Abatzoglou, J.T., Dobrowski, S.Z. & Parks, S.A. Multivariate climate departures have outpaced univariate changes across global lands. Sci Rep 10, 3891 (2020). https://doi.org/10.1038/s41598-020-60270-5

"We calculate annual multivariate climate departures during 1958–2017 relative to a baseline 1958–1987 period that account for covariance among four variables important to Earth’s biota and associated systems: annual climatic water deficit, annual evapotranspiration, average minimum temperature of the coldest month, and average maximum temperature of the warmest month. Results show positive trends in multivariate climate departures that were nearly three times that of univariate climate departures across global lands. Annual multivariate climate departures exceeded two standard deviations over the past decade for approximately 30% of global lands. Positive trends in climate departures over the last six decades were found to be primarily the result of changes in mean climate conditions consistent with the modeled effects of anthropogenic climate change rather than changes in variability. These results highlight the increasing novelty of annual climatic conditions viewed through a multivariate lens and suggest that changes in multivariate climate departures have generally outpaced univariate departures in recent decades. The largest positive σd trends occurred in regions of historically low variance (e.g., equatorial regions), in regions such as southern Europe that have seen large changes in climate trends in the variables considered, and in boreal regions that saw joint increases in AET (annual evapotranspiration) and D (annual climatic water deficit) that are orthogonal to the historical covariance of these variables.”

See also data in an interactive way at Perry World House Global Climate Security Atlas https://global.upenn.edu/perryworldhouse/global-climate-security-atlas