Features and Requirements

• PDAF is implemented in Fortran90 with some features from Fortran 2003. The standard interface also supports models that are written in other languages like C or C++.
• The parallelization uses the MPI (Message Passing Interface) standard. The localized filters use, in addition, OpenMP-parallelization.
• The core routines are fully independent of the model code. They can be compiled separately and can be used as a library.

Filter algorithms

PDAF provides the following algorithms for data assimilation. All filters are fully implemented, optimized and parallelized.

Local filters:

• LSEIK (Nerger et al., 2006)
• LETKF (Hunt et al., 2007)
• LESTKF (Local Error Subspace Transform Kalman Filter, Nerger et al., 2012, see publications)

Global filters:

• SEIK (Pham et al. (1998a, 2001), the implemented variant is described in more detail by Nerger et al. (2005))
• ETKF (The implementation follows Hunt et al. (2007) but without localization, which is available in the LETKF implementation)
• SEEK (The original formulation by Pham et al. (1998))
• EnKF (The classical formulation with perturbed observations by Evensen (1994), Burgers et al. (1998))
• ESTKF (Error Subspace Transform Kalman Filter, Nerger et al., 2012, see publications)

Starting from version 1.9 of PDAF, smoothers algorithms are provided for the following algorithms

• ESTKF & LESTKF
• ETKF & LETKF
• EnKF

Simplifying the implementation

PDAF simplifies the implementation of data assimilation systems using existing model code by the following:

1. PDAF provides fully implemented, parallelized, and optimized ensemble-based algorithms for data assimilation. Currently, these are ensemble-based Kalman filters like the LSEIK, LETKF, and EnKF methods.
2. PDAF is attached to the model source code by minimal changes to the code. These changes only concern the general part of the code, but not the numerics of the model. In addition, a small set of routines is required that are specific to the model or the observations to be assimilated. These routines can be implemented like routines of the model.
3. PDAF is called through a well-defined standard interface. This allows, for example, to switch between the LSEIK, LETKF, and LESTKF methods without additional coding.
4. PDAF provides parallelization support for the data assimilation system. If your numerical model is already parallelized, PDAF enables the data assimilation system to run several model tasks in parallel within a single executable. However, PDAF can also be used without parallelization, for example to test small systems.
5. PDAF does not require that your model can be called as a subroutine. Rather PDAF is added to the model and the formed data assimilation system can be executed pretty much like the model-program would without data assimilation.
6. PDAF also offers an offline mode. This is for the case that you don't want to (or even cannot) modify your model source code at all. In the offline mode, PDAF is compiled separately from the model together with the supporting routines to handle the observations. Then the model and the assimilation step are executed separately. While this strategy is possible, we don't recommend it, because it's computationally less efficient.

Requirements

• Compiler
  To compile PDAF a Fortran compiler is required. PDAF has been tested with a variety of compilers like gfortran, ifort, xlf, pgf90.
• BLAS and LAPACK
  The BLAS and LAPACK libraries are used by PDAF. For Linux there are usually packages with these libraries. With commercial compilers the functions are usually provided by optimized libraries (like MKL, ESSL).
• MPI (optional)
  If the assimilation program should be executed with parallelization, an MPI library is required (e.g. OpenMPI). The assimilation program can also be compiled and run without parallelization. For this, PDAF provides functions that mimic MPI operations for a single process.
• make
  PDAF provides Makefile definitions for different compilers and operating systems.

Test machines

PDAF has been tested on various machines with different compilers and MPI libraries. Current test machines include:

• IBM p575 with Power6 processors, AIX6.1, XLF compiler 12.1, ESSL library, POE parallel environment
• IBM BladeCenter with Power6 processors, AIX5.3, XLF compilers 10.1 to 13.1, ESSL library, POE parallel environment
• Linux Desktop machine, Ubuntu 12.04, ifort compiler
• Linux Desktop machine, Ubuntu 12.04, gfortran, OpenMPI
• Notebook Apple MacBook, Mac OS X 10.11, gfortran, OpenMPI
• SGI Altix UltraViolet, SLES 11 operating system, ifort compiler, SGI MPT
• Cray XC30 and XC40, Cray compiler CCE 8.3.5, MPICH

Test cases

The regular tests use a rather small configuration with a simulated model. This model is also included in the test suite of the downloadable PDAF package. In addition, the scalability of PDAF was examined with a real implementation with the finite element ocean model (FEOM, Danilov et al., A finite-element ocean model: Principles and evaluation. Ocean Modeling 6 (2004) 125-150). In these tests up to 4800 processor cores of a supercomputer have been used (see Nerger and Hiller (2013)). To examine PDAF's behavior with large-scale cases, experiments with the simulated model have been performed. By now the biggest case had a state dimension of 1.2^.10¹¹. An observation vector of size 2.4^.10⁹ was assimilated. For these experiments, the computations used 12000 processor cores. In this case, the distributed ensemble array occupied about 2 GBytes of memory for each core. The overall largest number of processes with which PDAF was validated was 16800 processor cores.