ERT - Ensemble based Reservoir Tool - is designed for running ensembles of dynamical models such as reservoir models, in order to do sensitivity analysis and data assimilation. ERT supports data assimilation using the Ensemble Smoother (ES), Ensemble Smoother with Multiple Data Assimilation (ES-MDA) and Iterative Ensemble Smoother (IES).
ERT currently supports localization via a workflow defined in semeio [1] that uses an experimental feature called row_scaling [2]. This method is quite complicated to set-up and is for that and probably other reasons not begin used.
Localization is an important part of using ensemble based methods so ERT needs a decent implementation.
The goal of this milestone is to provide ERT with good localization capabilities.
Correlation based localization was implemented and tested in this feature branch:
https://github.com/equinor/ert/pull/4243
Based on this implementation and studies done by NORCE and the Troll-field, correlation based localization is deemed useful and is therefore to be implemented into ERT.
ERT currently supports localization via a workflow defined in
semeio
[1] that uses an experimental feature calledrow_scaling
[2]. This method is quite complicated to set-up and is for that and probably other reasons not begin used. Localization is an important part of using ensemble based methods so ERT needs a decent implementation. The goal of this milestone is to provide ERT with good localization capabilities.Correlation based localization was implemented and tested in this feature branch: https://github.com/equinor/ert/pull/4243 Based on this implementation and studies done by NORCE and the Troll-field, correlation based localization is deemed useful and is therefore to be implemented into ERT.
[1] https://github.com/equinor/semeio/tree/main/semeio/workflows/localisation [2] https://github.com/equinor/iterative_ensemble_smoother/blob/main/src/iterative_ensemble_smoother/experimental.py
Issue list: https://github.com/equinor/ert/milestone/23