Precipitation whiplash variability and drivers - case studies from CONUS hotspots

Precipitation variability affects how much, when, and where water is available, and water is one of our most valuable natural resources. Historically, studies on precipitation have tended to focus on either excess (pluvial), or lack (drought), but in more recent work, there is recognition of the compounding hazards that both extremes can pose when they occur in rapid succession, including infrastructure damage, erosion and wildfire activity. Temporally sequenced flips between unusually dry and pluvial states are termed ‘whiplash’ events. In this study, we use two definitions of whiplash (SPEI thresholds, precipitation thresholds) over three timescales (1-month, 3-month, annual) to build a geospatial climatology over CONUS from 1895-2022, contrasting the results of each. Subsequently, using an aggregate metric for whiplash that is composed of each definition, we identify regional hotspots of whiplash activity, ultimately selecting a geographically-diverse subset to examine further. For these locations, we evaluate temporal trends, using time-series analysis, regression, and decomposition to evaluate longer-term drivers of variability. Subsequently, we use random forest regression to identify the degree to which known more of natural variability and/or seasonal to inter-annual dynamics influence the frequency of whiplash occurrence. Preliminary results suggest that a) high-frequency dynamical variations in the storm track (Geopotential heigh anomalies) play a dominant role, but that some Pacific and Atlantic modes of variability (e.g., Pacific-North-American Pattern, ENSO, Atlantic Multi-decadal Oscillation) also influence whiplash, b) whiplash is generally more frequent during periods of drought c) whiplash predictors are in many cases distinct from key predictors for pluvial or drought events independently. Establishing whether there are predicable precursor environments for whiplash, particularly at seasonal timescales, could be helpful to numerous sectors and thus this work aims to further that line of inquiry.