Pardeep Pall: Characterising distributions of extremes between pre- and post-peak climate overshoot periods in ACCESS-ESM1-5
Overshooting the Paris Agreement target of limiting global warming to 1.5°C now seems unavoidable. This increasingly raises questions regarding whether changes in the climate system reverse if we manage to return and stabilise to a 1.5°C-or-below climate, after overshooting and peaking in global warming. A common practice is to compare changes in a post-peak period that has stabilised at some level of global warming, against a pre-peak period possessing that same level of global warming. However, this latter period very often spans a transient, relatively rapid, warming phase uncharacteristic of a stable climate. Thus, one might expect the characteristics of statistical distributions of climate variables to differ between this transient pre-peak period and stable post-peak period — on top of any changes due to potential climate irreversibility. Furthermore, one might expect this behaviour to be exacerbated for climate extremes. To date, no overshoot studies have explicitly addressed this issue, in part due to a lack of simulations to adequately sample the distributions. Here, we do this using daily resolution data from a 40-member ensemble of simulations with the ACCESS-ESM1-5, run under the SSP5-3.4 overshoot scenario to 2300. This unusually large dataset affords a robust characterisation of changes, and consideration of a variety of climate extremes. It also allows better exploration of the influence of pre- and post-peak period length on our results. With a new round of more comprehensive overshoot simulations planned as part of CMIP7, our study may be important in informing how changes in those simulations should be interpreted.