I will be at the "AGU fall meeting" taking place in San Francisco, USA from Monday (December 12th) to Friday (December 16th). Please contact me, if you want to chat with me in person. And please check out and come and talk to me at the presentation I will be giving on Monday and on Thursday
Looking forward to meet you there!
Asymmetry in Spatial Dependence Structures
The poster was upgraded to a presentation, which will take place on Monday morning:
Monday, 12 December 2016 09:15 - 09:30 Moscone South - 302
In this poster I can show you how non-Gaussian dependence (non symmetric dependence) can be quantified, is encountered frequently in data, can be modelled (simulations of spatially distributed random variables), and what impacts such an improved description has on dependent parameters (such as solute transport behaviour based on high-dimensional hydraulic conductivity fields.
Link to the AGU calendar.
| date | Monday, 12 December 2016 |
| time | 13:40-18:00 |
| location | Moscone South - Poster Hall |
| number | NG13A-1699 |
| title of talk | “Asymmetry in Spatial Dependence Structures” |
| Authors | Claus P. Haslauer, András Bárdossy |
Predict Breakthrough Curves based on Structural Parameters of Heterogeneous K Fields
It is desirable to predict the shape of breakthrough curves downgradient of a solute source from subsurface structural parameters (as in the small-perturbation macrodispersion theory) both for realistically heterogeneous fields, and at early time, before any sort of Fickian model is applicable. Using a combination of a priori knowledge, large-scale Monte Carlo simulation, and regression techniques, we have developed closed-form predictive expressions for pre- and post-Fickian flux-weighted solute breakthrough curves as a function of distance from the source (in integral scales) and variance of the log hydraulic conductivity field. Using the ensemble of Monte Carlo realizations, we have simultaneously computed error envelopes for the estimated flux-weighted breakthrough, and for the divergence of point breakthrough curves from the flux-weighted average, as functions of the predictive parameters. We have also obtained implied late-time macrodispersion coefficients for highly heterogeneous environments from the breakthrough statistics. This analysis is relevant for the modelling of reactive as well as conservative transport, since for many kinetic sorption and decay reactions, Laplace-domain modification of the breakthrough curve for conservative solute produces the correct curve for the reactive system.
Link to the AGU calendar.
| date | Thursday, 15 December 2016 |
| time | 14:25-14:40 |
| location | Moscone West - 3014 |
| number | H43N-04 |
| title of talk | “Prediction of Breakthrough Curves for Conservative and Reactive Transport from the Structural Parameters of Highly Heterogeneous Media” |
| Authors | Scott K. Hansen, Claus P. Haslauer, Olaf A. Cirpka, Velimir V. Vesselinov |