Improving our ability to model crystalline aquifers using field data combined with a regionalized approach for estimating the hydraulic conductivity field - Archive ouverte HAL Accéder directement au contenu
Article Dans Une Revue Journal of Hydrology Année : 2021

Improving our ability to model crystalline aquifers using field data combined with a regionalized approach for estimating the hydraulic conductivity field

(1, 2, 3) , (1) , (1) , (1, 3) , (1) , (1) , (1, 3)
1
2
3

Résumé

Modelling of heterogeneous aquifers, such as crystalline aquifers, is often difficult and, flow and transport predictions are always uncertain, suffering of our imperfect knowledge of the spatial distribution of aquifer parameters. This paper aims to test the robustness of a first-order hydraulic conductivity map estimated from both a detailed water-table map and hydraulic-conductivity statistics on a granitic watershed in Brittany (57 km(2)), compare it to local estimates and assess it through various numerical models. Map values range on four orders of magnitude (10(-7) to 10(-4) m/s, mean: 3.9x10(-6) m/s) and their comparison to local estimates from various sources (pumping tests, MRS measurements, streamflow) gives satisfactory results. Four hydraulic conductivity fields were assessed through numerical modelling under steady-state condition. Model 1 used the regionalized hydraulic conductivity field directly, Model 2 used a uniform value (average of Model 1), Model 3 used a hydraulic conductivity field obtained by inverse modelling of the water table and Model 4 used two zones of uniform values based on the analysis of Model 1 and Model 3 fields. Model results were analysed based on their ability to reproduce the observed water-table levels and the groundwater flow directions. Modelled groundwater discharges to streams at sub-catchment scale were compared to spatial streamflow measurements performed during low water condition, which help validating models. The comparison between the fields obtained from Model 3 and that from the regionalized method (Model 1) shows that they are close in terms of mean values and spatial distribution. Model 1 reproduces rather well the water-table map and the groundwater flow directions. Model 2 shows the less good results. Model 4 has led to satisfactory results and shows that the hydraulic conductivity is higher (2.1x10(-6) m/s) where the water table is located in the fractured zone, and lower (3.3x10(-7) m/s) where it is located in the saprolites (highly weathered rock), which is expected for such aquifer system. Modelled groundwater discharges to streams are comparable in all models to streamflow measurements in most sub-catchments, but the models overestimate them in certain places, mainly because of sub-surface drains in a forest capturing part of the groundwater that can no longer return to streams (drains were not considered in the models). In addition, experiment on a second watershed (40 km2) shows how with much less field data the methodology can already provide interesting information on the hydraulic conductivity field (values and spatial distribution). Results are very encouraging and open up prospects for using quantitative and qualitative information from the mapping of hydraulic conductivity to constrain the spatialization of hydrodynamic parameters on models and thus our ability to model such complex aquifers.
Fichier principal
Vignette du fichier
Dewandel_et_al_HYDROL39896_REV1_Unmarked_juillet2021_HA1.pdf (4.39 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03292193 , version 1 (15-09-2021)

Identifiants

Citer

Benoît Dewandel, Alexandre Boisson, Nadia Amraoui, Yvan Caballero, Bruno Mougin, et al.. Improving our ability to model crystalline aquifers using field data combined with a regionalized approach for estimating the hydraulic conductivity field. Journal of Hydrology, 2021, 601, pp.126652. ⟨10.1016/j.jhydrol.2021.126652⟩. ⟨hal-03292193⟩
123 Consultations
157 Téléchargements

Altmetric

Partager

Gmail Facebook Twitter LinkedIn More