The following was submitted by pre-eminent climate scientist Kevin Trenberth, who was employed by the National Center for Atmospheric Research from 1984 to 2019, to the National Science Foundation.
Comment on Dear Colleague Letter on the National Center for Atmospheric Research
This letter focusses on the second topic, “NCAR weather modeling and atmospheric observing capabilities.”
I am writing as an individual who is a U.S. citizen (and also a New Zealand citizen). I lived in the United States for over 45 years and was employed by NCAR until 2019. Earlier, for 7 years from 1977 to 1984, as a professor at the University of Illinois, I visited NCAR every summer and had my students using the NCAR computers and data archives remotely.
I believe my credentials are second to none as a climate scientist:
My publication record includes a total (as of January 2026) of 80 books or book chapters, and 308 journal articles, plus 26 non-technical articles on The Conversation and 22 on New Zealand’s Newsroom, for a total of 662 publications plus 4 videos. On Google Scholar, there are > 137,500 citations and an H index of 142 (or 89 since 2021). I have been heavily involved in the World Climate Research Programme.
I am appalled at the attacks on NCAR and attempts to break it up. That is not to say that its operations could not be improved, but much of the U.S. Administration’s rhetoric is misguided and scientifically quite wrong.
NCAR is a national and even international center for research on the atmosphere and Earth system. It plays a major role as a center for 125 or so universities in education of students and provides facilities for research. Many students visit over the summer and participate in specially designed programs and seminars, and many doctoral theses are written jointly with NCAR scientists as co-advisors. Altering NCAR would be a major setback for the entire community, and it would also ultimately negatively impact every person in the U.S. and on Earth, and the impact would be felt for decades to come. NCAR activities include maintaining a major community super-computing center along with enormous data archives; heavily instrumented airplanes and many other instruments for field programs and exploring the Sun, and maintaining and developing community weather and Earth system models that are used widely in the US and around the world. The latter range from specialized models, such as for FAA and aircraft and airport management, a hydrology model, a major weather model (Weather Research and Forecasting Model: WRF), and the Community Earth System Model. The CESM has evolved from being a climate model to much more. A key aspect of all of these is that the atmosphere is global and it interacts with the land and oceans, and solar radiation. The land includes all of the complexity of the topography and vegetation, as well as ice and snow cover. Accordingly, it is an exceedingly challenging problem and through NCAR the entire community can participate fully.
NCAR’s science is diverse, and involves all aspects of atmospheric sciences, including atmospheric chemistry, and also the upper atmosphere and the Sun. It increasingly goes further to embrace all Earth sciences, including the entire climate system (atmosphere, oceans, land and ice), and social science aspects on how to best utilize this information for societal benefit. Climate change is a tiny component of all this, but has to be included because all the Earth systems are changing, and now most of that change is driven by human activities. But even in the absence of climate change, climate study and understanding is essential for all of human endeavours, especially farming, forestry, air travel, shipping and transport.
It has been argued that NCAR should not do climate but should focus on improving weather forecasting. However, the main challenges and scope to improve weather forecasts on all time scales are by improving the way interactions of the atmosphere occur with the land hydrology, vegetation, ice, and ocean. Rainfall fundamentally involves the hydrological cycle. It includes evaporation from the surface and oceans, transport of water vapor onto land, precipitation, stream flow back to the ocean, and storage in ponds and lakes. These are inherently all climate processes.
As another critical example, in early days dealing with hurricanes was regarded as a problem for meteorologists. Now hurricanes are recognized to depend on heat and moisture from the oceans which fuel the storms and result in heavy rainfalls and potentially flooding on land. A new major review highlights the complexities and why hurricanes are inherently a climate problem.
Ma, Z., L. Cheng, S. Camargo, K.E Trenberth, I.I. Lin, G. R Foltz, D. R. Chavas, D. Zhang, E. A Ritchie, J. Fei, C. Pasquero, K. J. E. Walsh, Z. Tan, R. L. Sriver, H. Ye and L. Zhou, 2026: Interactions of tropical cyclones with global energy and water cycles. Nature Reviews Earth & Environment, doi:10.1038/s43017-026-00770-6, in press.
I am included among the mix of authors here, involving 16 different institutions, which highlights also the global nature of the scientific issues- and that leadership is moving towards China.
None of the statements from the administration nor from NSF make sense physically and would amount to a major step backwards and loss of progress in improving skill in forecasting on all time scales. The observations, data processing, assimilation, modeling, and simulations must all be strongly linked. Climate and the atmosphere are inherently global, and their simulation have necessarily always utilized the biggest and fastest super computers available.
NCAR has been a unique leader and major center in the global and coupled aspects of climate and Earth system science, as well as in training early career scientists and graduate students. No doubt some aspects of NCAR could be improved in some way, but the component links are there because they reflect the physical and biogeochemical links in the real world and the global nature of the atmosphere.
Kevin E Trenberth