Ongoing and projected future intensification of North American pluvial and drought extremes relative to the pre-Industrial millennium

Assessing precipitation non-stationarity beyond the modern, instrumental record is valuable for disentangling the impacts of anthropogenic climate change from natural climate variability. This study evaluated changes in 3-month meteorological drought and pluvial extremes by merging tree-ring reconstructions, observations, and climate model simulations spanning 850 – 2100 CE across North America; to determine whether the Industrial era and projected future changes are outside the range of natural climate variability. To accomplish this, we utilized a non-stationary version of the commonly used Standardized Precipitation Index (SPI), modified to capture the slow progression of underlying statistical distributions through time. Within this non-stationary framework, multi-dimensional splines simultaneously model annually recurring seasonality and seasonally specific trends, constrained to mimic the WMO 30-year reference period, for each distribution parameter. The non-stationary SPI framework was further developed to merge tree-ring proxy data, 20th century observations, and CMIP6 climate model output into a common millennial-scale model by accounting for seasonal and data-specific biases.

Results show that many regions of North America have already experienced significant intensification of drought and pluvial extremes relative to the previous 1,000 years of presumed natural climate variability. These appear as widespread exacerbation of both extremes, especially summer drought and winter pluvials with consistent spatial signals: overall drying trends in the west and south, wetting trends in the northeast, and increased interannual variability across the east and north. Climate change projections indicate a continued intensification of these trends by 2100. These results underscore the need for reassessing severities of recent drought and pluvial events relative to a changing climatological baseline and a need for incorporating climate non-stationarity when assessing future drought and pluvial risk.