EMPIRE: Employing MPI for Researching Ensembles
A model developer walks into a bar and sees a data assimilator. "I have a model which I want to assimilate data into", says the model developer. "Add the EMPIRE functionality to your model and come back and buy me a drink" replies the data assimilator.
EMPIRE is the name given to a way of changing the source code of a dynamical model so that it can interface with sequential data assimilation methods.
EMPIRE should be one of the quickest and easiest ways in which to modify the source code of the model to use data assimilation.
Please contact us for help, specifically if there are any errors in the below documentation and source codes.
Follow this link to join the EMPIRE mailing list, for news, updates, questions etc.
EMPIRE Data Assimilation Codes
These are now available with full download instructions and documentation in html and pdf.
Models connected to EMPIRE
Implementation Documentation
| Document | File type | Date modified |
| Fortran implementation guide | 17 Feb 2015 | |
| Python implementation guide | 17 Feb 2015 |
Example model codes
The following codes are distributed under the MIT license. See below for details.
| Document | File type | Date modified |
| Fortran minimal example codes | .tar | 12 Mar 2015 |
| Fortran Lorenz 1963 model without EMPIRE | .f90 | 12 Nov 2014 |
| Fortran Lorenz 1963 model with EMPIRE | .f90 | 12 Mar 2015 |
| Python minimal example codes | .tar | 17 Feb 2015 |
| Python Lorenz 1963 model without EMPIRE | .py | 12 Nov 2014 |
| Python Lorenz 1963 model with EMPIRE | .py | 17 Feb 2015 |
| Python Lorenz 1963 model input file | .dat | 12 Nov 2014 |
| Python Lorenz 1996 model without EMPIRE | .py | 12 Nov 2014 |
| Python Lorenz 1996 model with EMPIRE | .py | 17 Feb 2015 |
| Python Lorenz 1996 model input file | .dat | 12 Nov 2014 |
