Today’s weather and climate scientists are tasked with analyzing a massive tidal wave of data in order to better understand and predict significant changes affecting the climate. Whether trying to predict the future implications of global warming or studying the effects of human activities on the environment, climatologists are awash with data promising to drive new levels of innovation in the field of climate research.
Research centers are increasingly leveraging advanced computing systems to help them support large data volumes and quickly reveal new insights. In climate science, data must be gathered from a variety of sources including atmospheric and oceanographic records, geological datasets like borehole temperatures, floral and faunal histories, and records of past sea levels. These sources are producing tremendously large and complex datasets, which require high performance computing (HPC) solutions capable of supporting very large databases and executing compute-intensive models. These powerful machines are helping climate researchers improve the accuracy of their models, understand past and present climate trends, and better predict future climate changes.
Supercomputing systems are currently in widespread use across the field of climate research; however, as data grows and climate patterns become even more complex, researchers are ramping up their investments in some of the largest and most powerful systems available today. For example, the National Center for Atmospheric Research (NCAR) recently deployed a new supercomputer named Cheyenne in order to further their critical climate change efforts. The new system, which is roughly the size of a house, is currently one of the top 20 fastest supercomputers in the world, capable of executing 5.3 quadrillion calculations per second. This type of supercomputer is helping researchers better understand climate change in a number of key ways.
Improve model accuracy
Supercomputers allow scientists to vigorously test climate models to ensure accuracy before applying them to predict future trends. Climate models are first tested using a process called hindcasting, where the models are asked to simulate historical conditions and then those results are compared to actual observations. If the models are able to correctly simulate past conditions, then scientists can have confidence that their models will be able to accurately predict the future.
Gain a comprehensive view of past and present trends
Climate records that date back thousands of years can help researchers build models of past climate conditions in order to establish a trend line for how the climate is evolving. For example, a recent study used reconstructed temperature data dating back 2 million years to analyze the relationship between historical temperatures and carbon dioxide levels. This analysis helped researchers gain a better understanding of historical climate conditions, as well as revise their findings on the future effects of large climate changes like global warming.
Predict future climate change
Climate science combines historical observations and theoretical models to understand past climate trends and then use those trends to predict what might be likely to occur in the future. For example, climate models cannot predict if it will snow on a certain day like a weather forecast can, but they can reveal general trends like the average number of snowy days over a period of days, months, or years.
Refining and improving climate models is dependent on ever-growing volumes of data from both historical and current observations, which require massive compute and storage resources. The highly data-intensive nature of today’s climate research field is broadening the demand for optimize compute resources that are more flexible, scalable, fast, and reliable. These tools are helping researchers better understand and predict climate change, and ultimately allowing us to gain a better understanding of the world we live in.
Please follow me on Twitter at @Bill_Mannel to learn more about the supercomputers that are driving innovation in the field of climate science.