EKI

Demonstrate feasibility of ensemble Kalman filter based inversion of a simplified wave equation

Quick Facts

Type

Mini App

Status

Planned

Work Packages

WP4

Frameworks

PETSc, GeoS

1. Overview

Demonstrate feasibility of ensemble Kalman filter based inversion of a simplified wave equation

2. Technical Stack

2.1. Frameworks Used

  • PETSc

  • GeoS

2.2. Parallel Frameworks

  • MPI

3. Methods & Algorithms

3.1. WP1: Discretization

  • spectral element

  • dG/hdG

3.2. WP4: Inverse Problems & Data Assimilation

  • stochastic inverse problem

  • stochastic data assimilation: ensemble

3.3. WP6: Uncertainty Quantification

  • uncertainty propagation

4. Data Flow

4.1. Inputs

  • Gmsh

  • in-house

4.2. Outputs

  • in-house

5. Benchmarking

5.1. Metrics

  • Manufactured Solution

  • Weak Scalability

5.2. Benchmark Scope

  • Ensemble Mode

6. Timeline

Milestone Date

Specification Due

2026-10-01

Prototype Due

2027-01-15

7. Team

Partners: Inria BXSO, U Grenoble Alpes

Responsible: M. Asch