Introduction
The term “weather whiplash” has appeared frequently in recent media
reports describing abrupt shifts from one type of weather extreme to
another. These shifts typically consist of a severe cold spell being
replaced by a period of above-normal temperatures, a prolonged drought
followed by intense precipitation, or the reverse sequence in either
case. A striking example of a whiplash event occurred in early September
2020 when a prolonged record-breaking heatwave over a large region of
the central Rocky Mountains in the U.S. abruptly ended with a
temperature drop exceeding 60oF in some areas along
with several inches of snow. In 2018 a six-week cold spell in eastern
North America flipped to a February heatwave that broke high temperature
records from northern Maine to New Orleans and sent Bostonians flocking
to the beaches
(https://www.wunderground.com/cat6/summer-february-80-massachusetts-78-nyc).
The disruptive “false spring” in March 2012 that struck the Midwest
and lasted for several weeks was followed by killing freezes in April
that wreaked havoc on fruit farmers, whose crops blossomed too early in
March and were then damaged by the anomalous cold spell
(https://medium.com/dose/whats-a-false-spring-b64cb977d59). Only a
handful of studies has investigated these types of disruptive weather
shifts, and as yet there is no consistent definition of a weather
whiplash event, as distinct from the passage of fronts associated with
progressive synoptic weather systems.
Metrics of variability are one approach used to assess whiplash events.
An investigation of atmospheric temperatures during recent (1988/89 –
2014/15) winters (DJF) in the northern hemisphere by Cohen (2016) found
that, consistent with previous studies (e.g., Screen 2014), daily winter
near-surface temperature variability decreased in high latitudes
(60oN-90oN), as would be expected
with a weaker poleward temperature gradient owing to amplified Arctic
warming. In contrast to most studies, Cohen (2016) also found that
variability in creased in low- to mid-latitudes
(0-50oN), which was interpreted as a possible
indication of increased weather whiplash. Larger variability in daily
temperatures, however, suggests a timescale associated with fast-moving
synoptic weather systems, such as frontal passages and changes in wind
direction, rather than a shift from one persistent regime to another.
Loecke et al. (2017) also focused on extreme precipitation events to
define a weather whiplash index associated with drought-to-flood
transitions. Focusing on the Upper Mississippi River basin, they
calculated the index using the total precipitation from January to June
of each year minus the total from July to December of the previous year
divided by the total over both periods. They applied the index to
identify events in projections by 30 of the models that participated in
the climate model intercomparison project, version 5 (CMIP5) forced with
representative concentration pathway (RCP) 8.5. They found that 19 of
the models exhibited a robust positive trend in their index while the
others had no significant trend.
The study by Swain et al. (2018) focused on regional precipitation
extremes in California’s rainy season (Nov.-Mar.), coining the term
“precipitation whiplash” as a transition from anomalously dry to
anomalously wet seasons in consecutive years. They analyzed a large
number of climate simulations (forced with RCP8.5 conditions) created
with the National Center for Atmospheric Research (NCAR) Community Earth
System Model, CESM1. A whiplash event was identified when one rainy
season with precipitation totals below the 20thpercentile during the pre-industrial period was followed by a season
with totals exceeding the 80th percentile. They found
a significant increase in events over most of southern California,
Mexico, and parts of northern California through the
21st century, which implies increasing challenges for
agencies that manage freshwater resources in that region.
He and Sheffield (2020) also focused on precipitation to develop a
metric of whiplash, defining a drought-pluvial seesaw similar to Swain
et al. (2018), in which dry spells were followed by wet spells
(pluvials). They analyzed data from the past nearly seven decades
(1950-2016) on a global scale for the spring/summer (April-September)
season and for fall/winter (October-March). Unlike Swain et al. (2018)
who focused on consecutive wet seasons, this study investigated
three-month lags in the drought-pluvial seesaw. They calculated the
ratio of the frequency of events during the past 30 years (1987-2016) to
that in the first 30-year period (1950-1979). Over North America, they
found increased ratios over more than half of the area during the warm
season along with a three-fold increase in frequency over about
one-fifth of the region during the cold season, mainly in the central
U.S. The authors note challenges, however, in interpreting the results
owing to differing consistency of regional metrics for droughts and
pluvials as well as disentangling the roles of natural variability and
climate change.
We build on this growing body of research into disruptive and abrupt
shifts in weather extremes, with a focus on a domain spanning North
America and the eastern North Pacific Ocean. We demonstrate a novel
approach to assess weather whiplash based on abrupt shifts in the
large-scale circulation regime following the persistent dominance of one
pattern. This method does not rely on measurements or simulations of
precipitation or temperature, thereby avoiding uncertainties introduced
by instrument error, local heterogeneity, and model physics associated
with precipitation processes. For this study, we adopt the following
definition of a weather-whiplash event (WWE): a long-lived (4 or more
consecutive days), continental-scale pattern in the upper-level
circulation that shifts abruptly (over 1-2 days) to a substantially
distinct pattern, bringing a stark end to persistent weather conditions
throughout the region. This definition eliminates confusion in the
possible misidentification of WWEs caused by sharp, localized weather
changes owing to synoptic features such as fronts, discreet disturbances
(e.g., squall lines, tropical storms), and shifts in low-level winds
from differing surface types (e.g., from onshore to offshore, downslope
to upslope, or forest to grasslands). We submit that a WWE should be
identified when one persistent and anomalous circulation pattern is
replaced abruptly by a very different one, as distinct from the passage
of fronts on synoptic time scales.