TY - JOUR
T1 - A heterogeneity model comparison of highly resolved statistically anisotropic aquifers
AU - Siirila-Woodburn, Erica R.
AU - Maxwell, Reed M.
N1 - Funding Information:
Data for this paper is available on request from the corresponding author. Funding for this work was provided by DOE NETL Grant No. DE-FE0002059 and EPA STAR Grant No. RD-83438701-0 . This research was supported in part by the Golden Energy Computing Organization at the Colorado School of Mines using resources acquired with financial assistance from the National Science Foundation and the National Renewable Energy Laboratory . This research has been supported in part by a grant from the US Environmental Protection Agency’s Science to Achieve Results (STAR) program . Although the research described in the article has been funded wholly or in part by the US Environmental Protection Agency’s STAR program through Grant RD-83438701-0 , it has not been subjected to any EPA review and therefore does not necessarily reflect the views of the Agency, and no official endorsement should be inferred.
Publisher Copyright:
© 2014 Elsevier Ltd.
PY - 2015/1/1
Y1 - 2015/1/1
N2 - Aquifer heterogeneity is known to affect solute characteristics such as spatial spreading, mixing, and residence time, and is often modeled geostatistically to address aquifer uncertainties. While parameter uncertainty is often considered, the model uncertainty of the heterogeneity structure is frequently ignored. In this high-resolution heterogeneity model comparison, we perform a stochastic analysis utilizing spatial moment and breakthrough curve (BTC) metrics on Gaussian (G), truncated Gaussian (TG), and non-Gaussian, or "facies" (F) heterogeneous domains. Three-dimensional plume behavior is rigorously assessed with meter (horizontal) and cm (vertical) scale discretization over a ten-kilometer aquifer. Model differences are quantified as a function of statistical anisotropy, ε, by varying the x-direction integral scale of hydraulic conductivity, K, from 15 to 960 (m). We demonstrate that the model is important only for certain metrics within a range of ε. For example, spreading is insensitive to the model selection at low ε, but not at high ε. In contrast, center of mass is sensitive to the model selection at low ε, and not at high ε. A conceptual model to explain these trends is proposed and validated with BTC metrics. Simulations show that G model effective K, and 1st and 2nd spatial moments are much greater than that of TG and F models. A comparison of G and TG models (which only differ in K-distribution tails) reveal drastically different behavior, exemplifying how accurate characterization of the K-distribution may be important in modeling efforts, especially in aquifers where extreme K values are often not measured, or inadvertently overlooked.
AB - Aquifer heterogeneity is known to affect solute characteristics such as spatial spreading, mixing, and residence time, and is often modeled geostatistically to address aquifer uncertainties. While parameter uncertainty is often considered, the model uncertainty of the heterogeneity structure is frequently ignored. In this high-resolution heterogeneity model comparison, we perform a stochastic analysis utilizing spatial moment and breakthrough curve (BTC) metrics on Gaussian (G), truncated Gaussian (TG), and non-Gaussian, or "facies" (F) heterogeneous domains. Three-dimensional plume behavior is rigorously assessed with meter (horizontal) and cm (vertical) scale discretization over a ten-kilometer aquifer. Model differences are quantified as a function of statistical anisotropy, ε, by varying the x-direction integral scale of hydraulic conductivity, K, from 15 to 960 (m). We demonstrate that the model is important only for certain metrics within a range of ε. For example, spreading is insensitive to the model selection at low ε, but not at high ε. In contrast, center of mass is sensitive to the model selection at low ε, and not at high ε. A conceptual model to explain these trends is proposed and validated with BTC metrics. Simulations show that G model effective K, and 1st and 2nd spatial moments are much greater than that of TG and F models. A comparison of G and TG models (which only differ in K-distribution tails) reveal drastically different behavior, exemplifying how accurate characterization of the K-distribution may be important in modeling efforts, especially in aquifers where extreme K values are often not measured, or inadvertently overlooked.
KW - Breakthrough curve
KW - Geostatistics
KW - Heterogeneous
KW - High resolution
KW - Spatial moment
KW - Uncertainty
UR - http://www.scopus.com/inward/record.url?scp=84913529417&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84913529417&partnerID=8YFLogxK
U2 - 10.1016/j.advwatres.2014.10.011
DO - 10.1016/j.advwatres.2014.10.011
M3 - Article
AN - SCOPUS:84913529417
SN - 0309-1708
VL - 75
SP - 53
EP - 66
JO - Advances in Water Resources
JF - Advances in Water Resources
ER -