Integrating models and observations

OpenDA is an open interface standard for (and free implementation of) a set of tools to quickly implement data-assimilation and calibration for arbitrary numerical models. OpenDA wants to stimulate the use of data-assimilation and calibration by lowering the implementation costs and enhancing the exchange of software among researchers and end-users.

A model that conforms to the OpenDA standard can use all the tools that are available in OpenDA. This allows experimentation with  data-assimilation/calibration methods without the need for extensive programming. Reversely, developers of data-assimilation/calibration software that make their implementations compatible with the OpenDA interface will make their new methods usable for all OpenDA users (either for free or on a commercial basis).

OpenDA has been designed for high performance. Hence, even large-scale models can use it. Also, OpenDA allows users to optimize the interaction between their model and the data-assimilation/calibration methods. Hence, data-assimilation with OpenDA can be as efficient as with custom-made implementations of data-assimilation methods.

OpenDA is an Open Source project. Contributions are welcome from anyone wishing to participate in the further development of the OpenDA toolset.

OpenDA Webinar on Wednesday 11 April 2012 (17:00 CEST)

Deltares is proud to announce that Martin Verlaan (Deltares/TU Delft) and Stef Hummel (Deltares) will present the Webinar on Wednesday 11 April 2012 (17:00 CEST) entitled Delft3D numerical model calibration with OpenDA. In this OpenDA webinar, we will demonstrate how to use OpenDA for calibration of your Delft3D FLOW model parameters. Read more...

Release of version 2.0

The release of version 2.0 of OpenDA has been publicly announced in Februari 2012. OpenDA is available through sourceforge.net. There are binary versions available for 32 bit Linux and Windows. The complete sources also contain precompiled native components for 64 bit Linux and Mac.

Features of OpenDA Currently available methods Data-assimilation methods Ensemble KF (EnKF) Ensemble SquareRoot KF (EnSR) Steady State KF Particle Filter 3DVar DudEnKF (still under research) DudEnSR (still under research) Parameter estimation (calibration) methods: Dud Sparse Dud Simplex Powell Gridded full search Shuffled Comples Evolution (SCE) Generalized Likelihood Uncertainty Estimation (GLUE) (L)BFGS Conjugate Gradient: Fleetjer-Reeves, Polak-Ribiere, Steepest Descent Language interfaces C/C++ Java Fortran77/90

OpenDA is supported by