scholarly journals Scalability and some optimization of the Finite-volumE Sea ice-Ocean Model, Version 2.0 (FESOM2)

2019 ◽  
Author(s):  
Nikolay V. Koldunov ◽  
Vadym Aizinger ◽  
Natalja Rakowsky ◽  
Patrick Scholz ◽  
Dmitry Sidorenko ◽  
...  

Abstract. A study of the scalability of the Finite-volumE Sea ice-Ocean circulation Model, Version 2.0 (FESOM2), the first mature global model of its kind formulated on unstructured meshes, is presented. This study includes an analysis of main computational kernels with a special focus on bottlenecks in parallel scalability. Several model enhancements, improving this scalability for large numbers of processes, are described and tested. Model grids at different resolutions are used on four HPC systems with differing computation and communication hardware to demonstrate model's scalability and throughput. Furthermore, strategies for improvements in parallel performance are presented and assessed. We show that in terms of throughput FESOM2.0 is on par with the state-of-the-art structured ocean models and in realistic eddy resolving configuration (1/10° resolution) can produce about 16 years per day on 14 000 cores. This suggests that unstructured-mesh models are becoming extremely competitive tools in high-resolution climate modelling. It is shown that main bottlenecks of FESOM parallel scalability are the two-dimensional components of the model, namely the computations of external (barotropic) mode and the sea-ice model. It is argued that these bottlenecks are shared with other general ocean circulation models.

2019 ◽  
Vol 12 (9) ◽  
pp. 3991-4012 ◽  
Author(s):  
Nikolay V. Koldunov ◽  
Vadym Aizinger ◽  
Natalja Rakowsky ◽  
Patrick Scholz ◽  
Dmitry Sidorenko ◽  
...  

Abstract. A study of the scalability of the Finite-volumE Sea ice–Ocean circulation Model, Version 2.0 (FESOM2), the first mature global model of its kind formulated on unstructured meshes, is presented. This study includes an analysis of the main computational kernels with a special focus on bottlenecks in parallel scalability. Several model enhancements improving this scalability for large numbers of processes are described and tested. Model grids at different resolutions are used on four high-performance computing (HPC) systems with differing computational and communication hardware to demonstrate the model's scalability and throughput. Furthermore, strategies for improvements in parallel performance are presented and assessed. We show that, in terms of throughput, FESOM2 is on a par with state-of-the-art structured ocean models and, in a realistic eddy-resolving configuration (1/10∘ resolution), can achieve about 16 years per day on 14 000 cores. This suggests that unstructured-mesh models are becoming very competitive tools in high-resolution climate modeling. We show that the main bottlenecks of FESOM2 parallel scalability are the two-dimensional components of the model, namely the computations of the external (barotropic) mode and the sea-ice model. It is argued that these bottlenecks are shared with other general ocean circulation models.


2019 ◽  
Author(s):  
Patrick Scholz ◽  
Dmitry Sidorenko ◽  
Ozgur Gurses ◽  
Sergey Danilov ◽  
Nikolay Koldunov ◽  
...  

Abstract. The evaluation and model element description of the second version of the unstructured-mesh Finite-volumE Sea ice–Ocean circulation Model (FESOM2.0) is presented. The model sensitivity to arbitrary Lagrangian Eulerian (ALE) linear and nonlinear free surface formulation, Gent McWilliams eddy parameterisation, isoneutral Redi diffusion and different vertical mixing schemes is documented. The hydrographic biases, large scale circulation, numerical performance and scalability of FESOM2.0 are compared with its predecessor FESOM1.4. FESOM2.0 shows biases with a magnitude comparable to FESOM1.4 and it simulates a more realistic AMOC. Compared to its predecessor FESOM2.0 provides clearly defined fluxes and a three times higher throughput in terms of simulated years per day (SYPD). It is thus the first mature global unstructured-mesh ocean model with computational efficiency comparable to state-of-the-art structured-mesh ocean models. Other key elements of the model and new development will be described in following-up papers.


2016 ◽  
Author(s):  
Sergey Danilov ◽  
Dmitry Sidorenko ◽  
Qiang Wang ◽  
Thomas Jung

Abstract. Version 2 of the unstructured-mesh sea ice – ocean circulation model FESOM is presented. It builds upon FESOM1.4 (Wang et al., 2014, Geosci. Mod. Dev., 7, 663–693) but differs by its dynamical core (finite volumes instead of finite elements) and is formulated using the Arbitrary Lagrangian Eulerian (ALE) vertical coordinate, which increases model flexibility. The model inherits the framework and sea ice model from the previous version, which minimizes the efforts needed from a user to switch from one version to the other. The ocean states simulated with FESOM1.4 and FESOM2.0 driven by CORE-II forcing are compared on a mesh used for CORE-II intercomparison project. Additionally the performance on an eddy-permitting mesh with uniform resolution is discussed. The new version improves numerical efficiency of FESOM in terms of CPU time by at least three times while retaining its fidelity in simulating sea ice and ocean. From this it is argued that FESOM2.0 provides a major step forward in establishing unstructured-mesh models as valuable tools in climate research.


2013 ◽  
Vol 6 (3) ◽  
pp. 3893-3976 ◽  
Author(s):  
Q. Wang ◽  
S. Danilov ◽  
D. Sidorenko ◽  
R. Timmermann ◽  
C. Wekerle ◽  
...  

Abstract. The Finite Element Sea ice-Ocean Model (FESOM) is the first global ocean general circulation model based on unstructured-mesh methods that has been developed for the purpose of climate research. The advantage of unstructured-mesh models is their flexible multi-resolution modelling functionality. In this study, an overview of the main features of FESOM will be given; based on sensitivity experiments a number of specific parameter choices will be explained; and directions of future developments will be outlined. It is argued that FESOM is sufficiently mature to explore the benefits of multi-resolution climate modelling and that it provides an excellent platform for further developments required to advance the field of climate modelling on unstructured meshes.


2014 ◽  
Vol 7 (2) ◽  
pp. 663-693 ◽  
Author(s):  
Q. Wang ◽  
S. Danilov ◽  
D. Sidorenko ◽  
R. Timmermann ◽  
C. Wekerle ◽  
...  

Abstract. The Finite Element Sea Ice-Ocean Model (FESOM) is the first global ocean general circulation model based on unstructured-mesh methods that has been developed for the purpose of climate research. The advantage of unstructured-mesh models is their flexible multi-resolution modelling functionality. In this study, an overview of the main features of FESOM will be given; based on sensitivity experiments a number of specific parameter choices will be explained; and directions of future developments will be outlined. It is argued that FESOM is sufficiently mature to explore the benefits of multi-resolution climate modelling and that its applications will provide information useful for the advancement of climate modelling on unstructured meshes.


2017 ◽  
Vol 10 (2) ◽  
pp. 765-789 ◽  
Author(s):  
Sergey Danilov ◽  
Dmitry Sidorenko ◽  
Qiang Wang ◽  
Thomas Jung

Abstract. Version 2 of the unstructured-mesh Finite-Element Sea ice–Ocean circulation Model (FESOM) is presented. It builds upon FESOM1.4 (Wang et al., 2014) but differs by its dynamical core (finite volumes instead of finite elements), and is formulated using the arbitrary Lagrangian Eulerian (ALE) vertical coordinate, which increases model flexibility. The model inherits the framework and sea ice model from the previous version, which minimizes the efforts needed from a user to switch from one version to the other. The ocean states simulated with FESOM1.4 and FESOM2.0 driven by CORE-II forcing are compared on a mesh used for the CORE-II intercomparison project. Additionally, the performance on an eddy-permitting mesh with uniform resolution is discussed. The new version improves the numerical efficiency of FESOM in terms of CPU time by at least 3 times while retaining its fidelity in simulating sea ice and the ocean. From this it is argued that FESOM2.0 provides a major step forward in establishing unstructured-mesh models as valuable tools in climate research.


1997 ◽  
Vol 25 ◽  
pp. 111-115 ◽  
Author(s):  
Achim Stössel

This paper investigates the long-term impact of sea ice on global climate using a global sea-ice–ocean general circulation model (OGCM). The sea-ice component involves state-of-the-art dynamics; the ocean component consists of a 3.5° × 3.5° × 11 layer primitive-equation model. Depending on the physical description of sea ice, significant changes are detected in the convective activity, in the hydrographic properties and in the thermohaline circulation of the ocean model. Most of these changes originate in the Southern Ocean, emphasizing the crucial role of sea ice in this marginally stably stratified region of the world's oceans. Specifically, if the effect of brine release is neglected, the deep layers of the Southern Ocean warm up considerably; this is associated with a weakening of the Southern Hemisphere overturning cell. The removal of the commonly used “salinity enhancement” leads to a similar effect. The deep-ocean salinity is almost unaffected in both experiments. Introducing explicit new-ice thickness growth in partially ice-covered gridcells leads to a substantial increase in convective activity, especially in the Southern Ocean, with a concomitant significant cooling and salinification of the deep ocean. Possible mechanisms for the resulting interactions between sea-ice processes and deep-ocean characteristics are suggested.


2006 ◽  
Vol 3 (3) ◽  
pp. 637-669 ◽  
Author(s):  
S. Natale ◽  
R. Sorgente ◽  
S. Gaberšek ◽  
A. Ribotti ◽  
A. Olita

Abstract. Ocean forecasts over the Central Mediterranean, produced by a near real time regional scale system, have been evaluated in order to assess their predictability. The ocean circulation model has been forced at the surface by a medium, high or very high resolution atmospheric forcing. The simulated ocean parameters have been compared with satellite data and they were found to be generally in good agreement. High and very high resolution atmospheric forcings have been able to form noticeable, although short-lived, surface current structures, due to their ability to detect transient atmospheric disturbances. The existence of the current structures has not been directly assessed due to lack of measurements. The ocean model in the slave mode was not able to develop dynamics different from the driving coarse resolution model which provides the boundary conditions.


Sign in / Sign up

Export Citation Format

Share Document