 |
| Titel |
Is Gaussian normal? |
| VerfasserIn |
J. Jaime Gómez-Hernández, Teng Xu |
| Konferenz |
EGU General Assembly 2014
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 16 (2014) |
| Datensatznummer |
250093596
|
| Publikation (Nr.) |
 EGU/EGU2014-8483.pdf |
|
|
|
|
|
| Zusammenfassung |
| For many years, hydraulic conductivity was considered to be lognormal. More so, in
stochastic hydrogeology it was postulated that the best multivariate model for logconductivity
was the multiGaussian one. However, evidence has proven that in many cases, hydraulic
conductivity is not normal and that modeling using a multiGaussian distribution can
introduce unwanted spatial patterns. This presentation will discuss some algorithms to
address the problem of non Gaussianity of logconductivity in the context of inverse
modeling. More precisely, the normal-score ensemble Kalman filter (NS-EnKF), and the
ensemble pattern matching (EnPAT) will be commented with its advantages and pitfalls. |
|
|
|
|
|
|