mengjintao/enkf-c

EnKF code for DA with large-scale layered geophysical models.

EnKF-C

EnKF-C provides a compact generic framework for off-line data assimilation (DA) into large-scale layered geophysical models with the ensemble Kalman filter (EnKF). Following are its other main features:

• coded in C for GNU/Linux platform;

• model-agnostic;

• can conduct DA either in EnKF or ensemble optimal interpolation (EnOI) mode;

• permits multiple model grids;

• can handle rectangular or curvilinear horizontal grids, z, sigma or hybrid vertical grids.

The code is designed to handle as large systems as possible, using shared memory capabilities of MPI-3. Here is a snapshot of ensemble spread of sea surface temperature from the 96-member EnKF ocean forecasting system with MOM5 based OFAM3 model (51 x 1500 x 3600 grid), assimilating about 14M super-observations at each 3-day cycle.

For more information see README and user guide. (An older version of the user guide is also available from arXiv.) Have a feel for how the code works by running the included example.

Checkout EnKF-C by running git clone https://github.com/sakov/enkf-c or svn checkout https://github.com/sakov/enkf-c.