top of page

Why climate models, though imperfect, are essential for policy makers

Ron Stouffer

Adjunct Professor of University of Arizona

In 1976, he received Pennsylvania State University BS Meteorology. In 1977, he received Pennsylvania State University MS Meteorology.

From 1977 to 2016, he worked at the Geophysical Fluid Dynamics Laboratory. From 2009 to 2012, he served as the group head of Climate and Ecosystems Group at GFDL. From 2012 to 2016, he served as the Senior Scientist for Earth System Modeling and Science at GFDL. From 2016 to 2019, he worked at the ClimBiz (a private climate consulting company). From 2016 to present, he served as the adjunct professor at University of Arizona.

His research interests focus on the Climate Change, Global warming and Climatology. His most recent publication is Climate Change 2021—The Physical Science Basis and Climate model projections from the Scenario Model Intercomparison Project (ScenarioMIP) of CMIP6.

Why climate models, though imperfect, are essential for policy makers

Climate models are the best tool climate scientists employ to understand past and present observed climate and predict future climate changes. Climate models are very similar in construction to weather forecasting models and are based on physical principles. These are then coded and run on a computer. Due to the long-time integrations needed for climate, the largest, fastest computers are used to study climate. For this reason, most climate centers are directly funded by their local federal government.

To answer both questions posed above, one must define the scale of the region in view. Is it the whole world or just your backyard? One also needs to define the time scale in view. Is it for a specific date in the future or some long term average (say 20 years or more). Many studies have shown that climate models can provide reliable information to inform decisions. The longer the time scale in view and the larger the area in view, the confidence in the model results increase. Also in the near term (next 10 to 20 years), changes in the greenhouse gas concentrations are not a big determinate to those changes. The built-in forcing changes and the state of the atmosphere-ocean system are the biggest players. As one goes further out in time, the greenhouse gas emissions (present day and future) play a larger and larger role in the predicted changes.

Another factor to consider in developing an answer to the questions above, is what social-economic future is in view. The longer the time into the future, the more important this factor (i.e. future emissions) becomes. Finally, climate models have a range of climate responses for the same input future emission scenarios. To account for this uncertainty, it is highly recommended to use a multi-model average to develop any future climate scenario. The use of any one model integration as “the right answer” will lead to many problems. Because of natural variability and model response uncertainty, the IPCC Reports contain plots with a range for any future climate change estimate.

Studies of the predicted climate changes made in the 1980’s and early 1990’s have been shown to be reliable at large scales (continental and larger) when compared to the observations of present day climate. For example, those early results indicated that the northern North Atlantic Ocean would not warm very much in the future. Observations now show a region of little warming in that region. They also showed that land areas would warm more than the adjacent ocean areas. Those early results also indicated that there would be more record high temperatures and fewer record low temperatures which has also been observed. The models predicted that summers would get longer leading to more droughts. Surprisingly, models also showed that the warmer atmosphere could hold more water vapor leading to more intense rainfall, resulting in more floods when it does rain in specific regions.

In summary, climate models are useful but not perfect. Climate scientists try to communicate what types of questions can be addressed by using present day models and what questions can not yet be answered. I encourage anyone to find and read the latest IPCC Working Group 1 Summary for Policy Makers. More details are found in the underlying reports which are referenced in the Summary. As is the case in any complex science, it is useful to consult with climate scientists (i.e. experts) before engaging in any large policy change.


Contacter: Zhifen Luo

Interviwer: Yaoyuan Fang

Proofreader: Siting Hou

Editor: Siting Hou

bottom of page