Climate research looks at weather patterns and evaluates the interaction of elements such as atmosphere, land surface, ocean and sea ice systems. Climate modeling research uses sophisticated mathematical algorithms and complex calculations on high-performance supercomputers to gain a better understanding of weather patterns and to predict changes in the Earth’s climate.
The U.S. Department of Energy (DOE) is at the forefront of climate research and climate modeling. DOE has a Climate Modeling Program whose mission is to improve climate change projections using state-of-the-science coupled climate and earth system models, on time scales of decades to centuries and spatial scales of global to regional. DOE plays a vital and unique role in the climate modeling enterprise in the U.S., primarily through two offices within DOE’s Office of Science: the Office of Biological and Environmental Research (BER) as well as the Office of Advanced Scientific Computing Research (ASCR).
Climate Modeling at DOE
The climate modeling program in the DOE Office of Science sponsors projects that develop, test, and apply state-of-the-science coupled climate and earth system models, based on theoretical climate change science foundations. According to Anjuli Bamzai DOE Program Manager, Climate Change Prediction Program, Climate and Environmental Sciences Division, “In order to enable sound decision-making on issues pertaining to future energy use and technology options, credible high-resolution climate change simulations are required at a regional scale. To achieve such high-resolution simulations, both the accuracy and throughput need to be dramatically increased; thus the climate modeling activity takes advantage of emerging high performance computing (HPC) and information technologies. An example of climate modeling using HPC and supercomputers can be found in the DOE Leadership-class Computing Facility.”
Why Climate Modeling is Important
Bamzai indicates that improved climate information at high spatial and temporal resolution is of immense significance to society and decision makers. Understanding future variability and predictability of the climate system, such as changes in major modes of climate variability, climate extremes, detecting and attributing the regional manifestations of climate change, remain significant challenges.
Specialized Facilities and Labs used in Research
There are a variety of universities, national laboratories and special computer facilities involved in climate modeling research. ASCR delivers leading edge computational and networking capabilities to scientists nationwide, enabling advances in computer science and the development of specialized software tools necessary to research major scientific questions in climate change science. Climate model simulations are currently being undertaken at the DOE Leadership-class Computing Facility (LCF) at Oak Ridge National Laboratory and Argonne National Laboratory. In addition, the National Energy Research Scientific Computing (NERSC) provides computational and computing support for climate researchers. Computing time is awarded to research groups based on peer review of submitted proposals. Basic research accomplished at these facilities covers a wide range of disciplines including climate modeling.
The Energy Sciences Network (ESnet) enables researchers at laboratories, universities and other institutions to communicate with each other using collaborative capabilities that are unparalleled. This high-speed network enables geographically distributed research teams to collaborate effectively on some of the world’s most complex problems. BER and ASCR, through its Computational Partnerships program, partner to support Scientific Discovery Through Advanced Computing (SciDAC), a unique program that provides the innovations in computational research and development for petascale computational and data management endeavors, including climate research.
Importance of HPC and Supercomputers in Climate Modeling Research
Climate modeling research requires complex algorithms which must run on state-of-the-art high performance super computers. Bamzai indicates that in climate modeling, throughput refers to the number of simulated years in a day; it’s measure of productivity of the code, so to speak. One may construct an earth system model that includes perfect physics, biology and dynamics, however, if careful attention is not paid to the computational aspects, and throughput is low, it would be a futile exercise since the simulation results would not be obtained in a timely manner. “Thus in order for climate and earth system models to be useful tools to study and assess long-term climate change, judicious use of HPC and supercomputers need to be invoked and integrated in the research plan. As we enter an exciting era of innovative technological advances in HPC, we can anticipate new discoveries and breakthroughs in climate change science,” states Bamzai.
Linda Barney owns Barney and Associates, a technical and marketing writing firm in Beaverton, Oregon that provides writing and web content for the high tech, government, medical and scientific communities. They specialize in presenting deeply technical information so that it is interesting and easy to understand. Readers can reach her at [email protected]