The ensemble Kalman filter (EnKF) and its many variants have been proven effective for data assimilation in large models, including those in atmospheric, oceanic, hydrologic, and petroleum reservoir systems. By bringing together technical experts, practitioners, researchers and students for presentations and informal interchange of information, we aim to share research results and suggest important challenges that have yet to be addressed.

Venue

Venue: May 2 - 5, Thon Hotel Sandven, Norheimsund, Norway

Program

Workshop program

The cultural event/excursion will likely consist of a hike to the Steindal waterfall (with an optional extension with some beautiful vistas), and a tasting session from a local apple cider farm. Another option will be a guided tour of the Hardanger Fartøyvernsenter (Maritime center), with the option of renting a rowboat thereafter.

Invited speakers

Ricardo Baptista
Caltech, MIT, USA 		Toward consistent nonlinear filtering and smoothing via measure transport
Feda Curic
Equinor, Norway 		Attempts at adaptively localising the Troll field
Geir Evensen
NORCE, Norway 		Consistent ensemble formulation and solution of the history-matching problem
Xiao-Hui Wu
ExxonMobil, USA 		Stein Variational Gradient Descent for Reservoir History Matching Problems

Background

The EnKF is a data assimilation method was co-invented and has been continuously developed by researchers at NORCE and NERSC. In the past two decades, the EnKF and related “ensemble” methods have been established as a school of viable and popular methods for data assimilation in very large models, and have made immense impacts on the advancements of various disciplines and promoted value creations for relevant industries.

Purpose

Although the basic concept ensemble methods is straightforward, successful practical implementations often require modifications that are problem-specific. By bringing together experts from diverse areas, we aim to explore the bases and connections among EnKF-related methods that are proven to work in different environments, so that the resulting applications are more robust and efficient. Other than the aspect of practical applications, this workshop also aims to exchange and communicate novel research ideas, methods, algorithms and/or workflows that have the potential of further improving the performance of EnKF and its related methods.

History

The first EnKF workshop took place in Voss (Norway) in 2006. Since then, the EnKF workshop has been held annually (except for a disruption in 2020 due to COVID-19). The EnKF workshop has always accommodated participants coming from diverse scientific disciplines (e.g., meteorology, oceanography, hydrology, to name a few). The communication and exchange of scientific progresses and advancements have led to even more fruitful discussions and raised the scientific quality of the workshop to a very high level. As such, the annual international EnKF workshop has now become one of the most influential events within the data assimilation community.

Sponsors

REMEDY

Please contact enkf@data-assimilation.no if you want to sponsor the workshop.

Scientific committee

  • NORCE: Geir Evensen, Dean Oliver, Patrick N. Raanes
  • NERSC: Laurent Bertino
  • Equinor: Remus Hanea

Photos

Group photo from outside hotel Dinner Hiking Hiking Talk by Geir Talk by Mathias Talk by Lars Cider tasting session Discussion panel 1 Discussion panel 2 At the Steinsdalsfossen waterfall At the Steinsdalsfossen waterfall
Photos from the workshop (click for larger image)

Presentations

Patrick N. Raanes talk 'Ensemblized' linear least squares (LLS)
Maximilian Ramgraber talk Adaptive localization in nonlinear ensemble transport filtering
Maximilian Ramgraber poster Adaptive localization in nonlinear ensemble transport filtering
Maximilian Ramgraber script Supplementary material: how I generate my histograms.
Moha Gharamti talk A Randomized Dormant Ensemble Kalman Filter: Dealing with Extreme Sampling Errors
Moha Gharamti poster A Quantile Conserving Ensemble Filtering Framework: Next Generation Nonlinear and Non-Gaussian Data Assimilation Capabilities for DART
Mathieu Le Provost talk An ensemble filter for heavy tailed t-distributions
Mathieu Le Provost poster Regularization of the ensemble Kalman filter for non-local observations: application to elliptic observations
Hamed Ali Diab Montero poster A Particle Flow Filter for Estimating Future Earthquake Occurrences
Samantha S.R. Kim poster Optimal proposal for the assimilation of geodetic data with a particle filter: an application to compacting reservoirs
Antoine Bernigaud poster Lp-norm regularization in variational data assimilation
Xiaoling Jin poster Ensemble Kalman-based data-driven identification of stochastic model from state data
Heng Xiao poster Inference of relative permeability curves in reservoir rocks with ensemble Kalman method
Manon Verberne poster Ensemble smoother with multiple data assimilation to disentangle shallow and deep subsidence
Rolf J. Lorentzen poster Ensemble-based history matching of the Edvard Grieg field using 4D seismic data
Kjersti Solberg Eikrem poster Ensemble-based optimization applied to an offshore wind farm layout problem
Geir Evensen talk Consistent ensemble formulation and solution of the history-matching problem
Mathias Methlie Nilsen talk An Exponential Class of Ensemble based Optimization Algorithms
Lars Nerger talk A hybrid nonlinear-Kalman ensemble transform filter for data assimilation in systems with different degrees of nonlinearity
Lars Nerger poster Coupled assimilation of satellite temperature and chlorophyll observations for improved ecosystem predictions in the Baltic Sea
Femke Vossepoel talk Quantifying information loss in a Particle Method for Subsidence Estimation
Luxi Yu talk Soft sensing of intracellular states in bioprocessing with Ensemble Kalman Filters
Luxi Yu poster Soft sensing of intracellular states in bioprocessing with Ensemble Kalman Filters
Yue Ying talk Assimilating observations of deformation to improve short-term ensemble forecasts of sea ice features
Sophie Mauran talk A kernel extension of the Ensemble Transform Kalman Filter
Jiping Xie talk Clutch effect in an Arctic coupled ocean and sea ice data assimilation system
Marina Durán Moro talk Data assimilation of SIC satellite observations in the Barents Sea region
Ricardo Baptista talk Toward consistent nonlinear filtering and smoothing via measure transport
Feda Curic talk Attempts at adaptively localising the Troll field
Andreas Størksen Stordal talk Marginalized Iterative Ensemble Smoothers for Data assimilation
Xiao-Hui Wu talk Stein Variational Gradient Descent for Reservoir History Matching Problems
Xiaodong Luo talk Hyper parameter optimization for improving the performance of localization in an iterative ensemble smoother