Geophysical surveys have been done at the Red River floodway and Letellier sites in order to provide more information on the genesis of the gypsum rosettes and to assess the application of geophysics as an exploration tool for new deposits.

    The physical properties of the Agassiz units, as well as overlying and underlying units, have been characterized in a number of geotechnical studies on the Winnipeg area (e.g. Baracos, 1977; Graham and Shields, 1985).  Seismic refraction surveys conducted during the construction of the Red River floodway (Hobson et al., 1964) defined the seismic velocity response of the different geological units and revealed considerable variation in the depth to bedrock along the floodway associated with karstification of the bedrock.  Near the floodway site, the bedrock is at 22 m depth but there is a sharp decrease to 12 m depth, several hundred metres to the north.  Previous geoelectric studies in the Winnipeg area have indicated that the conductivity of the upper several metres of soil is ~100 mS/m (millisiemens per metre), and the conductivity of the Agassiz units is 125-200 mS/m.

    Surveys at the Red River floodway and Letellier sites.  Geophysical surveys at the Red River floodway site included electromagnetic/electrical (EM31, EM38, DC-resistivity, PROTEM47) and seismic refraction methods.  The seismic refraction results resolved the thickness of the soil layer, the combined thickness of the clay units, and the thickness of the till, but not resolve the interface between Agassiz Unit 1&2 and Unit 3.  Small-scale EM measurements in the excavation pit revealed the conductivity of the Agassiz Unit 1&2 to be 200 mS.m-1 and that of Agassiz Unit 3 to be 120-150 mS.m-1. Surface DC resistivity measurements identified three layers corresponding to more resistive soil and Agassiz Unit 3, conductive Agassiz Unit 1&2, and more resistive till/bedrock.

    The most surprising results of the geophysical surveys were the lateral variations in the electrical response.  Surveys using an EM31 in horizontal dipole mode (penetration 3 m) and vertical dipole mode (5 m), DC resistivity profiling (5-10 m), and time-domain electromagnetics (5-20 m) all revealed a broad resistive zone containing a more narrow conductive zone coincident with the excavation pit (Fig. 3).  These observations suggest that the location of the deposit may be related to structures within the overlying Agassiz Unit 3.

    Initial geophysical surveys at the Letellier site have revealed a  conductive anomaly coincident with the gypsum deposit.  EM31 readings increase from 100 mS.m-1 in background areas to 250 mS.m-1 in the anomaly.  Conductivity readings correlate inversely with, and may be partially controlled by, a corresponding 75 cm variation in surface topography.  Readings in the dugout, where gypsum rosettes were collected, are in the range of 190-220 mS.m-1, similar to those from the excavation pit at the Red River floodway, but less conductive than the overlying units.  At both the floodway and Letellier sites, there is an indication that the location of the gypsum deposits is related to linear conductive features in the overlying units.
 

Interpretation and discussion

    One significant control on lateral variations in the Agassiz clays is iceberg scouring of Agassiz Unit 1&2.  The scouring produced low-angle faults below the scour trough, which extended at least 3 m below the scour, and high-angle normal faults outside the scour (Woodworth-Lynas & Guigné, 1990).  The scours can be recognized on the present-day ground surface as linear features that are on the order of 100 m wide and have topographic relief of several decimetres.  Shallow troughs at the margins of the scours have been deepened locally by farmers to control surface runoff.  The scours appear to influence the near-surface hydrology, and the associated deeper structures, particularly the normal faults, likely influence the deeper hydrological conditions.  In the absence of vertical structures, such as faults, the Agassiz clays form an aquitard, with low hydraulic permeability across the laminae (Baracos et al., 1980).

    The geophysical results suggest a relationship between gypsum rosette accumulations and linear anomalies in near-surface units; iceberg scours can produce such linear anomalies.  Hence, we have started to investigate the geophysical response of iceberg scours in the Agassiz clays.  To date, geophysical surveys have been completed over two scours in the Lorette area that were investigated by Woodworth-Lynas & Guigné (1990).  Electromagnetic survey results show conductivity variations of a magnitude and form similar to those observed at the gypsum sites.