Efficient ensemble data assimilation for coupled models with the Parallel Data Assimilation Framework: example of AWI-CM (AWI-CM-PDAF 1.0)

Abstract

Abstract. Data assimilation integrates information from observational measurements with numerical models. When used with coupled models of Earth system compartments, e.g., the atmosphere and the ocean, consistent joint states can be estimated. A common approach for data assimilation is ensemble-based methods which utilize an ensemble of state realizations to estimate the state and its uncertainty. These methods are far more costly to compute than a single coupled model because of the required integration of the ensemble. However, with uncoupled models, the ensemble methods also have been shown to exhibit a particularly good scaling behavior. This study discusses an approach to augment a coupled model with data assimilation functionality provided by the Parallel Data Assimilation Framework (PDAF). Using only minimal changes in the codes of the different compartment models, a particularly efficient data assimilation system is generated that utilizes parallelization and in-memory data transfers between the models and the data assimilation functions and hence avoids most of the file reading and writing, as well as model restarts during the data assimilation process. This study explains the required modifications to the programs with the example of the coupled atmosphere–sea-ice–ocean model AWI-CM (AWI Climate Model). Using the case of the assimilation of oceanic observations shows that the data assimilation leads only to small overheads in computing time of about 15 % compared to the model without data assimilation and a very good parallel scalability. The model-agnostic structure of the assimilation software ensures a separation of concerns in which the development of data assimilation methods can be separated from the model application.

Data Assimilation of Volcanic Aerosols using FALL3D+PDAF

“…In this work, the model state for each ensemble member is propagated by the FALL3D dispersal model. The DA system builds upon an efficient implementation by coupling FALL3D and the Parallel Data Assimilation Framework (PDAF), an opensource software environment for ensemble data assimilation providing fully implemented and optimised data assimilation algorithms, including ensemble Kalman filters (KF) such as EnKF, ETKF, and LETKF (Nerger et al, 2005(Nerger et al, , 2020, see also Sect. A).…”

“…New developments have led to improved quality of forecasts, enabled to quantify model uncertainties, and laid the foundations for the incorporation of ensemble-based DA techniques into future releases of FALL3D. This work presents a new data assimilation system based on the coupling between FALL3D and the Parallel Data Assimilation Framework (PDAF, Nerger et al, 2005Nerger et al, , 2020, available in the last code release (version 8.2) of FALL3D. The proposed methodology can be efficiently implemented in operational environments by exploiting High Performance Computing (HPC) resources.…”

Developing a common, flexible and efficient framework for weakly coupled ensemble data assimilation based on C-Coupler2.0

“…The ensemble Kalman filter (EnKF; Houtekamer and Mitchell, 1998;Evensen, 2003;Lorenc, 2003a;Anderson and Collins, 2007;Whitaker, 2012) is a widely used DA method that depends on an ensemble run of members. Other DA methods such as the nudging method (Hoke and Anthes, 1976;Vidard et al, 2003), optimal interpolation (OI; Gandin, 1966), ensemble OI (EnOI; Oke et al, 2002;Evensen, 2003), three-dimensional variational analysis (3D-Var; Anderson et al, 1998;Courtier et al, 1998;Gauthier et al, 1999;Lorenc et al, 2000) and four-dimensional variational analysis (4D-Var; Courtier et al, 1994;Kalnay, 2002;Lorenc, 2003b;Rabier et al, 2007) can be technically viewed as a special case of ensemble-based methods with only one member in the ensemble when we attempt to design and develop a software framework for data assimilation. Moreover, hybrid DA methods, such as hybrid ensemble and 3D-Var (Hamill, 2000;Etherton and Bishop, 2004; Wang et 2636 C. Sun et al: Weakly coupled ensemble data assimilation based on C-Coupler2.0 , 2013Ma et al, 2014) and ensemble-based 4D-Var schemes (Fisher, 2003;Bishop and Hodyss, 2011;Bonavita et al, 2012Bonavita et al, , 2016Buehner et al, 2015), also depend on the ensemble run of members from the same model.…”

“…Such an implementation (called offline implementation hereafter) can guarantee software independence between the models and the DA methods, so as to achieve flexibility and convenience in software integration; however, the extra I/O accesses of disk files, as well as the extra initialization of software modules introduced by the data exchange and the restarts, are time-consuming and can be a severe performance bottleneck under finer model resolution (Heinzeller et al, 2016;Craig et al, 2017). The Parallel Data Assimilation Framework (PDAF; Nerger et al, 2005;Nerger and Hiller, 2013;Nerger et al, 2020) and the Employing Message Passing Interface for Researching Ensembles (EMPIRE; Browne and Wilson, 2015) framework have shown that MPI (Message Passing Interface)-based data exchanges between the model ensemble members and DA pro-cedures can produce better performance for DA systems because they do not require disk files or the restart operations.…”

Comment on acp-2021-747

“…This work presents a new data assimilation system based on the coupling between FALL3D and the Parallel Data Assimilation Framework (PDAF, Nerger et al, 2005Nerger et al, , 2020, available in the last code release (version 8.2) of FALL3D. The proposed methodology can be efficiently implemented in operational environments by exploiting High Performance Computing (HPC) resources.…”