, 1989). These sculpted forms, or s-forms, which commonly occur in bedrock- floored rivers, are also commonly observed off esker, even on higher ground (e.g., Gilbert, 2000). In order for s-forms to be generated, particles in the flow must spontaneously move at high angles to the mean flow over short distances (centimetres to metres) without the aid of pre- existing obstacles (e.g., bedrock asperities). In other words, the flow has to be turbulent. Flows in the atmosphere and hydrosphere are almost invariably turbulent because of the high inertia-to- viscosity ratios (Reynolds numbers) of naturally flowing air and water. Glacier ice, however, deforms in an extremely slow, laminar (non-turbulent) fashion (as would any subglacial “till slurry”) because of the extremely low inertia-to-viscosity ratio of flowing ice. Macroscopic 16 transverse motion may occur, but only if instigated by a pre-existing obstacle4. As such, s-forms are interpreted to have been eroded by flowing meltwater at the base of the glacier.
The bedrock surface generates a laterally continuous, high-amplitude seismic reflection that is typically well resolved, except in some places where thick gravel overlies bedrock. In locations some distance from the Ottawa River (south of the Watson Road pit), the bedrock reflection is relatively flat. By contrast, closer to the Ottawa River (north of the Watson Road pit), the bedrock reflection becomes highly irregular. Here, bedrock ridges (5–10 m high) oriented parallel the Ottawa River (and parallel to the structural fabric) protrude locally through the mud plain (Fig. 12), and bedrock cliffs up to 10 m high outcrop on the river-facing sides of paleo-islands in the Ottawa River incised valley. In addition, a ~60-metre-deep bedrock valley is interpreted to extend north- south across the area based on seismic and well log data (see Stop 1 in Section 6). Given these observations, it seems probable that a number of flows (ice and meltwater) coming from both the north (when the area was subglacial?; see Sharpe and Shaw, 1989; Sharpe and Pugin, 2007) and down ancestral courses of the Ottawa River (when the area was ice free?; see Teller (1988) and below) sculpted the bedrock near the Ottawa River over multiple glaciations, accentuated the structural grain locally, and generated a complex, composite, irregular erosional surface.
Figure 16. Striae on limestone near the St. Lawrence River (east of Brockville). Where unweathered, the bedrock surface is commonly striated. The compass points northward. Two sets of striae exist, an older one that trends southeast and a younger one that trends south. Note that these striae directions are at odds with other striae from nearby areas, which suggest that regional southward ice-flow was followed by flow parallel to the St. Lawrence River, the message being that striae data paint a complicated story that is not yet completely resolved. From Terasmae (1965).
The bedrock surface forms the effective barrier between fresh (meteoric) water in Quaternary sediments derived from precipitation, and the relatively salty (connate) water in the sedimentary rock that was likely trapped at the time of deposition (Charron, 1978). Any enhanced permeability across this surface is associated with fractures. Although data are limited, the fracture permeability appears to be most extensively developed in the uppermost several metres of bedrock (hydrogeological models commonly use a depth of 3 metres). At greater depths, permeability commonly decreases by an order-of-magnitude (George Gorrell, personal communication, 2007).
4 Flow transverse particle motion is actually ubiquitous, but on a molecular scale. 17
Figure 17. Orientations of drumlins and selected striae on bedrock. Modified from Ross et al (2006).
Figure 18. Sculpted forms (sichelwannen) on Precambrian bedrock along Highway 15 near Joyceville (in between Kingston and Ottawa). This location is adjacent to, not underneath, an esker. Note that three nested scales of sichelwannen exist (as indicated by arrows of different size). Turbulent water acting on bedrock-floored river channels and on cohesive muds (e.g., turbidite flute casts; Allen, 1971) is known to generate identical forms. As such, these sculpted bedrock forms are interpreted to have been generated by fast-moving subglacial meltwater. Similar forms are found on the bedrock surface beneath nearby eskers (Henderson, 1988). 18
4.2.3 Diamicton (till)
Stiff diamicton (average 1–3 m thick; maximum drilled thickness 10 m) was intersected above bedrock in wells drilled adjacent to the Vars–Winchester esker and locally beneath the broad sandy flanks of the esker. By contrast, it was almost never intersected beneath gravel units within the esker. Based on new well and seismic data, the stiff diamicton appears to form a locally discontinuous yet areally extensive sheet that does not necessarily thin over bedrock highs or thicken into bedrock lows5. Stiff diamicton tends to form the surface-sediment layer on uplands.
In core, the diamicton is typically grey (Fig. 19) except at the Watson Road pit where it is orange- brown. The diamicton is much stiffer than Champlain Sea muds buried to the same depth. For example, it is too stiff to push a knife into with ease. As such, it is considered to be “overconsolidated”. It is also typically massive, except at one location near Winchester where well-sorted fine-sand layers (< 1 cm thick) are interstratified within the otherwise stiff diamicton. Carbonate-mudstone clasts of local derivation predominate, with subordinate amounts of igneous clasts that were likely derived from the Precambrian Shield to the north, and sandstone clasts that may have been derived from the Nepean Formation (Cambrian), which outcrops to the north (e.g., Buckingham) and east (Williams, 1991). Clasts are sub-angular to angular.
The stiff diamicton was not observed in gravel pits, but similar units have been described by previous workers in nearby outcrops, most notably at the St. Lawrence Seaway excavations near Cornwall. Here, workers distinguished two locally-superimposed diamicton sheets, the Malone Till, a very stiff, grey diamicton with at least 90% sedimentary clasts (presumably of local derivation), and the Fort Covington Till, a sandier diamicton that contains a greater abundance of igneous clasts (presumably of Shield derivation) (Terasmae, 1965). Two stiff-diamicton sheets are also commonly described in the greater Montreal area (Prest and Hode-Keyser, 1977; Ross et al. 2006); these diamictons may be either superimposed on top of one another or separated by several metres of stratified sediment. In the vicinity of the Vars–Winchester esker-aquifer, there is currently no evidence suggesting more than one stiff diamicton exists over bedrock (see also Johnston, 1917). Lateral changes in diamicton colour and texture may simply reflect lateral facies change within a single depositional unit.
Previous authors have interpreted the stiff diamicton unit to be a subglacial till (e.g., Johnston, 1917; Sharpe, 1979; Kettles and Shilts, 1987), which is supported by the observations presented above (e.g., angular clasts, massive appearance, overconsolidation, sheet-like geometry), in addition to the stratigraphic position of the stiff diamicton above bedrock and below eskers and Champlain Sea deposits. Massive diamictons are thought to form subglacially by two sometimes- coeval processes, lodgement (“plastering”) of sediment onto the substrate from basal ice, and/or by shear homogenization (deformation) of the substrate beneath basal ice (e.g., the shell-bearing, locally thrust-faulted St. Nicolas Till near Quebec; Cummings and Occhietti, 2001). Given that widespread subglacial erosion apparently preceded till deposition, the stiff diamicton sheet is most reasonably interpreted as a lodgement till deposited after the subglacial accommodation- space switched from being negative (sediment entrainment & bedrock abrasion) to positive (till deposition)6. Rare fine-sand layers present locally within the otherwise overconsolidated, massive diamicton may have been deposited by locally-flowing meltwater at the base of the glacier in between periods of lodgement. Their paucity, in addition to the overconsolidated nature of the diamicton, suggests that deposition did not occur by wholesale, regional melt out of debris
5 It is important to note, however, that it the unit is eroded locally. See Sharpe and Pugin (2007).
6 The reason for and timing of the switch is unclear, but presumably relate to a change in subglacial conditions associated with ice-margin retreat. 19 from stagnant, stratified basal ice (cf. Kettles and Shilts, 1987). Following the lead of previous studies (Sharpe, 1979; Kettles and Shilts, 1987), we term the diamicton the Fort Covington Till.
Figure 19. Representative photo of stiff diamicton (interpretation: lodgement till) from Ball Road well, 15.25 m depth below land surface. Where intersected in the new wells, the stiff diamicton unit invariably sharply overlies bedrock, and is sharply overlain by stratified sediment (sand, gravel or mud).
4.2.4 Top of stiff diamicton (drumlins)
The top surface of the regional till, where exposed, is sculpted into elongate, parallel, streamlined ridges that are oriented north-south, and are on average 3–10 m high, 0.5 km wide and 1 km long (Fig. 12). Although this surface was not investigated in detail, we concur with previous authors that the streamlined forms are drumlins (e.g., Sharpe, 1979). Given that s-forms commonly ornament the bedrock surface where till is absent, both beneath eskers (Henderson, 1988) and adjacent to eskers (Sharpe and Shaw, 1989; Gilbert, 2000), and that a thin (< 2 m) gravel sheet occurs above the regional till and beneath Champlain Sea mud in the north part of the study area (Sharpe and Pugin, 2007), the meltwater hypothesis, which argues that drumlins were carved by meltwater floods beneath the ice sheet (e.g., Shaw, 1996), cannot be ruled out. Arguments presented by proponents of the meltwater hypothesis (Sharpe and Shaw, 1989; Shaw et al., 1996; Sharpe and Pugin, 2007) therefore need to be weighed against those of workers who believe the drumlins formed by the action of ice (Ross et al., 2006). Although we find the physics- based arguments of proponents of the meltwater hypothesis (e.g., Shaw, 1996) in general more convincing than those presented by supporters of the ice hypothesis, and argue that large lakes do exist and can drain catastrophically beneath ice sheets (e.g., Wingham et al., 2006; Siegert et al., in press), in reality, the facies architecture relative to the form of the drumlins needs to be investigated before a final conclusion can be drawn. However, the thin, highly permeable gravel- 20 sheet present locally off-esker over till and below Champlain Sea mud is difficult to explain unless a sheet-like subglacial discharge of meltwater is invoked7.
It is possible that the hydrogeological significance of the top-till surface—and specifically the thin gravel-sheet that mantles it locally—has not been fully appreciated. The gravel-sheet has a high permeability. For example, at one cored well along French Hill Road (GSC-BH-07-06), circulation of drilling mud dropped considerably when this layer was intersected, leading drillers to call it a “thirsty” gravel. (In fact, more circulation was lost here than at a nearby well (GSC-BH-07-05) that intersected the esker.) Because the bedrock surface in the north end of the study area is irregular and protrudes locally through the mud plain, the permeable gravel-sheet may extend to surface at these locations. It is therefore possible that precipitation falling on the bedrock highs is readily transferred to the subsurface along the top of the till, which may in turn help recharge the esker aquifer (David Sharpe, personal communication, 2007). The degree of hydraulic connection between gravel sheet and esker is unknown. However, given their similar stratigraphic position, a connection is suspected.
4.2.5 Sand and gravel (esker)
An elongate, 40-km long, partially-buried ridge of sand and gravel extends north-south across the study area (Fig. 12). It is one of several such landforms in the Champlain Sea basin. Like previous authors (e.g., Gorrell, 1991), we interpret this landform to be an esker, and refer to it hereafter as the Vars–Winchester esker. Wells completed in the esker yield > 500 US gallons per minute (Anne-Marie Chapman, personal communication, 2007), which makes it one of the best aquifers in between Ottawa and Montreal. All other high-yield aquifers in the region are also eskers.
Although it is a distinct landform or “container”, the Vars–Winchester esker consists of two key elements, a gravely central-ridge (highly permeable) and a sandy-fan carapace (moderately permeable) (Fig. 13). Neither element generates coherent seismic reflections internally, but the boundaries of the esker “container” are commonly well resolved in seismic transects, an exception being where thick gravel obscures underlying sediment and/or bedrock (Fig. 20).
The gravely central-ridge of the Vars–Winchester esker is 2 to 20 metres high (average 15 m) and 100 to 200 metres wide (average ~150 m) (Fig. 20A). Its cross-sectional area varies, but does not systematically increase or decrease along the esker. The flanks of the gravely central- ridge dip between 10q and 30q. Clast lithologies are similar to the adjacent till: carbonate mudstone predominates, with minor percentages of igneous (granite), sandstone, and (surprisingly friable) shale clasts. Where intersected in wells, the gravely central-ridge typically overlies bedrock8. Locally, it appears to bifurcate and rejoin over several kilometres. Bifurcation and local widening of the gravel ridge is especially apparent in the north end of the study area, on northward-dipping bedrock surfaces. Although the gravely central-ridge is interpreted to be present in all new seismic transects that cross the esker, the continuity (and therefore hydraulic connectivity) of the gravels between seismic transects is difficult to ascertain, leaving open the possibility that breaks may be present. Where exposed, the gravely central-ridge consists of well- rounded pebbles and cobbles organized into thick (1.5–2.5 m) high-angle (dune) cross-stratified beds that dip southward, parallel to the esker axis. Thick (1–2 m) sand layers are present locally (see Stop 2 in Section 6).
7 Note that up to 2 metres of bouldery gravel was also intersected off-esker between till and Champlain Sea mud in long cores collected west of Montreal (see Ross et al. 2006).
8 Gorrell (1991) notes that the largest clasts (commonly rounded boulders) tend to occur at the base of the gravely central- ridges. A subtle fining upward trend was noted in one of the Route 300 cores, but it is unclear if such trends are present everywhere along the gravely central-ridge. 21
The gravely central-ridge is interpreted to have been deposited in a meltwater conduit (R- channel) that was thermally eroded (and corraded?) into ice at the base of the glacier. It is considered to be the most proximal element of the esker system. Lithologic similarity of its clasts and clasts in the till suggests the gravely ridge was sourced in part from the till; basal ice likely contributed sediment also, given volumetric considerations. Bifurcation and subtle widening of the ridge on northward-facing slopes may reflect a decrease in the wall-melting–to–viscous- heating ratio in the R-channel associated with the steeper, adverse bedrock inclines (Röthlisberger, 1972; Shreve, 1972, 1985). The flows must have been relatively fast and deep, given that they carried boulders up to 1 m in diameter and deposited dune cross-beds between 1.5 and 2.5 m in height. If scaling relationships for fluvial environments are assumed to apply, the flows likely traveled at speeds of several metres per second9 and that the R-channel was likely several 10s of metres high10. This in turn suggests discharges of 1000s of cubic metres per second11. Discharge may have therefore been similar to that of the modern Ottawa River (~2000 m3/s) and up to two orders-of-magnitude greater than R-channel discharge in modern (small) alpine glaciers during melt season (e.g., Østrem, 1975; Hooke et al. 1985).
Sandy-fan deposits with variable amounts of pebble gravel sharply overlie the gravely central- ridge locally, forming a carapace that is relatively wide (0.4–2 km) and gentle-flanked (1–5q) (Fig. 20B). The longitudinal extent of the sandy-fan carapace is poorly constrained, but based on seismic data it appears to be present in most places along the esker south of Watson Road pit. Core data suggest the distal parts of the sandy-fan carapace commonly overlie till. In outcrop, the sandy-fan carapace is composed of multiple, mound-shaped, upward-coarsening units (“lobes”) that are stacked compensationally on top of each other; sharp-based units, commonly gravely, interrupt this motif locally (see Stop 3 in Section 6)12. Climbing ripples, dunes (which also rarely climb), low-angle (antidune) cross-strata, and diffusely-laminated channel fills are common. In a continuous core from Route 300, marine-shell fragments (mostly Portlandia arctica)13 were observed at two intervals in the sandy-fan carapace (Figs. 20B, 21).
9 Calculated using the Helley method (Table 3 in Costa, 1983).
10 Calculated using the method outlined in Leclair and Bridge (2001).
11Assuming a flow width of 125 m, flow speed of 2 m/s, and a (conservative) flow depth of 20 m.
12 In adjacent eskers, where more outcrops along the esker exist, architectural elements in the sandy fan are commonly sharp-based and ungraded to upward-fining. It is likely that similar variability exists in fans of the Vars–Winchester esker, but that this variability is simply not observed due to limited exposure.
13 André Martel from the Canadian Museum of Nature has identified the shells as being from the Yoldidae and Nuculidae families. Using photos for comparison, many of the fragments appear to be Portlandia arctica. 22
Figure 20. Representative seismic transects across the esker in the north end (A) and south end (B) of the study area. Vertical exaggeration is 4X in both transects. (A) North of Watson Road pit, the sandy-fan carapace is absent, and the esker consists only of a narrow gravely ridge. By contrast, in most places south of Watson Road pit (B), the esker consists of a narrow gravely central-ridge with a broad sandy-fan carapace. Seismic images courtesy of André Pugin, Susan Pullan and Jim Hunter. 23
Figure 21. Shell fragments in subaqueous-outwash fan sands (23 m below land surface, GSC-BH-07-03 well). They are marine taxodont bivalve shells from the Yoldidae and Nuclidae families (André Martel, personal communication, 2007). Note that the periostracum is preserved on one of the shell fragments. Although not visible in this photo, the sand bed contains abundant granule- and pebble-sized mud rip-ups. Their presence may suggests that the marine bivalves proliferated on the subaqueous fan when R-channel discharge into proglacial water was low, salinity was higher, and mud was being deposited (during winter?), but that they became entrained, along with the mud, when R-channel discharge increased (during summer?).
The sandy-fan carapace is interpreted to have been deposited by sediment-laden jet–plume pairs (e.g., Fischer et al., 1979; Syvitski, 1989; Powell, 1990; Hoyal et al., 2003; Russell and Arnott, 2003) that were issued from the R-channel into standing, brackish water of the Champlain Sea (Fig. 22). The general impression gained from the sedimentology of outcrop exposures is one of rapid sedimentation from rapidly decelerating unidirectional flows: the presence of bedforms attests to the tractive nature of sediment transport, whereas climbing bedforms (ripples and some dunes) and diffuse stratification attest to high rates of suspended-sediment rain-out (e.g., Arnott and Hand, 1989). The upward-coarsening, mound-shaped depositional elements at the Watson Road pit (see Stop 3 in Appendix) are interpreted to be progradational fan lobes. Compensational stacking of these sand bodies is interpreted to have been produced by avulsion of the jet–plume depocenter14. Sharp-based, ungraded to upward-fining units either reflect progradation of proximal-fan channels over distal-fan deposits, or abrupt increases in discharge. In either case, these upward-fining units likely correlate downflow to fan lobes. Backstepping of the ice margin during esker deposition is interpreted to have superimposed subaqueous-fan sand over the gravely central-ridge (Fig. 13D). Finally, although we believe that flow-expansion can occur subglacially under certain circumstances (e.g., Oak Ridges and Oro moraines, Southern Ontario), given the occurrence of shells in the sandy-fan carapace of the Vars–Winchester esker, the hypothesis that all eskers (i.e., long eskers with a gravely central-ridge and a broad sandy-fan carapace) are formed fully subglacially (e.g., Shreve, 1985) needs to be re-examined.
14 The small size of these lobes at the Watson Road pit (Stop 3) may suggest deposition under lower-discharge conditions than at the Kemptville pit (Stop 5) (e.g., Powell, 1990). 24
Figure 22. Facies model for the Vars–Winchester esker system. The gravely central-ridge is interpreted to have been deposited in a subglacial meltwater channel (R-channel), and the sandy-fan carapace is interpreted to have been deposited where the R-channel debouched into deep, standing, proglacial, marine- influenced water. To fully understand the Vars–Winchester esker, this facies model—a snapshot of the depositional system in time—needs to be integrated with the sequence-stratigraphic model (Fig. 13D), which takes into consideration how longer-term changes in sediment supply and accommodation space generated the observed stratigraphic succession.
We understand our suggestion that the Champlain Sea existed proglacially during esker deposition will be met with controversy. Most researchers over the past 25 years have argued that a proglacial lake existed during esker deposition, and that the basin only became marine after the lake drained into an inland arm of the Atlantic Ocean, which is thought to have encroached up the St. Lawrence River valley through the glacier to a position near Quebec City (Thomas, 1977). The proglacial lake is believed to have been large: it is thought to have extended from Quebec City (Parent and Occhietti, 1988) to Lake Ontario (Clarke and Karrow, 1984), and from south of the Canada–US border up the Ottawa River (Pair and Rodgrigues, 1994). Support for this hypothesis is derived from the presence of rhythmically laminated sediment (“varves”) at the base of the Champlain Sea mud succession that commonly contain only Candona subtriangulata, a benthic freshwater ostracode (Anderson et al. 1985; Parent and Occhietti, 1988; Rodrigues, 1992; Pair and Rodrigues, 1993). Some authors disagree with this hypothesis, suggesting instead that the “varves” may be Champlain Sea deposits (e.g., Sharpe, 1988). Despite this, almost all workers agree that the “varves” are likely distal subaqueous- 25 outwash–fan deposits (e.g., Naldrett, 1988; Gadd, 1988; Pair and Rodrigues, 1993; Ross et al., 2006).
Given the above compelling arguments for the existence of a proglacial lake during esker deposition near Ottawa, we discuss our findings in greater detail. Marine-shell fragments were observed 2 metres and 11 metres below the top of the sandy-fan carapace in one of two long- cores (GSC-BH-07-03) collected just north of Embrun (Fig. 20B). The esker “container” is clearly resolved in the seismic data at this location (see also Stop 4 in Appendix). The shell-bearing strata are not reasonably interpreted as having slumped from the esker crest following deposition (11 m is too deep), and are not beach deposits formed during isostatic rebound (the esker is buried by 10 m of mud). The shells were also not introduced during the coring process because the core collected from 11 m below the top of the esker is pristine. It is also very unlikely that the shells were reworked from older sediments because they are extremely thin and fragile, and their periostracum (organic coating) is typically preserved. As such, we argue that the most reasonable interpretation, especially given that Portlandia arctica is known to thrive in modern, freshwater-influenced prodeltaic environments in Canadian high-arctic fjords (e.g., Syvitski, 1989; Aitken and Gilbert, 1996), is that the marine bivalves simply lived on the subaqueous-outwash fan.
Candona subtriangulata can tolerate high turbidity (Rodgrigues, 1992), but the species has never been identified in modern brackish or marine environments, and is known to tolerate only minor levels of sodium (1–14 mg/L), chloride (1.2–15 mg/L) and sulphate (0.1–27 mg/L) (L. Denis Delorme, personal communication, 2007). However, we believe that Champlain Sea microfaunal assemblages speak for themselves: in addition to occurring by itself, in some places Candona subtriangulata is found with or below benthic marine foraminifera in unbioturbated/undisturbed strata (e.g., Cronin, 1977; Hunt and Rathburn, 1988)15. Given that the Champlain Sea was a restricted, inland water body that received an enormous yet highly variable supply of meltwater from the ice-sheet (Marshall and Clarke, 1999), it is not unreasonable to think that near-bottom salinity was low or even fresh in parts of the basin for much of the time, that the salinity front fluctuated significantly in response to astronomic (diurnal, seasonal) and episodic (flood) forcing16, and that Candona subtriangulata was simply able to colonize the seafloor as a result.
As a final note, the (apparent) absence of mud in the subaqueous-outwash fans is striking. However, several sundry observations suggest that mud should actually be the dominant grain- size in the esker system sensu lato. 1) Mud constitutes between 55–75% of the total load in alpine-glacier streams (Østrem, 1975; Hammer and Smith, 1983; Gurnell, 1987). 2) The distal portions of fan-lobes in deepwater turbidite systems—which are somewhat reasonable process- analogues for subaqueous-outwash fans—are invariably muddy, even if they are considered to be coarse-grained systems (e.g., Reading and Richards, 1994). 3) R-channels should be extremely efficient at transferring mud from source to sink—possibly up to 10 times more so than a fluvial system—because they lack floodplain-like sediment-storage sites17. It is therefore highly likely that the sandy subaqueous-outwash fans exposed in the pits correlate to mud distally. The question is, how much?
15 Brackish-water ostracodes are observed locally in post-glacial sediments in the Lake Ontario basin (L. Denis Delorme, personal communication, 2007). Unlike freshwater-ostracode eggs, marine-ostracode eggs cannot dry out. They can only migrate into water bodies that are ocean-connected. This suggests the Champlain Sea was likely once confluent with Lake Ontario (e.g., Sharpe, 1979).
16 The salinity front in estuaries/deltas commonly moves 10s to 100s of km seasonally (e.g., Shanley et al., 1992). In the Amazon River, it is pushed 150 km onto the shelf during high discharge (Geyer et al., 2004). Meltwater discharge into the Champlain Sea may have been similar to that of the Amazon (200,000 cubic metres per second; Marshall and Clarke, 1999), but would have been much more seasonal.
17 30–90% of a river’s sediment load can be trapped in floodplains (Goodbred and Kuehl, 1999). 26
4.2.6 Mud with minor sand near bottom and/or top (Champlain Sea deposits)
Champlain Sea mud buries the esker locally, and forms the surficial sediment unit throughout much of the study area. Based on a cursory examination of cores, as they were being collected, three units are identified within the mud succession (bottom to top): 1) rhythmically laminated mud and sand (“varves”), 2) massive mud, and 3) stratified mud, locally with sand layers near or at the top. These three units are commonly stacked on top of each other in “complete” Champlain Sea successions between Ottawa and Montreal (Gadd, 1961, 1986; Shilts, 1994; Aylsworth et al., 2003; Ross et al., 2006).
Where intersected in the new GSC cores, the rhythmically-laminated unit consists of thin (< 1 cm), alternating layers of light grey silt or very-fine sand and dark grey mud (Fig. 23). Couplets tend to be sharp-based and normally graded. Bioturbation levels are low to moderate. Rare dropstones are present. The unit fines upward, and reacts strongly with dilute HCl. It sharply overlies the esker. The unit is less than 2 metres thick, which is near the limit of resolution of the seismic data.
Previous studies identify a similar rhythmically-laminated unit in the same stratigraphic position throughout the western Champlain Sea basin (e.g., Gadd, 1986; Pair and Rodrigues, 1993). These studies suggest the rhythmites may contain only Candona subtriangulata (e.g., Pair and Rodriguez, 1994; Ross et al., 2006), only marine–brackish water fauna (e.g., Shilts, 1994; Ross et al., 2006), or a mix of both (Cronin, 1977; Hunt and Rathburne, 1988). Porewater salinities are commonly low to moderate and may gradually increase upward (e.g., Shilts, 1994; Torrance, 1988).
The rhythmically laminated unit passes gradationally upward into dark grey massive mud that is intensely bioturbated (Fig. 23). The unit commonly reacts strongly with HCl, but less so than underlying rhythmites. Portlandia arctic shells are common. Black, vertical to horizontal “squiggles” (0.1–2 mm wide, <1 cm long) are visible on freshly cut surfaces; these disappear after several hours of exposure to the atmosphere. Freshly cut surfaces have a subtle sulphurous odor. In nearby wells, porewater salinity in the massive-mud unit approaches that of the modern ocean (Fig. 13B; Torrance, 1988). Like the rhythmites it overlies, the massive-mud unit appears to be present throughout the western Champlain Sea basin (e.g., Gadd, 1986; Pair and Rodrigues, 1993).
A stratified-mud unit sharply overlies the massive-mud unit. In core, the unit consists predominantly of gradationally-based couplets that grade from light-grey mud to dark-grey or pinkish-red mud. Black residue, similar to that in the massive-mud unit, is observed in light-grey bands near the base of the succession, and also rarely occurs as discrete layers in pinkish-red bands. Gadd (1986) suggests the light-grey bands are coarser than the dark-grey/red bands. Couplets increase in thickness upward, from <1 cm near the base of the unit to several 10s of centimetres near the top of the unit. The thickest couplets in French Hill Road cores reach 35 cm thickness, whereas the thickest couplets in Route 300 cores are 15 cm thick. Successive light- grey mud bands may also become sandier upward. Shells are present, but very rare. Massive bioturbated intervals (<50 cm in thickness) are rarely observed near the base of the unit. In comparison to underlying mud units, the stratified-mud unit reacts less strongly with dilute HCl, and commonly not at all, with red layers reacting slightly more than light-grey layers. The upper 1–5 m below ground surface is orange-brown, is stiffer and dryer than underlying mud, and may contain root traces, joint-like structures, and very fine sand beds (<20 cm thick) that consist of stacked current-ripple cross-sets.
Previous studies identify a similar stratified-mud unit in the same stratigraphic position throughout the northern part of the western Champlain Sea basin (e.g., Gadd, 1961, 1986; Aylsworth et al. 2003; Ross et al. 2006). The unit is apparently absent in the south half of the basin (e.g., Pair and Rodgrigues, 1994). Based on numerous cores from various localities, Fransham and Gadd (1977) conclude that the stratified-mud unit is finer grained than the underlying massive-mud unit. 27
Figure 23. Representative cores of the buried Vars–Winchester esker from the north end (A) and south end (B) of the study area. Bedrock was not intersected in the GSC-BH-07-03 well, but we have a high degree of confidence, based on the Route 300 seismic data and depth-to-bedrock in a nearby well (GSC-BH-07-02), that the base of GSC-BH-07-03 is within a metre or two of bedrock. See text for details.
We now interpret the Champlain Sea succession using standard sequence-stratigraphic concepts and principles. Sequence stratigraphy is process-oriented sedimentology at the largest scale; its goal is to relate basin-scale patterns in sedimentary strata to long-term changes in two key variables, sediment supply (S) and accommodation space for sediment (A) (Curray, 1964; Vail et al. 1977; Posamentier and Vail, 1988; Van Wagoner, 1990; Posamentier and Allen, 1999). In non-glacial sedimentary systems, horizontal translation of the shoreline forced by sea-level change is the main process that mediates S and A over long (“Milankovitch-scale”) time scales. Shoreline movement is assumed to be slow and gradual. Glacial systems are different in two ways. First, the key interface that mediates gradual changes in S and A is the ice-margin, not the shoreline (although shoreline translation is also important). Second, rapid events (meltwater floods) are common, and generate rapid changes in S and A. These two characteristics—that ice-margin translation causes gradual change in S and A, and that meltwater events punctuate this gradual change—are believed to be the hallmarks of glacial sedimentary systems.
Although the “complete” Champlain Sea succession between Ottawa and Montreal consists of three main lithostratigraphic units (“varves”–massive mud–stratified mud; e.g., Gadd, 1961), we argue that the succession is best subdivided into two genetically-related sediment packages, one 28 deposited as the sediment source (the ice-margin) backstepped northward through and out of the Champlain Sea, and one deposited as a new sediment-source, the ice-distal, meltwater-fed shoreline, moved back into the basin as the result of isostatic rebound (Fig. 13D). It is possible that one or more rapid meltwater-events may have punctuated this gradual, ice-mediated sedimentation pattern.
1) Backstepping system
Initially, the ice was in contact with the Champlain Sea. Retreat of the ice front through the sea caused a huge jump in accommodation space when the environment switched from subglacial to proglacial18. Continued backstepping of the ice-margin gradually reduced the caliber and supply of sediment to the seafloor, with an associated reduction in the environmental stress on benthic organisms (e.g., salinity stress, turbidity). The sedimentary result of this gradual backstepping is the upward-fining succession that starts with the esker gravel19 and ends with the massive bioturbated mud. Backstepping of the ice through marine water adequately explains both the outcrop-scale characteristics of the lower mud-package (change in fauna, grain-size, bioturbation level, porewater salinity) and its regional-scale characteristics (deposition of a mud “blanket” throughout the western Champlain Sea, which would be difficult to do if the sediment source was stationary and far away; see Dalrymple and Cummings, 2005; Cummings et al. 2005).
2) Forestepping system
Meltwater flux into to the Champlain Sea is interpreted to have increased abruptly at some point during northward ice retreat, causing unbioturbated, red-and-grey stratified mud to onlap the massive, intensely bioturbated mud unit throughout the basin. Data in Ross et al. (2006) suggest this may have started around 11 14C kyr BP20. Thinning of the stratified-mud unit southward, and its apparent absence south of the St. Lawrence River (Pair and Rodrigues, 1993) suggests that sediment was likely supplied by rivers that drained into the northern Champlain Sea (e.g., Ottawa River, smaller rivers that drain the Laurentian Highlands). Forced regression as a result of isostatic rebound is interpreted to have generated upward-coarsening and upward-thickening trends in the unit. If each red-and-grey couplet is a varve (e.g., Gadd, 1986), the sedimentation rate was initially millimetres to centimetres per year, and increased with time, reaching ~15 cm/yr along Route 300 and ~35 cm/yr along French Hill Road. The low level of bioturbation and paucity of shells support the inference that sedimentation rate was high (MacEachern et al. 2005). Isolated sand beds in the stratified mud unit (< 1 m thick), some of which are associated with higher-amplitude seismic reflections (e.g., Route 300), may have been deposited during meltwater outbursts into the basin.
Gadd (1986) interprets the stratified-mud unit to be a deltaic deposit. This is indeed reasonable, given the upward-coarsening and upward-thickening trends. However, no obvious downlap is observed in any of the seismic transects. Rather, the stratified mud unit appears to onlap the massive mud unit (e.g., Fig. 4A in Ross et al. 2006). This would not be expected if the unit was deposited by a delta: deltaic deposits consist of clinoforms, albeit commonly very low angle ones (<1q) if the system is mud-rich (e.g., Dalrymple and Cummings, 2005; Cummings et al. 2005). It is possible that seismic transects are not long enough to image the low-angle clinoforms, or that they are oriented obliquely to the progradational axis of the delta(s). An alternative explanation,
18 There was no “transgression” in the standard (non-glacial) sense of the word. Rather, water depth was maximum immediately following ice retreat, then decreased with isostatic rebound. Technically, therefore, Champlain Sea sediments are all forced regressive (falling stage) deposits.
19 The gravely central-ridge may also simply have aggraded in a long R-channel without ice retreat, followed by blanketing of fan sands over the gravel ridge with ice retreat.
20 Dates from marine shells. 29 given the apparent onlap relationship (i.e., depressions appear to fill with horizontal strata from the bottom upward), is that the stratified-mud unit was deposited by dense, muddy underflows that hugged bathymetric lows, a style of sedimentation that commonly results in a ponded basin- fill (e.g., Vail et al., 1977). Resolution of this question may be possible following a more detailed analysis of the cores, and acquisition of regional seismic data.
4.2.7 Top of Champlain Sea deposits (modern landscape)
Where underlain by Champlain Sea deposits, the modern landscape tends to be relatively flat and nearly horizontal, except where streams incise the ground surface (Fig. 12). Two types of incision are observed, one associated with a break of slope, and one that is not.
The only incised valley21 that lacks an obvious downstream break in slope is the Ottawa River incised valley, within which the modern Ottawa River sits (Fig. 11). (Note that the St. Lawrence River, by contrast, is not incised upstream of Montreal.) The Ottawa River incised valley is 15–30 metres deep, 5–20 kilometres wide, and is anastomosed just downstream of a bedrock constriction at Ottawa. (The modern Ottawa River is an order-of-magnitude smaller.) The downflow extent of the incised valley past Montreal is unknown, but it can be traced 300 km upstream of Montreal to Fort Coulonge, and maintains a relatively constant cross-sectional area along this distance. (A general lack of surficial sediment over bedrock hinders its identification upstream of this.) It is carved primarily into Champlain Sea mud. Paleoislands in the anastomosed section are also composed primarily of mud, with bedrock protruding through locally. Deep, elongate, flow-parallel scours (up to 20 m relief) eroded into Champlain Sea mud are present on the floor of the modern river in a lacustrine-like reach just upstream of Ottawa (Fig. 25A; Shilts, 1994), and streamlined, flow-parallel sediment ridges are present on the incised valley floor just downstream of Ottawa (Fig. 12). Streamlined boulder ridges have been reported on the floor of the incised valley in Hull, just north of Ottawa, that erosively overlie mud with marine fossils, and contain angular upstream-imbricated limestone slabs up to 3 metres in length (Keele and Johnston, 1913). In general, however, large clasts and/or bedforms are absent on the floor of the incised valley. A basal date from the Mer Bleu bog suggests incision occurred prior to 7650 14C yr BP (GSC-681) (see Aylsworth et al., 2000 for sample location).
The second type of stream incision occurs where tributaries of the modern Ottawa River (e.g., South Nation and Rideau rivers) cross the edge of the Ottawa River incised valley (Figs. 11, 12). Depth of incision is greatest at the break in slope, and decreases gradually upstream. An incised tributary that crosses the centre of the Route 300 seismic line (the depression in the middle of Figure 20B) truncates near-horizontal Champlain Sea mud reflections; no channel deposits are observed.
Interfluves outside of the incised valleys and large portions of land within the incised valleys have undergone pedogenic alteration to a depth of one to several metres. Modern and historic landslides occur on the flanks of both types of incised valley. For most historic landslides, wood samples collected within or below the landslide material yield dates that cluster tightly around 4550 14C yr BP (Aylsworth et al., 2000).
The Ottawa River incised valley is huge: it is an order-of-magnitude larger than most modern rivers (only the Amazon comes close), and is similar in width to the Lake Missoula outburst-flood channel (Fig. 24). Given that no obvious break in slope occurs downstream, and that the St. Lawrence River shows no comparable incision, the Ottawa River incised valley did not likely form by knickpoint migration upstream from a slope-break or by entrenchment related to isostatic rebound. Rather, the valley likely formed when water discharge down the Ottawa River was
21 An incised valley, by definition, is a channel that never tops its banks. It therefore has interfluves, not floodplains. See Posamentier and Allen (1999). 30 greater than today. The Ottawa River has long been suspected as having acted as a major continent-to-ocean meltwater conduit during the last deglaciation (MacPherson, 1968; Broecker et al., 1989; Teller, 1988; Teller et al., 2004). Teller (1988) estimates early Holocene discharges of up to 200,000 cubic metres per second from Lake Agassiz outburst floods. (Modern discharge is ~2,000 cubic metres per second.) However, if the simple scaling relationship outlined in Figure 24 is assumed to apply, the channel-forming discharge may have been upwards of 800,000 cubic metres per second, almost four times that predicted by Teller (1988)22. In any case, the incised valley is very large, and discharge must have been accordingly very high. Incised tributary streams likely formed differently, by knickpoint nucleation at the edge of the Ottawa River incised valley (e.g., Leeder and Stewart, 1996)23. Radiocarbon dates indicate that most landslides postdated erosion of the Ottawa River incised valley by several thousand years. Aylsworth et al (2000) argue that the cluster of landslide dates around 4550 14C yr BP indicates an earthquake trigger.
With respect to the groundwater system, the Ottawa River incised valley may seem like a geological footnote. However, it is not: in at least one location (in between Watson Road pit and Highway 417), it appears to completely truncate the Vars–Winchester esker-aquifer. In closing, these observations serve to highlight our contention that groundwater-aquifers are best understood (modeled) when high-quality data are collected (i.e., cores, seismic, outcrop, in addition to water-well data) and used to develop a regional sequence-stratigraphic context that connects local observations (i.e., the aquifer) to the larger-scale groundwater picture.
Figure 24. Simplified cross-sections of large modern rivers versus that of the Ottawa River incised valley.
22 Similar values are obtained using the slope–area method of Dalrymple and Benson (1967) (Dmitri Ponomarkenko, personal communication, 2007)
23 Similar yet less pronounced knickpoint incision is visible where tributaries cross a break in slope and enter the modern Ottawa River. 31
5.0 Hydrogeological implications Don Cummings, Geological Survey of Canada and South Nation Conservation Authority
5.1 How does water get into the esker?
Water could enter the Vars–Winchester esker aquifer in several ways, including downward movement of water from precipitation or streams, horizontal movement of water from adjacent aquifers (e.g., fractured-bedrock aquifer, thin gravel-sheet over till), and upward movement from subjacent aquifers (e.g., deep bedrock). Several of these mechanisms are more likely than others, as discussed below.
Much of the precipitation that falls on the exposed esker likely percolates downward into the aquifer. Approximately 3 million square metres of the esker is exposed24. If all precipitation that falls on this surface is assumed to enter the aquifer (~350 mm/yr following evapotranspiration25), this mechanism could potentially account for ~1 billion litres of recharge per year (~280 million US gallons or 1 million cubic metres). This is equivalent to a constant influx of 0.03 cubic metres of water per second. Given that the baseflow in the Castor River, the major river that crosses the esker, is typically two order-of-magnitude higher than this (~1 m3/s) in late summer when evapotranspiration tends to exceed precipitation (see Environment Canada website, http://www.wsc.ec.gc.ca/staflo/index_e.cfm?cname=flow_daily .cfm), it is possible that other mechanisms in addition to direct precipitation operate to recharge the esker aquifer. This is potentially a common theme for buried esker-aquifers in mud-rich glaciated basins (e.g., Caswell, 1988b).
Figure 25. Example of an incised stream that intersects the buried esker in an elevated part of the study area (on a paleoisland in the Ottawa River incised valley, just north of the Regimbald Road pit—see Stop 1 in Section 6). (A) Champlain Sea mud is present along most of the stream bed. (B) Esker gravels crop out in the stream bed, however, just north of the Regimbald Road gravel pit.
24 This is a rough “order-of-magnitude” estimate based on air-photo interpretation (Dmitri Ponomarenko, personal communication, 2007), field observations, and the work of Gorrell (1991).
25 This is an inherently difficult value to predict—see Section 3 and Telmer and Veizer (2000). 32
Streams that cross the esker may also in part recharge the aquifer. Some streams have a direct physical connection, for example, where they dissect the exposed esker landform (e.g., the stream just south of Magladry Road) or incise through Champlain Sea mud into the buried esker26. An example of the latter is observed just north of the Regimbald Road pit (Fig. 25). There is anecdotal evidence that suggests water from this stream recharges the esker each spring. Installation of tile drains in adjacent fields increased runoff to the stream. The following spring, a nearby resident’s well flooded into this basement for the first time in 20 years, and has continued to do so in each subsequent year. His house is located obliquely downslope of the creek, and his well is drilled directly into the buried esker. This suggests that the stream recharges the esker at this location during the spring freshet.
It is unlikely that a significant amount of water percolates through the Champlain Sea mud and into the esker. The esker water is fresh, similar to that of the precipitation, whereas porewater in the mud is typically brackish, and can reach fully-marine salinities in some locations (e.g., Torrance, 1988). The fractured-bedrock aquifer is also an unlikely source of recharge since it contains water that is salty and hard relative to the water in the esker (Charron, 1978). Well- sorted sediment between till and bedrock or “sub-till sediment” (e.g., Nixon and Veillette, 1981; Ross et al. 2006), if present, could supply water to the esker. However, such sediment was not intersected during drilling, and is therefore not believed to be a significant component of the groundwater system in the study area (see Section 3.6).
The thin (<3 m) buried gravel unit that occurs over till in off-esker locations (Sharpe and Pugin, 2007) is somewhat of a wildcard in the groundwater system. A gravel unit, possibly sheet-like, was intersected in two off-esker wells along French Hill Road (GSC-BH-07-05 and -07), and appears to be present in uncored wells along Ball Road, Rang St. Thomas, Watson Road and Maple Ridge Road (data not shown). Furthermore, boulders are strewn locally over the ground where till forms the surficial sediment, and are interpreted to locally overlie till/bedrock in several seismic transects (Sharpe and Pugin, 2007). Drillers suggest the off-esker gravel is “thirsty” because mud circulation became considerably reduced (GSC-BH-07-07) or stopped altogether (Rang St. Thomas) when it was intersected. It is possible, though unproven, that precipitation falling on the bedrock highs could be readily transferred via the gravel unit to the subsurface and into the esker (Sharpe and Pugin, 2007). Similar gravel units in nearby locations (e.g., west of Montreal; Ross et al., 2006) may have a similar hydrogeological function. This issue warrants further investigation.
5.2 How does water flow through the esker?
The gravely central-ridge of the esker has such a high permeability relative to the surrounding strata that it likely functions as a groundwater “highway” (Fig. 13C). In outcrop and core, the gravely central-ridge is devoid of baffles (aquitards) that would greatly impede groundwater flow (e.g., areally-extensive mud layers), as may be typical of gravely central-ridges in eskers in general (e.g., Brennand, 2000). What is less certain is whether breaks occur along its length. At least one break has been identified where the Ottawa River incised valley has eroded through the esker and into till. Other breaks may occur where the esker is buried and no subsurface data are available. However, the gravely central-ridge is present in all seismic transects collected over the esker, which suggests that it may be largely or even completely connected where buried.
The sandy-fan carapace of the esker has a lower permeability (Fig. 13C). Significant mud-baffles were not observed, but areally-extensive silty-sand climbing-ripple beds (hydraulic conductivities of 10-5; see Stop 5 in Section 6) that occur at lobe-downlap surfaces may partially compartmentalize the fan. It is possible that the sandy-fan carapace functions as a storage site
26 We have only walked out several streams. All streams that cross the esker need to be walked out to determine potential windows between surface and ground water. 33 or “sponge” that feeds the gravely central-ridge “highway” when the groundwater table falls, for example, as a result of intense pumping (George Gorrell, personal communication, 2007).
5.3 How does water get out of the esker?
In addition to anthropogenic extraction, several pathways by which the esker aquifer is recharged may also serve as discharge pathways. Specifically, water could exit the esker and provide baseflow to streams that cross the esker, could travel into the fracture-bedrock aquifer and/or sub-till sediments, or could travel into the gravel-sheet that locally occurs over till.
The hypothesis that esker aquifers discharge into streams seems to have supporting evidence. For example, rivers continue to flow when evapotranspiration exceeds precipitation (Chin et al. 1980), indicating a groundwater supply. Also, a conductivity/temperature metre dragged along the bed of the East Castor River in the summer identified an anomaly where the river crossed the esker (Fig. 26; Bustros-Lussier, 2006). Other pathways by which water leaves the esker (e.g., into the fractured-bedrock aquifer) are suspected to exist. Water chemistry data that are currently being collected will help test this hypothesis.
Figure 26. Example of two streams in a low-lying part of the study area into which groundwater from the esker-aquifer potentially discharges. Note that no temperature/conductivity anomaly was detected along the Castor River (Bustros-Lussier, 2006), likely because the esker is buried by Champlain Sea mud at this location (see sub-bottom profile). Electrical conductivity and temperature data are simplified from Bustros-Lussier (2006). Sub-bottom profile is courtesy of Marten Douma (Geological Survey of Canada). 34
6.0 Fieldtrip Stops
Stop 1. Seismic Section French Hill Rd East Pugin, André, Geological Survey of Canada
Objective:
Stop 1 provides an overview of shear-wave (SH-wave) reflection seismic technique applied in the South Nation study by the GSC. Data collection was initiated to better delineate the location of a buried segment of the Vars–Winchester esker.
Setting Overview:
Data collection was completed in an upland region that consists of bedrock outcrop, low relief till mounds, faint esker relief and extensive low relief muddy sediment. Facing the Ottawa River the upland has an escarpment of Paleozoic bedrock. The eastern flank of the upland is bordered by a paleo-channel and the bank consists of sandy surficial sediment. Water wells in the upland penetrate to bedrock, are locally flowing, and have yields of up to 200 gallons per minute. The upland forms a paleo-island in the incised-valley of the Ottawa River.
Description:
The ~2750 m long, east – west seismic line images bedrock and a number of seismic facies within the overlying surficial sediment secession. Boreholes drilled on the section (GSC-BH-07- 07 and GSC-BH-07-06) provide ground truth for the seismic interpretation. The section consists of three distinct entities: I) a shallow western portion of 1800 m length, ii) a 500 m wide bedrock valley, iii) an eastern portion of shallower bedrock (Fig. 30).
Bedrock signature: The bedrock interface slopes gently eastward from the beginning of the line to 1800 m. At this location the bedrock descends abruptly forming the western margin of an asymmetrical bedrock valley of 500 m width with a deep 200 m wide western trough and a 300 m wide shallower eastern shoulder. The valley has 80 m of relief along the axial trough where the bedrock reflection is weak to absent.
Till on Bedrock: Overlying the limestone bedrock there is a boulder lag and/or a 2–3 m thick till layer. The boulder lag is interpreted on the basis of parabolic refractions. Strong parallel reflections are interpreted to be produced from the top of till and the top of the underlying limestone bedrock.
Valley Fill: Fill of the buried bedrock valley consists of an 80 m thick succession characterized by a diffractive chaotic seismic facies. The shallower part is less chaotic, with discontinuous reflections forming a large trough shaped feature. The chaotic facies may be a signature of disturbed sediment that has been subject to liquefaction triggered by earthquake events. Similar signatures have been described elsewhere in the Ottawa basin. These disturbances occur when earthquake energy is amplified by bedrock depressions filled by thick basin mud sequences (Aylsworth et al. 2000).
Basin Mud: The lower basin mud is characterized by continuous reflections that can be subdivided into two subunits separated by a strong reflection. The two units have different velocity/density characteristics. Based on surface seismic data. the lower "Basin Mud I" unit has an average interval SH-wave velocity of ~160 m/s and is more transparent, whereas the upper more reflective unit "Basin Mud II" has an average interval velocity of 125 m/s. More definitive velocity measurements will come from geophysical borehole logging surveys. 35
Figure 27. Seismic reflection section from the east-end of French-Hill road. Note the buried bedrock valley and the blanketing marine sediment that forms a low relief landscape.
Discussion:
The seismic profile provides information on a number of points on the geological history of the area that are also of hydrogeological significance (Fig. 27). The most unexpected development was the discovery of a >80 m deep bedrock valley. The 500 m wide valley appears to be intercepted in a number of deep boreholes near the Ottawa River and to the south of the section. From this sparse borehole control, the valley is interpreted to extend north - south with a slight sinuosity in its course. Based on the seismic facies of the valley fill and the interpreted stratigraphic context, the feature is interpreted as tunnel valley that likely formed by the reoccupation of an existing bedrock valley. Gravel mounds on the eastern end of the seismic section are interpreted to be the Vars – Winchester esker. The esker appears to intersect the valley where it curves eastward. Immediately to the south of the valley, the esker is exposed in a number of small aggregate pits, two of which form subsequent stops. 36
Stop 2. Regimbald Road pit Don Cummings, Geological Survey of Canada and South Nation Conservation Authority
Objective Observe and interpret the sedimentology of the esker where it is narrow and composed entirely of gravel.
Setting The pit is excavated into a tributary-like arm of the esker (Fig. 12). At this location, the esker is partially overlain by Champlain Sea mud. In a nearby cored well, the esker rests directly on carbonate-mudstone bedrock.
Description
The coarsest (and presumably most proximal) sediment in the esker is exposed in the Regimbald Road pit. In the exposure, the gravel is organized into thick (1.5–2.5 m) dune cross-stratified beds that are stacked on top of each other (Fig. 28). Cross-strata dip at a high-angle (25–30q) towards the south, which is parallel to the esker long-axis. Clasts are well rounded and are typically pebble to cobble size, although boulders up to ~1 metre in diametre are present. Clast lithology is similar to that of the adjacent till: carbonate mudstone predominates, with minor percentages of igneous (granite), sandstone and shale. Within the cross-sets, cyclicity is observed on both a centimetre-scale (alternation of sandy cross-strata with gravely cross-strata; Fig. 29) and on a metre-scale (several reactivation-surface bound, downflow-fining packages of cross-strata within a single cross-set).
Figure 28. One of several thick, stacked, high-angle (dune) cross-sets exposed in the Regimbald Road pit. See Figure 12 for pit location. Note the reactivation surface separating sandy cross-strata from cobble cross-strata. Cross-strata dip roughly southward, parallel to the long-axis of the esker. Dmitri “Kung-fu” Ponomarenko for scale (~2 m height). 37
Figure 29. Close-up of centimetre-scale rhythmicity in the high-angle (dune) cross-strata.
Discussion x What was the source of the gravel—how did it “get into” the esker? x Are these subglacial deposits? x Why is the sandy-fan carapace absent? x What is the significance of the two scales of rhythmicity? x How deep was the flow? x How fast was the flow? x How long did it take to deposit the esker at this location? Days? Years? 100s of years? x With respect to the groundwater system, are we justified in upscaling observations from this pit to the whole esker-aquifer? 38
Stop 3. Watson Road pit Don Cummings, Geological Survey of Canada and South Nation Conservation Authority
Objective Observe and interpret the sedimentology and stratigraphy of the esker where it is wider, composed of a gravely central-ridge with a broad sandy-fan carapace, and partially buried by fossiliferous mud.
Setting The Watson Road pit is located immediately downflow of an apparent confluence between two tributary-like esker ridges (Fig. 12). The esker doubles in width at this location. It is flanked by Champlain Sea mud and crops out along an axial zone of 100–200 m width.
Figure 30. West flank of the esker ridge exposed in cross-section at the Watson Road pit. Note that succession fines up from cobble gravel (gravely central-ridge?) to sand and gravel (subaqueous-outwash fan) to Champlain Sea mud.
Description
The Watson Road pit is the only outcrop where all components of the esker system are exposed—the gravely central-ridge, the sandy-fan carapace and the mud that overlies the esker (Fig. 30). 39
At the base of the western side of the pit, a poorly-exposed, mound-like cobble gravel unit is exposed. Given its coarse grain-size and mound-like morphology, it is interpreted to be the gravely central-ridge, or possibly the most proximal portion of the subaqueous-outwash fan.
The sandy-fan carapace sharply overlies the gravely central-ridge. In the eastern half of the pit, the fan is well exposed, and is composed of mound-shaped, upward-coarsening sand and gravel units that are stacked compensationally on each other (Fig. 31). Sharp-based beds, commonly gravely, interrupt this motif locally. Climbing current-ripples and diffusely-laminated sand beds are common, suggesting deposition from tractive unidirectional flows with abundant sediment rain-out from suspension.
Champlain Sea mud overlies the sandy-fan carapace (Fig. 30). Two mud units are recognized. The lowermost unit is brownish, and contains numerous sand beds, some rich in marine shells (mostly Hiatella arctica). This unit downlaps the sandy-fan carapace at a very low angle; its clinoforms prograde outward at a normal/oblique angle from the crest of the landform. The second, overlying mud unit consists of red-and-grey laminated mud. Where exposed, its lower contact is sharp and truncates the underlying mud unit at a low angle. Strata are highly folded and contorted close to the crest of the landform, but undeformed off the crest of the landform. Shells are less common in this unit. A thin layer of gravel caps the succession.
Figure 31. East flank of the esker ridge exposed in cross-section at the Watson Road pit. Paleoflows are out of the page (southward). Note that the sandy fan carapace here is composed of several sharp-based, mound-shaped, upward-coarsening units (numbered 1 to 4). 40
Discussion x Why is sandy-fan carapace present here, but not at previous stop? x Note that sandy-fan carapace is composed of smaller, upward-coarsening, mound- shaped architectural elements. What are they? How did they form? x How was downlap relationship between Champlain Sea mud and esker generated? x Note that two mud units exist, a lower sandy/shelly one and an upper, locally deformed, shell-less, red-and-grey banded one. Why two units? Why are they so different? Why is their contact sharp? x What succession of events deposited the observed stratigraphy, taking into account the nature of the contacts between and within them? x Does water percolate through the mud into the esker aquifer? Does this mud layer prevent recharge? If so, how does water get into esker? 41
Stop 4. Route 300 seismic profiles and cores Susan Pullan and André Pugin, Geological Survey of Canada Don Cummings, Geological Survey of Canada and South Nation Conservation Authority
Objective
Stop 4 provides an opportunity to discuss and interpret compressional-wave (P-wave) and horizontally polarized shear-wave (SH-wave) seismic profiles collected along Route 300 (Fig. 12), and to compare these data with two continuous cores. Acquisition and processing methods, and the difference between P- and SH-wave data, will be outlined.
Setting Overview
Route 300 crosses an area of flat, low-relief agricultural land where the esker is completely buried by Champlain Sea mud.
Description
In 2006, the Geological Survey of Canada collected approximately 20 line-km of shallow seismic reflection data in order to delineate the three-dimensional structure of the buried Vars–Winchester esker-aquifer and the surrounding stratigraphy. P-wave data were collected using planted geophones and an in-hole shotgun source. SH-wave data were collected using a landstreamer and minivib source. Additional information on seismic reflection profiling and the minivib- landstreamer system can be found in the methodology section of this field guide.
Along this road (Route 300), both P- and SH- wave profiles have been acquired, and clearly delineate the buried esker and its cross-sectional architecture (Fig 32). Both sections show the relatively flat-lying, fine-grained Champlain Sea deposits overlying the lower-frequency and less- coherent reflections related to the esker deposits. The esker deposits are observed to be at least 20–25 m thick at the crest, and the central core of the esker is ~200 m wide. The flanks extend an additional >200 m on each side. Both sections display some evidence of “disturbed” (sandier?) sediment directly above the esker which suggests that there may be enhanced hydraulic connection between the surface and the esker in this region.
Seismic stratigraphy
Three main reflection-packages are observed in the Route 300 seismic profiles. 1) At the base of the profiles, a package of high- to moderate-amplitude reflections is observed at ~30–35 m depth (~35 metres above sea level). This represents the deepest interpretable seismic signal on the record. 2) A mound-shaped reflection package is present above this. It is relatively symmetric, ~800 m wide and ~20 m thick at its apex. Its flanks dip at low angles (2–5q). Reflections within the mound are generally low–moderate amplitude, often discontinuous, and somewhat chaotic. 3) At the top of the profiles, a package of low-amplitude, continuous, nearly-horizontal reflections is present (Fig. 32). The package varies in thickness from ~35 m off the mound to ~12 m over the crest of the mound. Reflections onlap the mound-shaped unit. A stream that crosses Route 300 (at break in P-wave profile – Fig. 32A, and at common mid-point (CMP) 640 - Fig. 32B) truncates reflections at the top of this unit. At ~10–14 m depth (~55 metres above sea level), a higher-amplitude reflection is observed is observed within this unit. This reflection parallels overlying and underlying reflections. 42
Correlation of core and seismic data
The three seismic units described above correlate with carbonate-mudstone bedrock (or bedrock plus a thin overlying till), the buried Vars–Winchester esker, and Champlain Sea mud, respectively. High-amplitude reflections between seismic units are generated by the till or bedrock surface and by the top of the esker. Stiff diamicton (till) was intersected by the borehole midway along the esker flank but not beneath the esker crest. However, given the unit’s thinness (less than several metres), this reflection is not easy to differentiate from the bedrock reflection at this location. The shallow higher-amplitude reflection within the Champlain Sea deposits is generated at a mud-on-mud contact (see below).
The seismic data clearly delineate the esker surface, and provides some information about its internal architecture (Figs. 32, 33). Core data indicate that the esker consists of two elements at this location, a gravely central-ridge and a broad sandy carapace. The gravely central-ridge is 12 m thick in the borehole near the esker crest. The seismic data suggest that it may extend laterally >100 m, though its top surface does not generate a distinct, coherent reflection. Clasts observed are predominantly carbonate-mudstone pebbles and cobbles, with minor percentages of granite and sandstone. A subtle fining-upward trend is observed in the uppermost several metres, which may help explain the lack of a distinct seismic reflection. The gravely central-ridge directly overlies bedrock in the borehole drilled at the esker crest, whereas in the borehole along the esker flank it overlies a thin till-sheet (Fig. 23B). (The latter core did not extend to bedrock, but based on the seismic section, the till unit is likely to be very thin.) The cores indicate that the gravely central-ridge is sharply overlain by the sandy carapace, which is 7 m thick where it overlies the crest of the gravely central-ridge. Its top contact generates the high-amplitude, mound-shaped seismic reflection. Marine shells (mostly Portlandia arctica) are observed 2 metres and 11 metres below the top of the sandy carapace.
Champlain Sea mud that overlies the esker can be subdivided into three lithostratigraphic units, 1) moderately bioturbated, upward-fining rhythmites, which pass gradationally upward into 2) massive bioturbated mud, which is in turn sharply overlain by 3) upward-coarsening red-and-grey stratified mud (Fig. 23B). The higher-amplitude reflection observed in the seismic data apparently correlates to a sand bed in the stratified-mud unit. Stratified mud overlying the sand bed is soft (“buttery”) relative to the stratified-mud below the sand bed.
Acquisition and processing
Figure 33 demonstrates the effect and importance of the final processing steps (migration, topographic correction and depth conversion), using the SH-wave profile as an example. Topographic corrections related to a small creek crossing are significant (note the flattening of shallow reflectors in Fig. 33c).
Significant differences exist between the P- and SH-wave sections, and are interesting to note. The SH-wave data have a significantly higher vertical resolution, particularly in the shallow subsurface (0–20 m depth; see Fig. 32b). Continuous, coherent reflections are observed in the SH-wave data at depths of less than 5 m, whereas the P-wave section contains little information in this depth range (in part this is a limitation of the larger geophone spacing and source offsets used in P-wave survey). Deeper in the section, the P-wave data show a more coherent and higher-amplitude reflection from the interpreted bedrock surface at ~30–35 masl, while the SH- wave data show significant reduction in reflected signal below coarse-grained units (see Fig. 32b, CMPs 450–800, where bedrock reflection is very weak). As well, there are significant differences between the P- and SH-data in the relative amplitudes of some reflectors; for example, the amplitude of the esker surface along the flanks of the feature remains high in the case of P-waves (see Fig. 32a, CMPs 150–250 on west end of profile and CMPs 350–500 on east end), whereas the amplitude of this reflection drops considerably on the SH-section (see Fig. 32b, CMPs 150– 350, 800–1200). These differences may be significant in terms of understanding the lithological 43 and geotechnical information that can be gleaned from the seismic data, and need to be further investigated through borehole sampling and logging.
Conclusion
Seismic reflection techniques have produced excellent high-resolution data that have significantly improved the understanding of the cross-section form of the buried eskers and surrounding basin stratigraphy. The new knowledge on esker dimensions permits better estimation of the aquifer scale and groundwater reserves. The seismic facies observed also provide an indication of the extent of a coarse-grained esker core, of the flanking esker sands, and of “mixed” zone above the esker crest. The SH-wave landstreamer-minivib system yielded higher-resolution data and has the added advantages of much faster data acquisition (2–3 times the data acquisition rate of the P-wave system), and fewer data gaps in survey lines (P-wave data cannot be obtained where shotholes cannot be drilled or where geophones cannot be planted; e.g. road/driveway/creek crossings, buried utilites etc.). The P-wave system provided better definition of the bedrock surface, as SH-waves do not seem to penetrate through overlying coarse-grained units (gravel or coarse-grained till?). However, the differences in reflection character between the P- and SH- data may provide important information related to the lithology or physical properties of the subsurface sediment. Further investigation and integration of groundtruth data, continuous core and downhole geophysics are required to understand these differences.
Discussion x Why is till absent beneath the gravely central-ridge? x What is the significance of marine shells in the sandy carapace of the esker? x What is the significance of the high-amplitude reflection within the Champlain Sea mud package?
Additional reading:
Pullan, S.E., Pugin, A.J-M, and Hunter, J.A., 2007. Shallow seismic reflection methods for the delineation and hydrogeological characterization of buried eskers in Eastern Ontario. In Proceedings, SAGEEP’07 (Symposium on the Application of Geophysics to Engineering and Environmental Problems), April 1-5, 2007, Denver, CO, CD-ROM edition, 9 p. 44
Figure 32. Comparison of a) P-wave and b) SH-wave processed seismic reflection sections (vert. exg.= 3x) over buried esker north of Embrun, Eastern Ontario. See discussion in text. The locations of the two boreholes drilled along this line in March 2007 are indicated by the red lines. c) Direct side-by-side comparion of short sections extracted from a) and b) showing higher resolution obtained with SH-waves, particularly in the upper 10–20 m. 45
Figure 33. Processed SH-wave seismic reflection section over buried esker: a) amplitude section in two-way travel time; b) section a after phase-shift migration; c) section b converted to depth and with topographic corrections (vert. exg.= 3x); d) interpreted section. 46
Stop 5. Kemptville/Loughlin Ridge esker – subaqueous fan Gorrell, George; Gorrell Resource Investigations Russell, Hazen; Geological Survey of Canada Cummings, Don; Geological Survey of Canada and South Nation Conservation Authority
Objective:
Stop 5 provides an overview of a much larger subaqueous fan deposit associated with the esker west of the Vars – Winchester esker, and provides insight into the vertical succession and lateral continuity of the sedimentary facies (Fig. 12). Deposits at the two sites (Lafarge Pit and Tackaberry Pit) demonstrate the contrast in sediment caliber and hence hydraulic conductivity between the proximal fan setting and the more distal mid fan. The fan at this site is likely an analogue for the similar sized but less extensively exposed deposit at the southern terminus of the Vars – Winchester esker at Maple Ridge.
Setting Overview:
The ridge is mapped as ice-contact stratified drift with flanking littoral sand (Richard, 1974). The deposit is part of an esker extending from the St Lawrence River northward to the Ottawa International Airport. Parts of this esker have been studied by investigators interested in esker – subaqueous fan sedimentology (Rust and Romanelli, 1975, Rust, 1977, 1988; Cheel and Rust, 1986; Gorrell, 1991; Gorrell and Shaw, 1991).
The subaqueous-fan exposed at Stop 5 is the source of ~40% of the concrete sand in the Ottawa area. To wash the sand, 100s if not 1000s of gallons per minute of groundwater are used.
Description:
Tackaberry Pit: This extensively worked pit currently has a ~ 500 m long, 6 m high, north trending face along the western margin of the operation. More than 80 % of the face was free of slump in the summer of 2006. The principal architectural elements are tabular bedsets of diffusely graded and plane-bed medium sand overlain by cross-stratified medium sand and small-scale cross-stratified fine sand. Locally scours of diffusely graded sand are prominent. Cross-strata have climbing forms at both the ripple- and dune- scale. A general fining trend can be identified from exposure in the northeast part of the pit to the southwest. This primary architecture has been modified locally by deformation features that include convolute bedding, load structures and fluidization pillars (Fig. 34). The underlying silty sand unit is commonly intact; however, local rupture and disturbance of the underlying unit can be observed (Fig. 35). Deformation is concentrated in the south part of the section, immediately next to the road, over lateral distance of 50 to 75 m.
Discussion:
Scours and fills: Cheel and Rust (1982) describe three styles of scour fills from a nearby subaqueous fan in the Stittsville area: i) horizontally stratified medium sand with imbricate pebbles and cobbles along the base and on bedding planes within the fill; (ii) massive sands; and (iii) horizontally stratified sands. Documented scours in other Champlain Sea fans are up to 10 m deep and 40 m across (e.g. Rust, 1977; Burbidge and Rust, 1988). In the Tackaberry pit, scours are generally relatively small, < 2 m deep and < 10 m across. Fills consist of diffusely graded fine - medium sand and plane-bed. In the western section no pebbles are observed; however, more proximally, fills are more texturally diverse. Generally some degree of faint stratification is present with massive fills being rare. 47
Figure 34. Representative images of the western exposure in the Tackaberry Pit at Loughlin Ridge. A) Convolute bedding with diaperic structures. B) Typical stratigraphic section of i) dune-scale cross-stratification, ii) climbing ripple-scale cross-lamination, iii) low-angle and dune-scale cross-stratification, iv) scours with diffusely graded fills. C) Toe-set climbing ripples with isolated mud laminae. K values are in m sec-1, and were calculated from grain size data using the Hazen method (Freeze and Cherry, 1979). Using D10 of the fine fraction. Dime scale is 1.8 cm. Metre stick in (A) and (B) is 110 cm long. 48
Figure 35. A) Section of diffusely graded sand overlain by climbing dune-scale cross-stratification and climbing ripple-scale cross-stratification. Sample locations indicated with bars. B) Close-up of diffusely graded sand fining upward to plane-bedded sand. C) Microscopic image of sediment peel from diffusely graded sand. Note distinct graded bands of grading and heavy mineral concentration. D) Grain size data for sample locations shown in (A). E) Histogram of grain size distribution of a sample of diffusely graded sand. All K values are in m sec-1, and were calculated from grain size data using the Hazen method (Freeze and Cherry, 1979). 49
The predominant fill is diffusely graded fine and medium sand that can locally be traced laterally and/or vertically to plane bed. Previous authors have attributed the scour and fills to slumping (e.g. Rust 1977; Cheel and Rust, 1982) or alternatively to scour beneath hydraulic jumps (Gorrell and Shaw, 1991). The scours and fills at the Tackaberry pit are interpreted to have formed where large vortices within a hydraulic jump impinged on the basin floor (e.g. Long et al. 1990). The absence of slump scars, the low depositional slope, and the upward gradation from diffusely graded to plane bedded sand suggest the scour fills were deposited from sediment laden, turbulent flows rather than from laminar, slump generated sandy debris flows. In flume experiments Leclaire and Arnott (2005) found that parallel lamination formed at bed-aggradation rates up to 4 mm s-1 and bedload-layer sediment concentration up to 0.36. Consequently, this suggests diffusely graded sediment is deposited under even higher aggradation rates and sediment concentrations.
Climbing Bed-Forms: Climbing small-scale cross-stratification is a ubiquitous element of many glacilacustrine and subaqueous fan deposits (e.g Jopling and Walker, 1968). Climbing dune-scale cross-stratification is less commonly reported. The development of this scale of climbing cross-strata further documents the high-rates of both suspension and bedload sedimentation in the subaqueous fan environment.
Deformation: Deformation features have been well described by Cheel and Rust (1986) from subaqueous fan deposits in the nearby Stittsville area (Sharpe, 1988). They place the deformed sediment into a slump facies that includes convolute bedding, ball and pillow structures and dish structures. Deformation is attributed to two mechanisms i) rotational slumping, and ii) dewatering or fluidization due to melt-out of underlying buried ice. In the Tackeberry pit, convolute bedding is interpreted to be related to dewatering of underlying units as a result of high sedimentation rates.
Groundwater Implication: Based on a small number of samples, average hydraulic conductivity (K) values are 10-3 to 10-5 m sec-1 for diffusely graded sand, dune-scale cross-stratified fine sand, and ripple-scale cross-laminated fine sand (Fig. 35). Local mud laminae are less permeable with K values of 10-8 m sec-1. Mud laminae in the mid fan setting are likely relatively discontinuous. The sandy fan may therefore provide significant reservoir capacity for coarser gravel deposits inferred to be buried beneath the fan, as they are observed to be elsewhere in Champlain Sea eskers (e.g. Route 300).
Depositional Model: Sediment exposed in the Tackaberry pit is interpreted to have been deposited on a subaqueous fan beneath a jet downflow of the hydraulic-jump (Fig. 36). Hydraulic jumps are characterized by high suspended sediment loads and intense turbulent fluidal flow. Depositional evidence for this is provided by diffusely graded sand and climbing dune- and ripple- scale cross- stratification. Diffusely graded sand may occur from hyperconcentrated flows, which may form as part of a basal layer within a thicker flow (Pierson and Scott, 1985; Best, 1992; Russell and Arnott, 2003). Climbing bedforms have been reported from flume studies in association with hydraulic jumps (Daub, 1996). Additional work needs to be completed at the Tackaberry pit to identify whether the diffusely graded sand facies bedding is from antidune stratification or not. Improved knowledge on this element of the bedding would further constrain the depositional model. 50
Figure 36. Depositional facies model of a grounding line fan with differential ice contact with the bed and lateral breaching of the conduit (Gorrell and Shaw, 1991). Sediment images in figures 48 and 49 illustrate the character of the central part of the model. 51
7.0 References
Aitken, A.E. and Gilbert, R. 1996. Marine mollusca from Expedition Fiord, Western Axel Heiberg Island, Northwest Territories, Canada. Arctic, v. 49(1), p. 29–43.
Anderson, T.W., Mott, R.J. and Delorme, L.D. 1985. Evidence for a pre-Champlain Sea glacial lake phase in Ottawa Valley, Ontario, and its implications in Current Research, Part A, Geological Survey of Canada Paper 85–1A, p. 239–245.
Arnott, R.W.C., and Hand, B.M., 1989, Bedforms, primary structures and grain fabric in the presence of suspended sediment rain: Journal of Sedimentary Petrology, v. 59, p. 1062-1069.
Artimo, A., Makinen, J., Berg, R.C., Abert, C.C. and Asalonen, V.-P. 2003. Three-dimensional geologic modeling and visualization of the Virttaankangas aquifer, southwestern Finland. Hydrogeology Journal, 11, 378-386.
Aylsworth, J.M, Hunter, J.A, Gorrell, G.A., and Good, R. 2003. Borehole investigation of earthquake- induced soft-sediment deformation at Lefaivre; A contribution to the Ottawa Valley Landslide Project. 3rd Canadian Conference on Geotechnique and Natural Hazards, Edmonton, Alberta, Canada. June 9–10.
Aylsworth, J.M., Lawrence, D.E. and Guertin, J., 2000, Did two massive earthquakes in the Holocene induce widespread landsliding and near-surface deformation in part of the Ottawa Valley, Canada? Geology, v. 28, pp. 903-906.
Best, J.L. 1992. Sedimentology and event timing of a catastrophic volvaniclastic mass flow, Volcan Hudson, southern Chile. Bulletin of Volcanology, 54, 299-318.
Bolduc, A., Paradis, S.J., Riverin, M.-N., Lefebvre, R. and Michaud, Y. 2006. A 3D Esker Geomodel For Groundwater Research: The Case Of The Saint-Mathieu - Berry Esker, Abitibi, Québec, Canada. In: Three-Dimensional Geologic Mapping for Groundwater Applications Workshop Extended Abstracts October 15, 2005 (Eds H.A.J. Russell, R.C. Berg and H.L. Thorleifson). Geological Survey of Canada, Open File 5048.
Brennand, T.A. 2000. Deglacial meltwater drainage and glaciodynamics: inferences from Laurentide eskers, Canada. Geomorphology, 32, 263-293.
Broecker, W.S., Kennett, J.P., Flower, B.P., Teller, J.T., Trumbore, S., Bonani, G. and Wolfli, W. 1989. Routing of meltwater from the Laurentide ice sheet during the Younger Dryas cold episode. Nature, v. 341, p. 318–321.
Burbidge, G.H. and Rust, B.R. 1988. A Champlain sea subwash fan at St Lazare, Quebec. In: The Late Quaternary Development of the Champlain Sea Basin (Ed N.R. Gadd), Special Paper, 35, pp. 47-61. Geological Association of Canada, St. John's, Newfoundland.
Bustros-Lussier, E. 2006.Identification of groundwater discharge in rivers using electrical conductivity. Unpublished technical report submitted to the Geological Survey of Canada, 24 p.
Caswell, B. 1988a. Time-of-travel in glacial aquifers. Water Well Journal, v. 42, p. 48–51.
Caswell, B. 1988b. Esker aquifers. Water Well Journal, v. 42, p. 36–-37.
Caswell, B. 1989. Realities of Wellhead Protection. Water Well Journal, v. 43, p. 36–39. 52
Charron, J.E. 1978. Hydrochemical study of groundwater flow in the interstream area between the Ottawa and St. Lawrence Rivers. Scientific Series No. 76. Inland Waters Directorate, Water Resources Branch, Ottawa, Canada.
Chebotarev, I.I. 1955. Metamorphism of natural waters in the crust of weathering. Parts 1, 2, 3. Geochim. Cosmochim. Acta 8:22-48, v. 137–170, p. 198–212.
CH2MHill, 2001. Eastern Ontario Water Resources Management Study (EOWRMS). Final Report, March 2001.
Cheel, R.J. and Rust, B.R. 1982. Coarse grained facies of glacio-marine deposits near Ottawa, Canada. In: Research in Glacial, Glacio-fluvial, and Glacio-lacustrine Systems (Eds A.R. Davidson, W. Nickling and B.D. Fahey), Proceedings - Guelph Symposium on Geomorphology, v. 6, p. 279-295. University of Guelph, Guelph, On, Canada.
Cheel, R.J., and Rust, B.R., 1986, A sequence of soft-sediment deformation (dewatering) structures in late Quaternary subaqueous outwash near Ottawa, Canada: Sedimentary Geology, v. 47, p. 77- 93.
Chin, V.I., Wang, K.T. and Vallery, D.J. 1980. Water resources of the South Nation River basin— Summary. Ontario Ministry of the Environment Water Resources Report 13, 7 p.
Clarke, P.U. and Karrow, P.F. 1984. Late Pleistocene water bodies in the St. Lawrence Lowland, New York, and regional correlations. Geological Society of America Bulletin, v. 95, p. 805–813.
Costa, J.E. 1983. Paleohydraulic reconstruction of flash-flood peaks from boulder deposits in the Colorado Front Range. Geological Society of America Bulletin, v. 94, p. 986–1004.
Cronin, T.M., 1977. Champlain Sea Foraminifera and Ostracoda: A systematic and paleoecological synthesis. Géographie physique et Quaternaire, v. 31, p. 107–122.
Cummings, D.I. and Arnott, R.W.C. 2005. Growth-faulted shelf-margin deltas: A new (but old) play type, offshore Nova Scotia. Bulletin of Canadian Petroleum Geology, v. 53(3), p. 211–236.
Cummings, D.I. and Occhietti, S. 2001. Late Wisconsinan sedimentation in the Quebec City region: evidence for energetic subaqueous-fan deposition during initial deglaciation. Géographie Physique et Quaternaire, v. 55, p. 257–273.
Cummings, D.I., Dalrymple, R.W. and Arnott, R.W.C. 2005. Why does the Naskapi mudstone “blanket” prograde the paleo–shelf-edge offshore Nova Scotia? AAPG Annual Meeting, Calgary, June 19– 22.
Curray, J.R. 1964. Transgressions and regressions in Miller, R.L. (ed.), Papers in Marine Geology. Macmillan, p. 175–203.
Dalrymple, R.W. and Cummings, D.I. 2005. The offshore transport of mud: Why it doesn’t happen and the stratigraphic implications. AAPG Annual Conference, June 18–24, Calgary, Canada.
Daub, G.A., 1996, Experimental study of deep-sea fan evolution and sedimentology [unpublished MSc thesis]: Colorado State University, Fort Collins, Colorado, 165 p.
Dix, R.G., Salad Hersi, O., Molgat, M., and Arnott, W. 1997. Lithostratigraphy and sequence stratigraphy of the Lower Paleozoic succession in the Ottawa Valley. Field Guide-book (A2). Geological Association of Canada - Mineralogical Association of Canada (GAC-MAC) Annual Meeting, Ottawa, Ontario, Canada. 53
Douma, M. and Nixon, F.M. 1993. Geophysical characterization of glacial and postglacial sediments in a continuously cored borehole near Ottawa, Ontario in Current Research, Part E, Geological Survey of Canada Paper 93-1E, p. 275–279.
Eyles, N., Clark, B.M., Kaye, B.G., Howard, K.W.F. and Eyles, C.H., 1985. The application of basin analysis techniques to glaciated terrains; an example from the Lake Ontario Basin, Canada. Geoscience Canada, 12(1): 22-32.
Fischer, H.B., List, J.E., Koh, C.R., Imberger, J. and Brooks, N.H. 1979. Mixing in Inland and Coastal Waters. Academic Press, 302 p.
Fransham, P.B. and Gadd, N.R. 1977. Geological and geomorphological controls of landslides in Ottawa valley, Ontario. Canadian Geotechnical Journal, v. 14(4), p. 531–539.
Freeze, R.A. and J.A. Cherry. 1979. Groundwater. Prentice Hall, 604p.
Gadd, N.R. 1961. Surficial geology of the Ottawa area, report of progress. Geological Survey of Canada Paper 61–19.
Gadd, N.R. 1986. Lithofacies of Leda clay in the Ottawa Basin of the Champlain Sea. Geological Survey of Canada, Paper 85-21, 44 p.
Gadd, N.R. 1988. Lithofacies relationships in a freshwater–marine transition of the Champlain Sea in Gadd, N.R. (ed.), The Late Quaternary Development of the Champlain Sea Basin. Geological Association of Canada Special Paper 35, p. 83–90.
Geyer, W.R., Hill, P.S. and Kineke, G.C. 2004. The transport, transformation and dispersal of sediment by buoyant coastal flows. Continental Shelf Research, v. 24, p. 927–949.
Gilbert, R.W. 2000. The Devil Lake pothole (Ontario): Evidence of subglacial fluvial processes. Geographie Physique et Quaternaire, v. 54(2), p. 245–250.
Gorrell, G.A. 1991. Buried sand and gravel features and blending sands in Eastern Ontario. Ontario Geological Survey Open File Report 5801, 77 p.
Gorrell, G. and Shaw, J. 1991. Deposition in an esker, bead and fan complex, Lanark, Ontario, Canada. Sedimentary Geology, 72, 285-314.
Gurnell, A.M. 1987. Suspended sediments in Glacio-fluvial Sediment Transfer, An Alpine Perspective, Gurnell, A.M. and Clark, M.J. (eds), p. 305–354.
Hammer, K.M. and Smith, N.D. 1983. Sediment production and transport in a proglacial stream: Hilda Glacier, Alberta, Canada. Boreas, v. 12, p. 91–106.
Harvey, F. E., Lee, D.R., Rudolph, D.L., and S. K. Frape. 1997. Locating groundwater discharge in large lakes using bottom sediment electrical conductivity mapping. Water Resources Research, 33(11): 2609-2616.
Henderson, P.J. 1988. Sedimentation in an esker system influenced by bedrock topography near Kingston, Ontario. Canadian Journal of Earth Sciences, 25, 987-999.
Hooke, R. LeB., Wold, B. and Hagen, J.O. 1985. Subglacial hydrology and sediment transport at Bondhusbreen, southwest Norway. Geological Society of America Bulletin, v. 96, p. 388–397. 54
Hoyal, D.C.J.D., van Wagoner, J.C., Adair, N.L., Deffenbaugh, M., Sun, T., LI, D., Huh, C., and Giffin, D.E., 2003, Sedimentation from Jets: A Depositional Model for Clastic Deposits of all Scales and Environments: AAPG, p. Search and Discovery Article #40082.
Hunt, A.S. and Rathburn, A.E. 1988. Microfaunal assemblages of southern Champlain Sea piston cores in Gadd, N.R. (ed.), The Late Quaternary Development of the Champlain Sea Basin. Geological Association of Canada Special Paper 35, p. 145–154.
Hunter, J.A., Pullan, S.E., Burns, R.A., Good, R.L., Harris, J.B., Pugin, A., Skvortsov, A., and Goriainov, N.N., 1998, Downhole seismic logging for high-resolution reflection surveying in unconsolidated overburden; Geophysics, 63, p. 1371-1384.
Inazaki, T., 2004. High resolution reflection surveying at paved areas using S-wave type land streamer. Exploration Geophysics, 35, 1-6.
Johnston, W.A. 1917. Pleistocene and Recent deposits in the vicinity of Ottawa, with a description of the soils. Geological Survey of Canada Memoir 101, 69 p.
Jopling, A.V., and Walker, R.G., 1968, Morphology and origin of ripple-drift cross-lamination, with examples from the Pleistocene of Massachusetts: Journal of Sedimentary Petrology, v. 38, p. 971-984.
Kay, G.M. 1942. Ottawa–Bonnechere Graben and Lake Ontario Homocline. Geological Society of America Bulletin, v. 53, p. 585–646.
Keele, J. and Johnston, W.A. 1913. The superficial deposits near Ottawa. Geological Survey of Canada Guidebook No. 3, p. 126–134.
Kettles, I. and Shilts, W.W. 1987. Tills of the Ottawa region in Fulton, R.J. (ed.), Quaternary Geology of the Ottawa Region, Ontario and Quebec. Geological Survey of Canada, Paper 86-23, p. 10–13.
Leclair, S.F. and Arnott, R.W.C., 2005. Parallel lamination formed by high-density turbidity currents. Journal of Sedimentary Research, 75(1): 1-5.
Leclair, S.F., and Bridge, J.S., 2001, Quantitative Interpretation of Sedimentary Structures Formed by River Dunes: Journal of Sedimentary Research, v. 71, p. 713-716.
Lee, D.R. 1977. A device for measuring seepage flux in lakes and estuaries. Limnology and Oceanography, 22:140-147.
Lee, D.R. and J.A. Cherry. 1978. A field exercise on groundwater flow using seepage metres & and mini-piezometres. Journal of Geological Education, 27: 6-10.
Leeder, M.R. and Stewart, M.D. 1996. Fluvial incision and sequence stratigraphy: alluvial responses to relative sea-level fall and their detection in the geological record in Hesselbo, S.P. and Parkinson, D.N. (eds.), Sequence Stratigraphy in British Geology. Geological Society Special Publication 103, p. 25–39.
Levasseur, D. and Prichonnet, G. 1995. La dispersion clastique des debris rocheux dans les eskers et le till adjacent de la region de Chapais-Chibougamau (Quebec) au Wisconsinien superieur. Canadian Journal of Earth Sciences, v. 32(5), p. 590-602.
MacPherson, J.B. 1968. The physiographic evolution of the Montreal archipelago. The Canadian Geographer, v. 12(4), p. 254–265. 55
Marshall, S.J. and Clarke, G.K.C. 1999. Modeling North American freshwater runoff through the last glacial cycle. Quaternary Research, v. 52, p. 300–315.
MacEachern, J.A., Bann, K.L., Bhattacharya, J.P. and Howell, C.D. Ichnology of deltas: Organism responses to the dynamic interplay of rivers, waves, storms and tides in River Deltas–Concepts, Models and Examples, Giosan, L. and Bhattacharya, J.P. (eds), SEPM Special Publication v. 83, p. 49–85.
Naldrett, D.L., 1988. Sedimentology, stratigraphy and micropaleontology of pre- and early- Champlain Sea fine-grained facies from the Foster sand pit, Ottawa, Ontario. Géographie physique et Quaternaire, v. 42, p. 181–189.
Østrem, G. 1975. Sediment transport in glacial meltwater streams in Jopling, A.V. and MacDonald, B.C. (eds), Glaciofluvial and Glaciolacustrine Sedimentation, SEPM Special Publication No. 23, p. 101–122.
Pair, D.L. and Rodgrigues, C.G. 1993. Late Quaternary deglaciation of the southwestern St. Lawrence Lowland, New York and Ontario. Geological Society of America Bulletin, v. 105, p. 1151–1164.
Parent, M. and Occhietti, S. 1988. Late Wisconinan deglactiation and Champlain Sea invations in the St. Lawrence Valley, Quebec. Géographic Physique et Quaternaire, v. 42(3), p. 215–246.
Pemberton, S.G., Spila, M., Pulham, A.J., Saunders, T., MacEachern, J.A., Robbins, D. and Sinclair, I.K. 2002. Ichnology and Sedimentology of Shallow to Marginal Marine Systems. Geological Association of Canada Short Course Notes Volume 15, 343 p.
Pierson, T.C. and Scott, K.M. 1985. Downstream dilution of a lahar: transition from debris flow to hyperconcentrated streamflow. Water Resources Research, 21, 1511-1524.
Posamentier, H.W. and Allen, G.P. 1999. Siliciclastic Sequence Stratigraphy – Concepts and Applications. SEPM Concepts in Sedimentology and Paleontology No. 7, 210 p.
Posamentier, H.W. and Vail, P.R. 1988. Eustatic controls on clastic deposition II. Sequence and systems tract models in Sea Level Changes: An Integrated Approach, Wilgus, C.K., Hastings, C.K., Kendall, S.G., Posamentier, H.W., Ross, C.A. and Van Wagoner, J.C. (eds), SEPM Special Publication 42, p. 125–154.
Powell, R.D., 1990, Glacimarine processes at grounding-line fans and their growth to ice-contact deltas, in Dowdeswell, J.A., and Scourse, J.D., eds., Glacimarine Environments: Processes and Sediments, Special Publication, 53: London, Geological Society, p. 53-73.
Prest, V.K. and Hode-Keyser, J. 1977. Geology and engineering characteristics of surficial deposits, Montreal Island and vicinity, Quebec. Geological Survey of Canada Paper 75-27, 34 p.
Pugin, A., Larson, T. and Phillips, A., 2002. Shallow high-resolution shear-wave seismic reflection acquisition using a land-streamer in the Mississippi River floodplain: potential for engineering and hydrogeologic applications. In Proceedings, SAGEEP’02 (Symposium on the Application of Geophysics to Engineering and Environmental Problems), February 10-14 Las Vegas, Nevada. CD-ROM edition, 7p.
Pugin, A.J.M., Larson, T.H., Sargent, S.L., McBride, J.H. and Bexfield, C.E., 2004. Near-surface mapping using SH-wave and P-wave seismic land-streamer data acquisition in Illinois, U.S. Leading Edge (Tulsa, OK), 23(7), p. 677-682. 56
Pullan, S.E. and Hunter, J.A., 1999, Land-based shallow seismic methods, Chapter 3 in A handbook of geophysical techniques for geomorphic and environmental research; by Gilbert, R; Geological Survey of Canada, Open File 3731, 1999; p. 31-55.
Reading, H.G. and Richards, M. 1994. Turbidite systems in deep-water basins classified by grain size and feeder type. AAPG Bulletin, v. 78(5), p. 792–822.
Richard, S.H. 1974. Surficial Geology, Winchester, Ontario. Geological Survey of Canada, Open File 213. scale 1:50,000
Rodrigues, C.G. 1992. Succession of invertebrate microfossils and the late Quaternary deglaciation of the central St. Lawrence Lowland, Canada and United States. Quaternary Science Reviews, v. 11, p. 503–534.
Ross, M., Parent, M., Benjumea, B. and Hunter, J. 2006. The late Quaternary stratigraphic record northwest of Montreal: regional ice-sheet dynamics, ice-stream activity, and early deglacial events. Canadian Journal of Earth Science, v. 43, p. 461–485.
Rothlisberger, H. 1972. Water pressure in intra- and subglacial channels. Journal of Glaciology, v. 11, p. 177–203.
Russell, H.A.J., and Arnott, R.W.C., 2003, Hydraulic Jump and Hyperconcentrated Flow Deposits of a Glacigenic Subaqueous Fan: Oak Ridges Moraine, Southern Ontario: Journal of Sedimentary Research, v. 73, p. 887-905.
Rust, B.R. 1977. Mass flow deposits in a Quaternary succession near Ottawa, Canada. Canadian Journal of Earth Sciences, 14, 175-184.
Rust, B.R. 1988. Ice-proximal deposits of the Champlain Sea at south Gloucester, near Ottawa, Canada. In: The Late Quaternary Development of the Champlain Sea Basin (Ed N.R. Gadd), 35, pp. 37-45. Geological Association of Canada, Special Paper, St John's, NF.
Rust, B.R. and Romanelli, R. 1975. Late Quaternary subaqueous outwash deposits near Ottawa, Canada. In: Glaciofluvial and Glaciolacustrine Sedimentation (Eds A.V. Jopling and B.C. McDonald), 23, pp. 177-192. Special Publication - Society of Economic Paleontologists and Mineralogists, Tulsa, Oklahoma.
Sharpe, D.R. 1979. Quaternary geology of the Merrickville area, Southern Ontario. Ontario Geological Survey Report 180, 54 p.
Sharpe, D.R. 1988. Glaciomarine fan deposition in the Champlain Sea. In: The Late Quaternary Development of the Champlain Sea Basin (Ed N.R. Gadd), 35, pp. 63-82. Geological Association of Canada, Special Paper, St. John's, NF.
Sharpe, D.R., Hinton, M.J., Russell, H.A.J. and Desbarats, A.J., 2002. The need for basin analysis in regional hydrogeological studies: Oak Ridges Moraine, Southern Ontario. Geoscience Canada, 29(1): 3-20.
Sharpe, D.R. and Pugin, A. 2007. Glaciated terrain and erosional features related to a proposed regional unconformity in Eastern Ontario; Geological Survey of Canada, Open File 5596p.
Sharpe, D.R and Shaw, J. 1989. Erosion of bedrock by subglacial meltwater, Cantley, Quebec. GSA Bulletin, v.101, p. 1011–1020. 57
Shaw, J., 1996, A meltwater model for Laurentide subglacial landscapes, in McCann, S.B., and Ford, D.C., eds., Geomorphology Sans Frontieres: Chichester, John Wiley & Sons Ltd, p. 181-236.
Shaw, J., Rains, B., Eyton, R. and Weissling, L. 1996. Laurentide subglacial outburst floods: landform evidence from digital elevation models. Canadian Journal of Earth Science, v. 33, p. 1154–1168.
Shilts, W.W. 1994. Lac Deschenes acoustic profiles and cores in Gilbert, R. (ed), A Field Guide of the Glacial and Postglacial Landscape of Southeastern Ontario and Part of Quebec. Geological Survey of Canada Bulletin 453, p. 58–62.
Shreve, R.L. 1972. Movement of water in glaciers. Journal of Glaciology, v. 11, p. 205–214.
Shreve, R.L. 1985. Esker characteristics in terms of glacier physics, Katahdin esker system, Maine. Geological Society of America Bulletin, v. 96, p. 639–646.
Siegert, M.J., Le Brocq, A. and Payne, A.J. in press. Hydrological connections between Antarctic subglacial lakes, the flow of water beneath the East Antarctic Ice Sheet and implications for sedimentary processes in Glacial Sedimentary Processes and Products, International Association of Sedimentologists
Spooner, I.S. and Dalrymple, R.W. 1994. Sediment facies relationships in esker-ridge/esker-fan complexes, Southeastern Ontario, Canada: Application to the exploration for asphalt blending sand. Quaternary International, v. 20, p. 81–92.
Steeples, D.W., 1998, Special issue: Shallow seismic reflection section - Introduction; Geophysics, 63, p. 1210-1212.
Syvitski, J.P.M. 1989. Baffin Island fjord macrobenthos: Bottom communities and environmental significance. Arctic, v. 42(3), p. 232–247.
Teller, J.T. 1988. Lake Agassiz and its contribution to flow through the Ottawa – St. Lawrence System in Gadd, N.R. (ed), The Quaternary Development of the Champlain Sea Basin, Geological Association of Canada Special Paper 35, p. 281–289.
Teller, J.T. and Leverington, D.W. 2004. Glacial Lake Agassiz: A 5000 yr history of change and its relationship to the į18O record of Greenland. Geological Society of America Bulletin, v. 116(5/6), p. 729–742.
Telmer, K. and Veizer, J. 2000. Isotopic constraints on the transpiration, evaporation, energy, and gross primary production budgets of a large boreal watershed: Ottawa River basin, Canada. Global Biogeochemical Cycles, v. 14(1), p. 149–165.
Terasmae, J. 1965. Surficial geology of the Cornwall and St. Lawrence Seaway project areas, Ontario. Geological Survey of Canada Bulletin 121, 54 p.
Thomas, R.H. 1977. Calving bay dynamics and ice-sheet retreat up the St. Lawrence Valley system. Geographic Physique et Quaternaire, v. 31(3,4), p. 347–356.
Torrance, J.K. 1988. Mineralogy, pore water chemistry, and geotechnical behaviour of Champlain Sea and related sediment in Gadd, N.R. (ed.), The Late Quaternary Development of the Champlain Sea Basin. Geological Association of Canada Special Paper 35, p. 281–290.
Vail, P.R., Mitchum, R.M., Todd, R.D., Widmier, J.M, Thompson, S., Sangree, J.B., Bubb, J.N. and Hatlelid, W.G. 1977. Seismic stratigraphy and global changes of sea level in Seismic 58
Stratigraphy–Applications to Hydrocarbon Exploration, Payton, C.E. (ed), AAPG Memoir 26, p. 149–212.
Van Wagoner, J.C., Mitchum, R.M., Campion, K.M., and Rahmanian, V.D., 1990, Siliciclastic sequences stratigraphy in well logs, cores, and outcrops, v. No.7, AAPG Methods in Exploration Series. Williams, D.A., 1991. Paleozoic Geology of the Ottawa–St. Lawrence Lowland, Southern Ontario. Ontario Geological Survey Open File Report 5770.
Walker, R.G. and James, N.P. (eds) 1992. Facies Models: Response to Sea Level Change. Geological Association of Canada, 409 p.
Williams, D.A. 1991. Paleozoic geology of the Ottawa-St. Lawrence Lowland, southern Ontairo. Ontario Geological Survey, Open File Report 5770p.
Wingham, D.J., Siegert, M.J., Shepherd, A.P. and Muir, A.S. 2006. Rapid discharge connects Antarctic subglacial lakes. Nature, v. 440, p. 1033–1036. 59
8.0 Appendix: Methods
A.1 Basin Analysis Approach
The study of the Vars-Winchester esker follows a basin analysis approach for regional hydrogeological applications in glaciated basins (Sharpe et al. 2002; Eyles et al. 1985). Initial work has focused on collection and synthesis of archival data and development of a conceptual geological model. Early data collection has focused on filling data gaps based on assessment of existing data.
Figure 37. Heterogeneity of Quaternary deposits at various scales, and data being collected to assess this heterogeneity. Asterisks indicate datasets that have been partially collected as of May 2007.
A.2 Database Development
Data collation and integration has been initiated in three key areas, i) digital library, ii) relational database, and iii) GIS coverage. South Nation Conservation Authority personnel visited municipal offices and agencies during May to July 2006. Relevant geotechnical reports, hydrogeological reports 60 and groundwater quality reports were catalogued in a relational database. The reports were also scanned and converted to a PDF format.
Hydrogeological data contained in reports are being entered into an MSAccess relational database for the watershed with a focus on the Vars-Winchester corridor. This archival source of geotechnical and hydrogeological data complements key data obtained from the South Nation’s Source Water Protection database, which includes the Ministry of Environment’s Water Well Information System. Integration of disparate geological descriptions is being facilitated by use of the GSC material coding scheme (Russell et al, 1998). Water & Earth Science Associates Limited (WESA) entered data in the relational database and South Nation Conservation Authority is continuing to add pertinent data.
Digital map coverage has been obtained from the Ontario Geological Survey seamless geological map. The mapped location of eskers has been captured from Gorrell (1991). Geological coverage is complemented by cultural coverage from the Ministry of Natural Resources. The Ontario provincial DEM (10 m resolution, Ministry of Natural Resources) is being used as a topographic datum for all datasets. Additional datasets have been obtained from the Ontario Ministry of Natural Resources (e.g. watershed boundaries), Environment Canada (e.g. stream baseflow, ecoregions), Ottawa Carleton Regional Municipality (e.g. orthophotography), and the Geological Survey of Canada Hazards Program.
A.3 Outcrop and core studies Cummings, Don; Geological Survey of Canada and South Nation Conservation Authority
Continuous cores were collected with a wireline-attached PQ-sized core barrel (11.3 cm diametre) using the BAROID mud rotary-drilling technique (see BAROID for information http://www.baroididp.com/baroid_idp.asp). This technique is taught at the BAROID Career Development Center in Houston, Texas. Key elements of this technique are the use of BAROID stabilizers/viscosifiers and recirculation of mud to surface. Core recovery using this technique approached 100%, even in sand and gravel.
Once collected, cores were analyzed using standard sedimentological and sequence-stratigraphic techniques and concepts (Curray, 1964; Van Wagoner, 1990; Walker and James, 1992; Pemberton et al., 2001). Outcrops were photographed in their entirety, and then analyzed in a similar fashion.
A.4 Shallow seismic reflection methods Pullan, Susan and Pugin, André; Geological Survey of Canada
Land-based seismic geophysical techniques use measurements of the time taken for acoustic energy to travel from a source on the surface through the subsurface and back to a series of receivers on the ground. Energy is refracted or reflected at boundaries where there is a change in acoustic impedance (the product of material density and seismic velocity). Because contrasts in acoustic impedance are generally associated with lithological boundaries, seismic techniques can be used to obtain subsurface structural information. This section briefly outlines the application of shallow seismic reflection methods to delineating the structure of unconsolidated sediment and the underlying bedrock surface.
Seismic reflection methods have been the primary geophysical tool used in oil and gas exploration for over 60 years. Because of the tremendous commercial importance of oil, much industrial research and development has been invested in this branch of geophysics. By the 1960s, specialized field procedures, digital magnetic tape recording, and computer processing of the data had become standard in the industry. Conventional seismic reflection techniques are highly sophisticated, but require considerable investment in both data acquisition and processing.
In the early 1980s, the development of digital enhancement engineering seismographs with high-pass filtering capabilities and the proliferation of increasingly powerful microcomputers, began to make the 61 application of seismic reflection methods to "shallow" problems a viable alternative. Over the last 20– 25 years, much experience and expertise in the application of shallow high-resolution reflection techniques have been gained, and today these methods are accepted and proven shallow geophysical tools. Seismic reflection techniques can be applied using compressional (P-wave) or shear (S-wave) energy. Compressional waves are those in which the particle motion and direction of wave propagation are the same, whereas shear waves are those in which the particle motion is normal to the direction of wave propagation (Fig. 38).
Figure 38. Schematic diagram showing the particle motions for compressional or P waves (upper panel) and shear or S waves (lower panel).
Seismic reflection methods involve measurement of the time taken for seismic energy to travel from the source at or near the surface, through the ground to an acoustical discontinuity, and back up to a receiver or series of receivers on the ground surface (Fig. 39a). Data are usually acquired continuously along a survey line, and processed to produce a seismic section which is a two-way travel time cross-section of the subsurface (Fig. 39b). Velocity-depth functions calculated from the data, or seismic logging of a nearby borehole(s), are used to translate the two-way travel time into depth.
Today, virtually all shallow seismic data are collected and processed based on the common midpoint (CMP) method (often also referred to as the common-depth-point, or CDP, method) which is an adaptation of the methods used by the petroleum industry. In CMP surveys, multi- (12, 24, or more) channel data are recorded for each shotpoint. During processing, the data are sorted according to their common midpoints or common depth points (Fig. 40). Each trace is corrected for offset according to a velocity–depth function determined from the data (normal moveout, or NMO, corrections). A standard sequence of CMP data processing steps includes trace editing, static corrections, bandpass filtering, gain scaling, velocity analyses, normal moveout corrections and finally, stacking of the NMO- corrected traces in each CMP gather to create a single trace on the final section. This stacking procedure is the essence of the CMP technique, and allows a potential improvement in the signal-to- noise ratio of the data according to the square root of the fold (number of traces summed to produce the final processed trace at a given point along the seismic profile).
The successful application of any shallow reflection survey depends on the detection of high-frequency energy reflected from velocity discontinuities within the subsurface. Earth materials, however, especially unconsolidated overburden materials, are strong attenuators of high-frequency energy. Thus, compressional (P) seismic waves in the 10–90 Hz range commonly used in petroleum 62 exploration may be reflected from depths of thousands of metres, but energy with frequencies above 100 Hz normally only have travel paths on the order of tens or hundreds of metres. The ability of a particular site to transmit high-frequency energy is a major factor in determining the quality and the ultimate resolution of a shallow reflection survey.
The optimum conditions for shallow reflection surveys (P-wave) are usually when the surface materials are fine-grained and water-saturated; reflections with dominant frequencies of 300–500 Hz can be obtained in such field situations. These frequencies correspond to seismic wavelengths in unconsolidated overburden materials on the order of 3–5 m, with a potential subsurface structural resolution of approximately 1 m. Experience has shown that excellent high-resolution, P-wave, seismic reflection data can be obtained where water-saturated, fine-grained Champlain Sea sediment is exposed at the surface (Fig. 41).
Shear wave reflections are commonly much lower in frequency (10–100 Hz) than shallow P-wave reflections. However, resolution of the seismic signals depends on the signal wavelength (higher resolution associated with shorter wavelengths). As the velocity of shear waves in unconsolidated materials can be an order of magnitude lower than the P-wave velocity of the same sediment (particularly if those sediment are water-saturated), the resolution of shear wave data can exceed that obtainable with P-wave data.
Seismic profiles are sections in two-way travel time (not depth). Velocity functions are estimated from the seismic data at intervals along the line during the processing sequence, in order to calculate the normal move out corrections applied to the data before the stacking procedure, and these velocities can also be used to convert the two-way travel time section to a depth section. However, velocities determined from reflection data can be subject to large uncertainties, depending on the move out of reflection events. Whenever possible, accurate downhole velocity data from borehole measurements should be obtained in support of the seismic reflection survey (Hunter et al., 1998). 63
Figure 39. Basic premise of seismic reflection methods. a) Seismic energy produced on the ground surface travels from the source down to an acoustic impedance (product of density and velocity) boundary, where it is partially transmitted and partially reflected back towards the surface. b) Data are usually acquired continuously along a survey line and the record of ground motion as a function of time is related to the subsurface structure.
Further discussion on the application of seismic methods to geomorphic and environmental problems can be found in Pullan and Hunter (1999). Steeples (1998) provides an overview of the development of shallow seismic reflection techniques, and the suite of papers in that special issue of Geophysics provides a summary of the state-of-the-art of shallow seismic reflection at the time. 64
Figure 40. Schematic diagram showing a) the subsurface travel paths of reflections from a field record and b) a common midpoint gather. The traces in the CMP gather will be processed and stacked together to form a single trace on the final CMP section (6-fold).
Figure 41. Example field shot gather obtained during the P-wave reflection survey using a 12-gauge shotgun source and 50 Hz vertical geophones at 3 m spacing: a) raw record, b) same record after high-pass filtering. These records show excellent reflection energy (hyperbolic events).
Seismic Landstreamer/Minivib System
Shallow seismic reflection surveys are a powerful tool for mapping detailed subsurface structure, with applications in a wide variety of groundwater, hazard, engineering and environmental investigations. More widespread use of this technique has been limited partly by the time and cost involved in acquiring and processing the data. The efficiency of data collection is largely dependent on the time required to individually plant every receiver (geophone) and to move and reconnect seismic cables as 65 the survey proceeds along a seismic line. As well, the ability to produce and record high-frequency energy for shallow seismic reflection surveys depends on the ground conditions, the effectiveness of ground coupling for both the receivers and the source, the frequency and energy of the seismic source, and the source and receiver spacings (which define the fold – see Fig. 40).
The Geological Survey of Canada has recently been successful in mating the IVI (Industrial Vehicles International, Inc) minibuggy minivib source (http://www.indvehicles.com) and landstreamer receiver arrays (both P-wave and horizontally-polarized shear (SH-) wave). The seismic landstreamer/minivib system is one way of addressing both the efficiency of data collection and data quality (improvement of signal-noise ratio by decreased source and receiver spacings).
Landstreamers consist of towed arrays of geophones fixed on sleds and have been demonstrated to be an efficient means of recording shear-wave reflection data (e.g. Inazaki, 2004, Pugin et al., 2004). The Geological Survey of Canada has built an SH-wave landstreamer array (24-48 channels) consisting of small metal sleds with 2 horizontal 8 Hz geophones per sled (Fig. 42) (Pugin et al. 2002). Typically, receiver spacing of 0.75 m are used, though the spacing can be adjusted according to the survey targets. The short spacing of the sleds avoids spatial aliasing of the surface waves for optimum results when FK-spatial filters are applied. For P-wave surveys, one vertical 40 Hz geophone is mounted on each sled and the sled spacing is typically 1.5-3 m. These landstreamers are designed for use along paved or gravel roads.
Figure 42. Photo of the minivib source and SH-wave landstreamer in operation, 2006. 66
Figure 43. Photos of the IVI minibuggy vibratory source in operation. In the lower photo the minivib is being operated in SH-mode (note weight above plate in mounted horizontally.
A.5 Simple, inexpensive methods for hydraulic head measurements and groundwater flux estimations Charland, Coralie: University of Ottawa
This section describes three simple techniques used to measure groundwater - surface water interactions. The quantification of seepage or discharge of groundwater in surface water bodies is an important element to water balance calculations. Miniature piezometres and seepage metres are devices that can be manually installed in saturated surficial sediment (rivers, lakes, ditches) overlying the aquifer of interest. Shallow piezometres are used to obtain hydraulic head and hydraulic conductivity of sediment, and seepage metres show direction (vertical gradients) and rate of groundwater flow. The last method described here uses an electrical conductivity probe to reveal groundwater discharge areas, which is useful to determine the whereabouts of possible discharge zones in rivers and lakes.
All calculations are based on Darcy’s Law, where Q=A K dh/dl, where Q is the groundwater flux [L3/T], A the area through which flow occurs [L2], dh/dl the hydraulic gradient [ ] and K the hydraulic conductivity of the sediment [L/T].
1) Shallow piezometres
Material needed: drive bolts, drive pipe, drive head, monkey hammer, PVS pipe, hand-made screens (for example: 15cm pipes, 7 notches or slots, mesh – size function of deposit - glued or welded around in place). From this device, we obtain: i) water levels of the free water table; ii) vertical hydraulic gradient (unit less) i=ǻh/ǻl (with ǻh the difference between the water level inside and outside the pipe, and ǻl the depth to which the pipe is set); iii) hydraulic conductivity, using slug or bail tests (Freeze and Cherry,1979). 67
2) Seepage metres
A) Open-top seepage metres
Material needed: large pipe, coupling to use as drive head if necessary, monkey hammer (post driver), small plastic piping for siphon, bag, and elastic bands.
An initial volume of water is placed in the bag and monitored over a period of time, permitting the calculation of flux in or out. Corrections must be done to take into account variations in surface water level.Three parametres can be calculated with the data obtained from this device: i) the vertical hydraulic gradient (dimensionless) i=ǻh/ǻl (with ǻh the difference between the water level inside and outside the pipe, and ǻl the depth to which the pipe is set); ii)the discharge rate (m3/d) r=V/ǻt with V the change in volume (corrected to take into account surface water level variation) and ǻt the time lapse between measures; and iii) the flux (m3/m2/day or m/day) Q= r/A, where A is the cross-sectional area of the seepage metre pipe.
Fig. 44. A) Installation of a shallow piezometre. A, casing driven into the sediment; B, plastic tube with screened tip inserted in the casing; C, plastic tube is a piezometre and indicates differential head (h) with respect to the surface water (Lee and Cherry, 1978). Secure loose pipe downwards (but not in the water) on a wooden stake. B) Open-top seepage metre. A, big piping; B, tube; C, small tube; D, elastic band; E, plastic bag. The flexible tubing is secured to the pipe to keep bag under water level. C) Closed-top seepage metre, using a tilted bucket, well pushed into sediment so that sides are sealed. A, plastic bag, open end sealed; B, elastic band; C, small tube; D, larger tube; E, rubber stopper; F, drum or bucket. (Lee, 1977; Lee and Cherry, 1978)
B) Closed-top seepage metres
Material needed: bucket, rubber stopper, plastic tubing, plastic bag, elastic bands. The data obtained from this device are the discharge rate (m3/d) r=V/ǻt with V the change in volume (corrected to take into account surface water level variation), and ǻt the time lapse between measures; and the vertical flux (m3/m2/day or m/day) Q= r/A, where A is the cross-sectional area of the bucket or drum. 68
3) Bottom sediment electrical conductivity mapping
Material needed: boat, EC&T probe.
This technique requires a probe capable of measuring electrical conductivity and temperature. By dragging the probe along the water-sediment interface changes the electrical conductivity of bottom water in a lake or river can be detected indicating possible areas of groundwater discharge (Harvey et al., 1997). Anomalies can be double-checked using seepage metres.
Figure 45. Conductivity survey for groundwater discharge at the base of rivers or lakes.