Probably one of the most important ideas to understand about climate impact modelling: small changes to the mean equate to huge changes on the tail of the distr

Probably one of the most important ideas to understand about climate impact modelling: small changes to the mean equate to huge changes on the tail of the distribution. And that distribution over time has direct causal links to the health and viability of all biological lifeforms. We can talk about a 1.5 C degree rise in global temperature. Using Kelvin (which has an absolute zero), this represents a change from ~290K to 291.5K - a 0.5% change. Big deal, right? Not even a percent difference. But life doesn't seem to be concerned with "average" temperatures. If we look at other metrics like the frequency of days with wet-bulb temperatures above a certain threshold, or the number of contiguous frost-free days, or the bioclimatic envelope of species - metrics that incorporate time as well as biological thresholds that are causally related to species health and spatial distribution - we see that the impacts of climate change are far more pronounced. Large and measurable multi-percentage point changes already. The response of biologically-relevant variables is quite obviously non-linear compared to the temperature rise. In some cases, the frequency of these "extreme" events now occurs so regularly that they are no longer considered extreme. The baseline has shifted so rapidly, in my lifetime, that the conditions of a stable Holocene climate might as well be an ancient mythology.