Method-uncertainty is essential for reliable confidence statements of precipitation projections

Peter F Uhe, Daniel M. Mitchell, Paul D Bates, Myles Allen, Richard A. Betts, Chris Huntingford, Andrew King, Benjamin Sanderson, Hideo Shiogama

Research output: Contribution to journalArticle (Academic Journal)


Precipitation events cause disruption around the world and will be altered by climate change. However, different climate modeling approaches can result in different future precipitation projections. The corresponding ‘method uncertainty’ is rarely explicitly calculated in climate impact studies and major reports, but can substantially change estimated precipitation trends. A comparison across five commonly-used modeling activities shows that there is low consensus between methods over half the land-surface, for significant changes in mean precipitation between the present climate and 2◦C global warming. We also find that none of the modeling activities capture the full range of estimates from the other methods. Our results serve as an uncertainty map to help interpret which regions require a multi-method approach. Our analysis highlights the risk of over-reliance on any single modeling activity and the need for confidence statements in major synthesis reports to reflect this ‘method uncertainty’. Considering multiple sources of climate projections should reduce the risks of policymakers being unprepared for impacts of warmer climates compared to using single-method projections to make decisions.
Original languageEnglish
JournalJournal of Climate
Publication statusSubmitted - 22 Apr 2020

Fingerprint Dive into the research topics of 'Method-uncertainty is essential for reliable confidence statements of precipitation projections'. Together they form a unique fingerprint.

  • Cite this

    Uhe, P. F., Mitchell, D. M., Bates, P. D., Allen, M., Betts, R. A., Huntingford, C., King, A., Sanderson, B., & Shiogama, H. (2020). Method-uncertainty is essential for reliable confidence statements of precipitation projections. Manuscript submitted for publication.