• Aucun résultat trouvé

Modelling of the paleo-hydrogeological evolution of a fractured-rock aquifer following the Champlain Sea transgression in the St. Lawrence Valley (Quebec).

N/A
N/A
Protected

Academic year: 2021

Partager "Modelling of the paleo-hydrogeological evolution of a fractured-rock aquifer following the Champlain Sea transgression in the St. Lawrence Valley (Quebec)."

Copied!
1
0
0

Texte intégral

(1)

Modeling of the palaeo-hydrogeological evolution of a fractured-rock aquifer

following the Champlain Sea Transgression in the St. Lawrence Valley (Quebec)

Marc Laurencelle

(1)*

, René Lefebvre

(1)

, John Molson

(2)

, Michel Parent

(3)

(1) Institut national de la recherche scientifique, Centre Eau Terre Environnement, Québec, Canada (2) Université Laval, Department of Geology and Geological Engineering, Québec, Canada

(3) Natural Resources Canada, Geological Survey of Canada (GSC), Québec, Canada

(*) (to contact first author: marc.laurencelle@ete.inrs.ca)

Abstract n°2343

① ② ③

1. I

NTRODUCTION

AND

O

BJECTIVES

The main objectives are to:

3. M

ETHODS

4. R

ESULTS

2. S

TUDY

AREA

&

ITS

INTRIGUING

BRACKISH

AREA

Qc

max. extent

Conceptualisation of the palaeo-problem:

Numerical modeling is required to represent this coupled flow & transport problem (water density depends on salinity), in a system with complex

geometry (e.g. topography) with spatial and temporal variations of model parameters and boundary conditions.

Numerical modeling approach:

5. C

ONCLUSION

i) Quantitative comparison of mass fluxes for different processes, alone (A, B, C) or coupled (A & C)

ii) A plausible reconstitution of the evolution of groundwater salinity and dynamics following deglaciation (A & C)

2Dxz(t) model flowchart for: C) variable density coupled flow & transport, with fully-transient

conditions at model surface 1Dz(t) model for: A) advective-dispersive salt

transport in and below the forming aquitard;

B) expulsion of saltwater from the forming

aquitard due to clay accumulation & consolidation

adi m . e le v. in a qui ta rd (-)

C. Density-driven free convection in the rock aquifer

de pt h i n ro ck (m ) de pt h i n ro ck (m )

time (kyr) time (kyr)

time (kyr) early time (kyr) time (kyr)

multi-dimensional axis (-) q m : m ass f lu x ( g/ a) q m (g /a ) q (g m /a ) de pt h i n ro ck (m ) t: time (kyr)

A. Aquitard salinity evolution

during and after sedimentation B. Consolidation and drainage during sedimentation (S&C) of clays

t ≈ 2.5 ka t ≈ 8.2 ka

t = 13.0 ka ≈ 0 BP This study aims to better understand changes in regional aquifer dynamics

that followed the Champlain Sea marine incursion, ≈13 000 years ago. 1. reconstruct the evolution of groundwater salinity and dynamics

following deglaciation, using physically-based numerical models; 2. identify the key palaeo-hydrogeologic processes involved;

3. explain the present-day state of the system, especially the persistence of brackish groundwater even at shallow depths;

4. improve our understanding of high-amplitude marine transgression-regression effects on groundwater systems in general.

Study area in Montérégie Est, Quebec, Canada

• The regional fractured-rock aquifer system

underwent a complex late glacial history that left a sedimentary record and ~2 200 km2 of

brackish groundwater.

• This brackish groundwater can be used as an

indicator of the historical aquifer dynamics.

• The study area also benefits from

high-quality data available from a regional

hydrogeological assessment (aka “PACES”).

 Density-driven free convection dominates salt transport in the rock aquifer, even in clay-confined areas.  Consolidation of clay generates advective mass fluxes that bring salt into the rock aquifer.

 Diffusive fluxes from the base of the clay aquitard to the rock aquifer are negligible compared to other fluxes.  Diffusive fluxes from the aquitard to the overlying sea / lake / surface are the dominant transport process

controlling the post-marine evolution of salinity in the aquitard and, hence, underneath.

 Density-driven saltwater fingers sink until they reach the brine “floor”; afterward, groundwater salinity is progressively homogenized by spreading and mixing.

 In higher areas not covered by clay, salty water leaching by fresh groundwater is substantial, whereas leaching is very limited under the aquitard, thus explaining the presence today of brackish groundwater in the aquifer. Canada SW St. Lawrence Valley

Montérégie Est

-120 -100 -80 -60 -40 -20 0 -500 0 500 1000 1500 ice sheet global sea level ice load (thickness) E lev at ion ( m )

Last glacial period timeline (cal ka BP)

(post-glacial epoch)

(LGM)

Surface conditions during the last glacial cycle

(non-local example data for Waterloo, Ontario, Canada; adapted from Lemieux et al., 2008)

t ≈ 1.0 ka

t

0

-13 simul. time cal ka BP 13 0

(c la y aq uit ar d) (r ock )

Brackish delimited groundwater Fresh GW with signif. salinity Thickness of the clay layer (m):

Montérégie Est study area

brackish gw zones

likely

(similar ctx) rel. continuous well-defined &

possible resid. sal. (St. L. River) 30 km N (wide cross-section) (wide cross-section)

Wide cross section A-B-C of St. L. Valley topography

(m)

(km)

Map source: Geol. Surv. Can. O.F. n°6960

red: salinity (TDS); qf: fluid flux (L3/T/L2); q

m: mass flux (M/T/L2) free convection, fingering + sinking, (hydrodyn. dispersion) qf → qm forced convection qf → leaching surface conditions subsurface processes RSL(t) clay acc.(t) consolidation qf → qm diffusion qm stagnant brines sea / lake clay aquitard REG. ROCK AQUIFER salinity(t)

Conceptualization of palaeo-conditions (at model surface):

RSL

sal

clay

Relative sea (or lake) level (RSL)

flooding lower elevations ≈ 2-stage linear decrease

170

0

Champlain

Sea Lake Lampsilis

2

0 4

RS

L

(m)

Relative time (ka)

Salinity of flooding water

(at bottom of the water bodies) ≈ constant at max. then linear decrease late in the C.S. stage

35 0.25 2 0 4 sa l ( g/L ) max. salinity seawater at sea bottom (stratified sea)

incr. vertical mixing of deep & shallow sea waters

lacustrine env. freshwater

Clay & silt accumulation at sea bottom (from settling of fines)

≈ linear thickening, mostly during the C.S. stage

30 0 2 0 4 cl ay (m

) & consolidated formed

0 30 tt=2ka n t2 t1 . . . moving clay top rock aquifer z+ B A time+ GW flow solution → p or hfw → (qx , qz)

ρ ← c ← solute transport solution changing BC's qz-dep., changing BC's ≥ csw,max ≤ cfw relative salinity

(uniform depth of model base = 1000 m)

A B C B C C B A excess pressure buildup Darcy velocities porosity reduction t+ c uc c uc c uc c uc di spe rs. CA RA adv. S&C CA → RA adv. S&C CA→RA (bis)

conv. inflow surface → RA (@uc: unconfined)

conv. inflow (surface →) CA → RA (@c)

"c" : confined by clay(t) cover "uc" : unconfined (clay-free)

Références

Documents relatifs

For uncompatible boundary conditions, the uniqueness of the stationary solutions of the equation corresponding to a special BGK approximation of the Boltzmann collision operator,

When using sampling to estimate statistical properties of RNA structures at the thermodynamic equilibrium, a recur- rent – crucial – question is to choose the number of gener-

Here, we present a phylogenetic analysis of core Tillandsioideae using the nuclear gene phyC and plastid data obtained from genome skimming.. We investigate evolutionary relationships

Our study on the immune monitoring of 18 patients with severe OA after local treatment by autologous ASC injection in the knee found slight but significant changes in

After a short description of the mechanisms involved in the degradation of cement paste, we present a reactive transport calculation in 3D-cylindrical geometry in

But it is well known that such set systems are distributive lattices O(N) (and so could be called poset convex geometries as well), hence closed under union and intersection

The evolutionary history of the taxa in X can then be represented by a rooted phylogenetic network N on X where the tree nodes correspond to speci- ation events and the reticulate

As discussed when commenting on example (xxi), even if we had relied solely on a logistic regression model to make predictions (we chose to rely on machine learning prediction),