*Hydrological sensitivity

The intensification of the hydrological cycle, involving the accelerated movement of water between the atmosphere, land, and oceans, is one of the important consequences of global warming, which profoundly affects food production, energy security, and economic development. Hydrological sensitivity (η) is a widely used metric for assessing the increase in global-mean precipitation resulting from surface warming, but its estimation remains uncertain and challenging. Previous research has identified a three-fold uncertainty in η estimated from global climate models (GCMs, an important tool used to simulate and predict the climate system), and attributed the source of uncertainty to different treatments of the atmospheric processes responding to the warming, such as clouds and water vapour simulations. 

*The warming pattern effect

Global warming not only leads to an increase in global-mean temperature but also causes changes in surface warming patterns. The recent Intergovernmental Panel on Climate Change (IPCC) assessment report (AR6) acknowledged the importance of this warming pattern effect on studying global climate change in terms of the Earth’s energy budget (IPCC AR6 Chapter 7: The Earth’s Energy Budget, Climate Feedbacks, and Climate Sensitivity). On the other hand, according to the atmospheric energy budget, the latent heat released from precipitation has to be balanced by atmospheric radiative cooling and surface fluxes. Given the close relationship between atmospheric energy budget and Earth’s energy budget, it is reasonable to expect that η is also dependent on warming patterns. Surprisingly, the effects of warming patterns on global η were not discussed in previous studies or in the IPCC chapter dedicated to precipitation and the water cycle (AR6 Chapter 8: Water Cycle Changes). 

Fig. 1. Sea surface temperature warming patch experiments. a, The geographical location of warming patches centre in the CAM5 model. b, The estimated probability density function (PDF) of η derived from the patch experiments.

*Filling the gap: the warming pattern effects on η

To fill this knowledge gap, our study investigated the influence of sea surface temperature (SST) warming patterns on η, using a combination of model simulations and observations. By varying the SST warming patterns within a single model, we were able to reproduce (and much beyond) the range of η estimated from multiple GCMs (Fig. 1), which is quite interesting and surprising. More interestingly, we found that warming in strong tropical ascending regions could produce η larger than the rate predicted by the Clausius-Clapeyron relationship (7% K⁻¹). This means that the global precipitation increase rate can exceed the water vapor increase rate under these specific (and idealized) circumstances. 

Fig. 2.  Attribution of hydrological sensitivity uncertainties in CMIP5 models. Values of hydrological sensitivity (black dots), warming pattern term (red dots), and atmospheric model difference term (blue dots) for 24 CMIP5 models normalized by the multi-model mean. 

Furthermore, our study found that differences in SST warming patterns have a comparable influence to differences in treatments of atmospheric processes (the focus of previous studies) when determining η in current GCMs (Fig. 2). Additionally, the reconstructed global-mean precipitation with pattern effect included agrees much better with observed annual global-mean precipitation than the reconstructed ones that only use global-mean temperature (i.e. the conventional approach) (Fig. 3). 

Fig. 3. Timeseries of observed precipitation anomaly from GPCP (solid black line), reconstructed precipitation anomaly without pattern effect (black dashed line), and reconstructed with pattern effect (orange line) based on HadCRUT observed SST datasets.

In a nutshell, our study brings attention to the importance of accounting for warming patterns when assessing η and predicting the effects of global warming on global precipitation. We also would like to deliver the point that η can vary over time, differ between climate models, and be influenced by experimental setups due to the pattern effect, instead of remaining unchanged. By unravelling the non-linear relationship between warming patterns and η, we move closer to a comprehensive understanding of how global warming shapes our planet's water cycle. Ultimately, our research aims to inform policymakers, scientists, and stakeholders, helping to improve their ability to better predict future climate and make mitigation policies accordingly.