Changes in the dependence between global precipitation and temperature from observations and model simulations

Abstract

Precipitation and temperature are physically related to each other with significant dependences that vary with seasons at regional and global scales. The dependence between precipitation and temperature (P–T dependence) plays a central role in characterizing the joint behaviour of the two variables. Along with extensive studies on the variation in the mean, variance, or extremes of precipitation and temperature under global warming, it is of particular interest to understand the climate‐change impacts on the covariability of precipitation and temperature. The aim of this study therefore is to assess climate change impacts on the P–T dependence over global land areas from observations, and the performance of the current‐generation Coupled Model Intercomparison Project (CMIP5) models in simulating these patterns. In general, the CMIP5 models correctly simulate the overall pattern of the P–T dependence (around 87–94% of global land areas). However, they capture both the sign and change of P–T dependence for relatively small regions (around 41–47% of global land areas). Results from this study may provide useful insights for the future development of climate models.

Projected increase in compound dry and hot events over global land areas

“…Although there are certain differences between the model results and observations (e.g., a smoother pattern of the modelled results because of the average of 19 models), generally similar spatial patterns in the distribution of LMF were shown by CMIP5 multimodel simulations. These results indicate that the likelihood of compound dry and hot events can generally be simulated well from the CMIP5 models, which is partly due to the relatively satisfactory performance in simulating precipitation‐temperature dependence (Wu et al ., 2013; Hao et al ., 2019).…”

“…We mainly focused on two seasons June–July–August (JJA) and December–January–February (DJF) (Coumou and Robinson, 2013; Hao et al ., 2019; Wang et al ., 2020). A serial correlation may exist in monthly precipitation and temperature data (Coumou et al ., 2013; Serinaldi, 2016), which can lead to an overestimation of occurrences.…”

“…The variation pattern of the frequency and spatial extent of compound dry and hot events was roughly similar to that of the 30th and 70th percentile thresholds (Figures S2‐S5). Differences may exist in results based on different observation datasets, which may result in uncertainties in validating the performance of model simulations (Trenberth et al ., 2014; Hao et al ., 2019). Thus, we also used another observation dataset of monthly precipitation and temperature from the University of Delaware (UD) to assess the potential uncertainty.…”

Half‐a‐Degree Matters for Reducing and Delaying Global Land Exposure to Combined Daytime‐Nighttime Hot Extremes

“…Another open issue is related to drivers underpinning the past and future transition of dominant type of hot extremes, that is, faster increases in frequency of combined events compared to daytime-/nighttime-only events. Physically, this question is promising to be interpreted in light of responsible dynamic and thermodynamic drivers' responses to anthropogenic global warming (Gibson et al, 2017;Hao et al, 2019;Miralles et al, 2014). Statistically, adopting bivariate-generalized extreme value distribution/generalized Pareto…”

“…Physically, this question is promising to be interpreted in light of responsible dynamic and thermodynamic drivers' responses to anthropogenic global warming (Gibson et al, 2017;Hao et al, 2019;Miralles et al, 2014). Physically, this question is promising to be interpreted in light of responsible dynamic and thermodynamic drivers' responses to anthropogenic global warming (Gibson et al, 2017;Hao et al, 2019;Miralles et al, 2014).…”