Skip to content

cgrudz/DataAssimilationBenchmarks.jl

Repository files navigation

DataAssimilationBenchmarks.jl

DataAssimilationBenchmarks.jl logo

docs-dev-img docs-stable-img status Total lines of code without comments Build Status codecov

Lines of code counter (without comments or blank lines) courtesy of Tokei.

Welcome to DataAssimilationBenchmarks.jl!

Description

This is a data assimilation research code base with an emphasis on prototyping, testing and validating sequential filters and smoothers in toy model twin experiments. This code is meant to be performant in the sense that large hyper-parameter discretizations can be explored to determine hyper-parameter sensitivity and reliability of results across different experimental regimes, with parallel implementations in native Julia distributed computing.

This package currently includes code for developing and testing data assimilation schemes in the L96-s model and the IEEE 39 bus test case in the form of the effective network model model equations. New toy models and data assimilation schemes are in continuous development in the development branch. Currently validated techniques are available in the master branch.

This package supported the development of all numerical results and benchmark simulations in the manuscript A fast, single-iteration ensemble Kalman smoother for sequential data assimilation.

Documentation

Please see the up-to-date documentation synchronized with the master branch or the stable documentation for the last tagged version in the Julia General Registries.