Coupled earth system modeling on heterogeneous HPC architectures with ParFlow in the Terrestrial Systems Modeling Platform

2021 ◽  
Author(s):  
Jaro Hokkanen ◽  
Stefan Kollet ◽  
Jiri Kraus ◽  
Andreas Herten ◽  
Markus Hrywniak ◽  
...  
Keyword(s):  
Land Surface ◽  
Memory Management ◽  
High Performance ◽  
Data Exchange ◽  
System Modeling ◽  
Hydrologic Model ◽  
Earth System ◽  
Single Node ◽  
Earth System Modeling ◽  
Domain Specific

<p>Rapidly changing heterogeneous supercomputer architectures pose a great challenge to many scientific communities trying to leverage the latest technology in high-performance computing. Implementations that simultaneously result in a good performance and developer productivity while keeping the codebase adaptable and well maintainable in the long-term are of high importance. ParFlow, a widely used hydrologic model, achieves these attributes by hiding the architecture-dependent code in preprocessor macros (ParFlow embedded Domain Specific Language, eDSL) and leveraging NVIDIA's Unified Memory technology for memory management. The implementation results in very good weak scaling with up to 26x speedup when using four NVIDIA A100 GPUs per node compared to using the available 48 CPU cores. Good weak scaling is observed using hundreds of nodes on the new JUWELS Booster system at the Jülich Supercomputing Centre, Germany. Furthermore, it is possible to couple ParFlow with other earth system compartment models such as land surface and atmospheric models using the OASIS-MCT coupler library, which handles the data exchange between the different models. The ParFlow GPU implementation is fully compatible with the coupled implementation with little changes to the source code. Moreover, coupled simulations offer interesting load-balancing opportunities for optimal usage of the existing resources. For example, running ParFlow on GPU nodes, and another application component on CPU-only nodes, or efficiently distributing the CPU and GPU resources of a single node between the different application components may result in the best usage of heterogeneous architectures.</p>

scholarly journals Leveraging HPC accelerator architectures with modern techniques — hydrologic modeling on GPUs with ParFlow

2021 ◽  
Author(s):  
Jaro Hokkanen ◽  
Stefan Kollet ◽  
Jiri Kraus ◽  
Andreas Herten ◽  
Markus Hrywniak ◽  
...  
Keyword(s):  
Memory Management ◽  
High Performance ◽  
Hydrologic Model ◽  
Third Party ◽  
Domain Specific ◽  
Scientific Institutions ◽  
Significant Performance ◽  
Performance Computing ◽  
Accelerator Architectures

AbstractRapidly changing heterogeneous supercomputer architectures pose a great challenge to many scientific communities trying to leverage the latest technology in high-performance computing. Many existing projects with a long development history have resultedin a large amount of code that is not directly compatible with the latest accelerator architectures. Furthermore, due to limited resources of scientific institutions, developing and maintaining architecture-specific ports is generally unsustainable. In order to adapt to modern accelerator architectures, many projects rely on directive-based programming models or build the codebase tightly around a third-party domain-specific language or library. This introduces external dependencies out of control of theproject. The presented paper tackles the issue by proposing a lightweight application-side adaptor layer for compute kernels and memory management resulting in a versatile and inexpensive adaptation of new accelerator architectures with little drawbacks.A widely used hydrologic model demonstrates that such an approach pursued more than 20 years ago is still paying off with modern accelerator architectures as demonstrated by a very significant performance gain from NVIDIA A100 GPUs, high developer productivity, and minimally invasive implementation; all while the codebase is kept well maintainable in the long-term.

scholarly journals Technical Note: The Modular Earth Submodel System (MESSy) - a new approach towards Earth System Modeling

2005 ◽  
Vol 5 (2) ◽  
pp. 433-444 ◽  
Author(s):  
P. Jöckel ◽  
R. Sander ◽  
A. Kerkweg ◽  
H. Tost ◽  
J. Lelieveld
Keyword(s):  
General Circulation ◽  
System Modeling ◽  
Circulation Model ◽  
Atmospheric Model ◽  
Ocean Model ◽  
Technical Note ◽  
Large Set ◽  
Earth System ◽  
Earth System Modeling ◽  
Domain Specific

Abstract. The development of a comprehensive Earth System Model (ESM) to study the interactions between chemical, physical, and biological processes, requires coupling of the different domains (land, ocean, atmosphere, ...). One strategy is to link existing domain-specific models with a universal coupler, i.e. an independent standalone program organizing the communication between other programs. In many cases, however, a much simpler approach is more feasible. We have developed the Modular Earth Submodel System (MESSy). It comprises (1) a modular interface structure to connect to a , (2) an extendable set of such for miscellaneous processes, and (3) a coding standard. MESSy is therefore not a coupler in the classical sense, but exchanges data between a and several within one comprehensive executable. The internal complexity of the is controllable in a transparent and user friendly way. This provides remarkable new possibilities to study feedback mechanisms (by two-way coupling). Note that the MESSy and the coupler approach can be combined. For instance, an atmospheric model implemented according to the MESSy standard could easily be coupled to an ocean model by means of an external coupler. The vision is to ultimately form a comprehensive ESM which includes a large set of submodels, and a base model which contains only a central clock and runtime control. This can be reached stepwise, since each process can be included independently. Starting from an existing model, process submodels can be reimplemented according to the MESSy standard. This procedure guarantees the availability of a state-of-the-art model for scientific applications at any time of the development. In principle, MESSy can be implemented into any kind of model, either global or regional. So far, the MESSy concept has been applied to the general circulation model ECHAM5 and a number of process boxmodels.

High-performance Earth system modeling with NASA/GSFC’s Land Information System

2007 ◽  
Vol 3 (3) ◽  
pp. 157-165 ◽  
Author(s):  
C. D. Peters-Lidard ◽  
P. R. Houser ◽  
Y. Tian ◽  
S. V. Kumar ◽  
J. Geiger ◽  
...  
Keyword(s):  
Information System ◽  
High Performance ◽  
System Modeling ◽  
Earth System ◽  
Earth System Modeling ◽  
Land Information System

scholarly journals Technical Note: The Modular Earth Submodel System (MESSy) – a new approach towards Earth System Modeling

2004 ◽  
Vol 4 (6) ◽  
pp. 7139-7166
Author(s):  
P. Jöckel ◽  
R. Sander ◽  
J. Lelieveld
Keyword(s):  
General Circulation ◽  
System Modeling ◽  
Circulation Model ◽  
Technical Note ◽  
Biological Processes ◽  
Earth System ◽  
Earth System Modeling ◽  
New Approach ◽  
Domain Specific ◽  
User Friendly

Abstract. Generally, the typical approach towards Earth System Modeling has been to couple existing models of different domains (land, ocean, atmosphere, ...) offline, using output files of one model to provide input for the other. However, for a detailed study of the interactions and feedbacks between chemical, physical, and biological processes, it is necessary to perform the coupling online. One strategy is to link the existing domain-specific models with a universal coupler. In many cases, however, a much simpler approach is more feasible. To achieve the online coupling, we have developed the Modular Earth Submodel System (MESSy). Data are exchanged between a and several within one comprehensive model system. MESSy includes a generalized interface structure for the standardized control of the and their interconnections. The internal complexity of the is controllable in a transparent and user friendly way. This provides remarkable new possibilities to study feedback mechanisms (by two-way coupling), e.g., by applying MESSy to a general circulation model (GCM).

scholarly journals Seasonal to multi-year soil moisture drought forecasting

2021 ◽  
Vol 4 (1) ◽  
Author(s):  
Musa Esit ◽  
Sanjiv Kumar ◽  
Ashutosh Pandey ◽  
David M. Lawrence ◽  
Imtiaz Rangwala ◽  
...  
Keyword(s):  
Soil Moisture ◽  
North America ◽  
Land Surface ◽  
System Modeling ◽  
Precipitation Variability ◽  
Surface Processes ◽  
Earth System ◽  
Drought Forecasting ◽  
Earth System Modeling ◽  
Order Of Magnitude

AbstractSoil moisture predictability on seasonal to decadal (S2D) continuum timescales over North America is examined from the Community Earth System Modeling (CESM) experiments. The effects of ocean and land initializations are disentangled using two large ensemble datasets—initialized and uninitialized experiments from the CESM. We find that soil moisture has significant predictability on S2D timescales despite limited predictability in precipitation. On sub-seasonal to seasonal timescales, precipitation variability is an order of magnitude greater than soil moisture, suggesting land surface processes, including soil moisture memory, reemergence, land–atmosphere interactions, transform a less predictable precipitation signal into a more predictable soil moisture signal.

Sign in / Sign up

Export Citation Format

Share Document