Nat. Geosci. https://doi.org/10.1038/s41561-019-0383-x (2019)
The Hadley circulation (HC) — defined by tropical ascent near the Equator and subtropical descent on either side — constrains tropical precipitation, subtropical dry zones and poleward atmospheric heat transport. Climate models predict that the HC will weaken in response to anthropogenic warming, estimating that this has already occurred in the past four decades. Reanalyses (assimilated data sets that serve as proxies for observations) instead show a strengthening HC since 1979. Elucidating the source of this disagreement in the historical past is necessary to accurately assess HC projections in models.
Rei Chemke and Lorenzo Polvani of Columbia University, USA, set out to understand this discrepancy. They use a multi-model archive and single-model initial condition ensemble to show that this disagreement is not a result of internal variability within the climate system, but rather an erroneous representation of latent heat of moist processes in reanalyses. While the origin of the flawed latent heating remains unclear, these findings suggest that reanalyses may be unreliable for analysing tropical circulation trends and associated moist processes. Furthermore, these results reinforce trust in model simulations of recent and long-term HC behaviour.
About this article
Cite this article
Langenbrunner, B. In models we trust. Nat. Clim. Chang. 9, 577 (2019). https://doi.org/10.1038/s41558-019-0550-5