A global environmental crisis 42,000 years ago

Abstract

Reversing the field
Do terrestrial geomagnetic field reversals have an effect on Earth's climate? Cooper et al. created a precisely dated radiocarbon record around the time of the Laschamps geomagnetic reversal about 41,000 years ago from the rings of New Zealand swamp kauri trees. This record reveals a substantial increase in the carbon-14 content of the atmosphere culminating during the period of weakening magnetic field strength preceding the polarity switch. The authors modeled the consequences of this event and concluded that the geomagnetic field minimum caused substantial changes in atmospheric ozone concentration that drove synchronous global climate and environmental shifts.
Science , this issue p. 811

Full text

PALEOENVIRONMENT
A global environmental crisis 42,000 years ago
Alan Cooper
1,2
*†, Chris S. M. Turney
3
*†, Jonathan Palmer
3
, Alan Hogg
4
, Matt McGlone
5
,
Janet Wilmshurst
5,6
, Andrew M. Lorrey
7
, Timothy J. Heaton
8
, James M. Russell
9
, Ken McCracken
10
,
Julien G. Anet
11
, Eugene Rozanov
12,13,14
, Marina Friedel
12
, Ivo Suter
15
, Thomas Peter
12
,
Raimund Muscheler
16
, Florian Adolphi
17
, Anthony Dosseto
18
, J. Tyler Faith
19
, Pavla Fenwick
20
,
Christopher J. Fogwill
21
, Konrad Hughen
22
, Mathew Lipson
23
, Jiabo Liu
24
, Norbert Nowaczyk
25
,
Eleanor Rainsley
21
, Christopher Bronk Ramsey
26
, Paolo Sebastianelli
27
, Yassine Souilmi
28
,
Janelle Stevenson
29,30
, Zoë Thomas
3
, Raymond Tobler
28
, Roland Zech
31
Geological archives record multiple reversals of Earth’s magnetic poles, but the global impacts of these
events, if any, remain unclear. Uncertain radiocarbon calibration has limited investigation of the potential
effects of the last major magnetic inversion, known as the Laschamps Excursion [41 to 42 thousand
years ago (ka)]. We use ancient New Zealand kauri trees (Agathis australis) to develop a detailed record
of atmospheric radiocarbon levels across the Laschamps Excursion. We precisely characterize the
geomagnetic reversal and perform global chemistry-climate modeling and detailed radiocarbon dating of
paleoenvironmental records to investigate impacts. We find that geomagnetic field minima ~42 ka, in
combination with Grand Solar Minima, caused substantial changes in atmospheric ozone concentration
and circulation, driving synchronous global climate shifts that caused major environmental changes,
extinction events, and transformations in the archaeological record.
Over the recent past, Earth’smagnetic
field has steadily weakened (~9% in
the past 170 years), and this, along
with the current rapid movement of
the magnetic North Pole, has increased
speculation that a field reversal may be im-
minent (1,2). The estimated economic impacts
of such a reversal have focused on the in-
creased exposure to extreme solar storms,
with multibillion-dollar daily loss estimates
(3) likely to be conservative. One of the best
opportunities to study the impacts of ex-
treme changes in Earth’s magnetic field is the
Laschamps Excursion (hereafter Laschamps)—
a recent, relatively short-duration (<1000 year)
reversal ~41 thousand years ago (ka) (4). Sedi-
mentary and volcanic deposits indicate a
weakening of the magnetic field intensity to
<28% of current levels during the reversed
phase of the Laschamps and, notably, as little
as 0 to 6% during the preceding transition as
polarity switched (Fig. 1 and supplementary
materials) (1,2,5).
Studies of Greenland ice cores have failed
to reveal marked impacts in high-latitude
paleoclimate associated with Laschamps (5,6),
and this observation has underpinned the
current view that there is no relationship be-
tween geomagnetic reversals and climate or
environmental changes. However, the markedly
increased levels of solar and cosmic radiation
reaching Earth’s atmosphere because of the
weakened geomagnetic field are likely to
have increased atmospheric ionization and
decreased stratospheric ozone levels, poten-
tially generating regional climatic impacts,
particularly in lower latitudes (7–9). In this
regard, it is notable that many environmental
records around the Pacific Basin appear to
detail a major (and often sustained) change
in behavior ~40 to 42 ka, including local gla-
cial maxima in Australasia and the Andes
(7,10), long-term shifts in atmospheric circu-
lation patterns (11,12), and continent-wide
aridification and megafaunal extinction in
Australia (4,13–16). The same periodin North
America saw the rapid, pronounced expan-
sion of the Laurentide Ice Sheet (LIS) from a
local minimum close to 42 ka (17–19). Many of
these records document a long-term phase
shift into a glacial state that persisted until
the transition into the Holocene (~11.6 ka), in
direct contrast to the Atlantic Basin records of
millennial-scale abrupt and extreme changes
associated with stadial-interstadial events.
Although the Pacific Basin environmental
changes appear broadly coincident with the
Laschamps, the lack of knowledge about the
exact timing and duration of the geomagnetic
excursion has greatly limited the ability to
examine whether it played any role. In addi-
tion, chronological uncertainties are complicated
in radiocarbon-dated terrestrial and marine
records around the Laschamps because of the
elevated production of
14
Cand
10
Be, cosmo-
genic radionuclides resulting from the substan-
tial increase in high-energy cosmic radiation
reaching the upper atmosphere. The high
10
Be
flux has been well described from Greenland
and Antarctic ice core records (6,20,21), which
reveal synchronous century-long
10
Be peaks
across the Laschamps that appear to reflect
a series of pronounced Grand Solar Minima
(GSM; prolonged periods of low solar activity
similar to the Spörer and Maunder Minima:
1410 to 1540 CE and 1645 to 1715 CE), with
unknown climate impacts (20,21). By con-
trast, the associated atmospheric
14
Cchanges
remain poorly constrained (22), preventing
precise calibration (23).
Radiocarbon changes across the Laschamps
In this study, we performed detailed radiocarbon
analyses of ancient kauri (Agathis australis)trees
preserved in northern New Zealand wetlands
(24) to generate a contiguous reconstruction
of atmospheric
14
C across the Laschamps (see
supplementary materials). We compared a
series of radiocarbon measurements across
multiple kauri trunk cross sections to identify
variations in atmospheric radiocarbon at a
highly resolved level. A 1700-year record from
a tree recovered from Ngāwhā, Northland, is
RESEARCH
Cooper et al., Science 371, 811–818 (2021) 19 February 2021 1of8
1
South Australian Museum, Adelaide, SA 5000, Australia.
2
BlueSky Genetics, PO Box 287, Adelaide, SA 5137, Australia.
3
Chronos
14
Carbon-Cycle Facility, and Earth and Sustainability Science
Research Centre, University of New South Wales, Sydney, NSW 2052, Australia.
4
Radiocarbon Dating Laboratory, University of Waikato, Hamilton 3240, New Zealand.
5
Landcare Research, PO
Box 69040, Lincoln, New Zealand.
6
School of Environment, University of Auckland, Private Bag 92019, Auckland 1142, New Zealand.
7
National Institute of Water and Atmospheric Research Ltd,
Auckland 1010, New Zealand.
8
School of Mathematics and Statistics, University of Sheffield, Sheffield S3 7RH, UK.
9
Department of Geological Sciences, Brown University, Providence, RI 02912,
USA.
10
University of New South Wales, Sydney, NSW 2052, Australia.
11
Zurich University of Applied Sciences, Centre for Aviation, 8401 Winterthur, Switzerland.
12
Institute for Atmospheric and
Climatic Science, ETH Zurich, 8006 Zurich, Switzerland.
13
Physikalisch-Meteorologisches Observatorium Davos and World Radiation Center, 7260 Davos, Switzerland.
14
Department of Physics of
Earth, Faculty of Physics, St. Petersburg State University, St. Petersburg 198504, Russia.
15
Swiss Federal Laboratories for Materials Science and Technology (Empa), 8600 Dübendorf,
Switzerland.
16
Department of Geology, Quaternary Sciences, Lund University, 22362 Lund, Sweden.
17
Alfred Wegener Institute, Helmholtz Centre for Polar and Marine Research, 27570
Bremerhaven, Germany.
18
Wollongong Isotope Geochronology Laboratory, School of Earth, Atmospheric and Life Sciences, University of Wollongong, Wollongong, NSW 2522, Australia.
19
Natural
History Museum of Utah and Department of Anthropology, University of Utah, Salt Lake City, UT 84108, USA.
20
Gondwana Tree-Ring Laboratory, PO Box 14, Little River, Canterbury 7546, New
Zealand.
21
School of Geography, Geology and the Environment, University of Keele, Keele, Staffordshire ST5 5BG, UK.
22
Department of Marine Chemistry and Geochemistry, Woods Hole
Oceanographic Institution, Woods Hole, MA 02543, USA.
23
Centre of Excellence for Climate System Science, University of New South Wales, Sydney, NSW 2052, Australia.
24
Southern University
of Science and Technology, Department of Ocean Science and Engineering, Shenzhen 518055, China.
25
Helmholtz Centre Potsdam, GFZ German Research Centre for Geosciences, Section 4.3,
14473 Potsdam, Germany.
26
Research Laboratory for Archaeology and the History of Art, School of Archaeology, University of Oxford, OX1 3TG, UK.
27
Faculty of Mathematics, Astronomy and
Physics (FAMAF), National University of Cordoba, X5000HUA, Argentina.
28
Australian Centre for Ancient DNA, University of Adelaide, Adelaide, SA 5000, Australia.
29
Archaeology and Natural
History, School of Culture History and Language, ANU College of Asia and the Pacific, Canberra, ACT 2601, Australia.
30
Australia ARC Centre of Excellence for Australian Biodiversity and
Heritage, Australian National University, ACT 2601, Australia.
31
Institute of Geography, Friedrich-Schiller-University Jena, 07743 Jena, Germany.
*These authors contributed equally to this work.
†Corresponding author. Email: alan.cooper@blueskygenetics.com (A.C.); c.turney@unsw.edu.au (C.S.M.T.)
on February 18, 2021 http://science.sciencemag.org/Downloaded from

particularly important because it spans the
period of greatest change in
14
C, including an
apparent weakening of the magnetic field
before the Laschamps. The growth of the
Ngāwhātree is relatively suppressed com-
pared with both modern kauri and other
late Pleistocene kauri, and there is a marked
decrease in tree-ring width that coincides
with the weakest phase of the geomagnetic
field (supplementary materials). We spliced
the kauri tree
14
C series into the radiocarbon
dataset reported from the
230
Th-dated Hulu
Cave speleothem (22) to provide an absolute
(calendar) time scale (Fig. 1). Our 40-year–
resolved reconstruction (Fig. 1) shows major
changes in atmospheric radiocarbon before
and during the Laschamps (23), closely match-
ing reconstructions of the virtual geomagnetic
pole [positions and geomagnetic intensity
(1,5)]. A comparison of the kauri-Hulu
14
Cwith
the paleomagnetic intensity data indicates that
Cooper et al., Science 371, 811–818 (2021) 19 February 2021 2of8
Fig. 1. Atmospheric radiocarbon
changes across the Laschamps
geomagnetic excursion and com-
pared with key environmental
datasets. (Aand B) Kauri
14
C ages
and D
14
C values before and through
the Laschamps (colored symbols)
compared with Hulu Cave radiocar-
bon values (open symbols) (22). The
arrow denotes the peak in D
14
C
coincident with a prominent GSM
(see below). A short
14
C plateau
42.20 to 42.04 ka occurs around
halfway through the steep rise in
14
C,
which is consistent with Cariaco
Basin (40) and demonstrates that
this is a robust feature in the record.
B.P., before present (1950 CE); Mag.,
magnetic. Error bars indicate 1s.
(C) Relative paleomagnetic intensity
curve aligned to the Greenland ice
core record reported from Black Sea
sediments (5). (D) Normalized
Greenland
10
Be flux (light blue line)
(20) compared with modeled
14
C
production rates from the kauri-Hulu
dataset (thick black line). The ampli-
fied peaks in
10
Be during the
weakened paleomagnetic field are
consistent with increased ionizing
radiation during GSM (arrows; see
supplementary materials). (E) North
Greenland Ice Core Project (NGRIP)
d
18
O record reported on the GICC05
(+265 years) time scale B.P.;
Greenland interstadial (warming)
events 11 to 9 are shown (41), along
with the weak Greenland stadial
(GS-10), which may represent a local
Greenland signal of abrupt cooling
interrupting a larger interstadial event
originally consisting of GI-10 and GI-9
(29,30). (F) Sediment total reflec-
tance (refl-L*) measurements from
Cariaco Basin (250-point running
mean) showing the absence of a clear
GS-10 signal (30). The reversed
geomagnetic polarity (light-gray col-
umn) and flanking transition phases
(dark gray) are indicated, with the
latter being the weakest periods
of Earth’s magnetic field and closely
coincident with GS-11 and GS-10.
RESEARCH |RESEARCH ARTICLE
on February 18, 2021 http://science.sciencemag.org/Downloaded from

the reversed phase of the geomagnetic field
(and associated partial recovery) defining the
Laschamps sensu stricto occurred at 41.56 to
41.05 ka (supplementary materials).
By modeling
14
C-production rates from our
kauri D
14
C record, it is possible to precisely
aligntotheicecoretimescalebyusing
10
Be
records (21). Across this period, we infer that
theGreenlandicecore2005(GICC05)time
scale is 265 years younger than the Hulu Cave
time scale (95.4% range: 160 to 310 years)
(Fig. 1 and fig. S15), which is considerably
more precise than previous comparisons (21).
Notably, the steep rise in D
14
Ccommencesat
42.35 ka, witha peak value of 782 per mil (‰)
occurring at 41.8 ka, 300 years before the
full Laschamps reversal. This is the highest
atmospheric
14
C concentration yet reported
of the pre-anthropogenic radiocarbon time
scale (22,23,25) (see supplementary ma-
terials). The peak D
14
C value reported here
occurs during the most weakened phase of
the geomagnetic field (5) and is associated
with a prominent GSM recorded by
10
Be flux
(20) (Fig. 1 and supplementary materials), when
the weakened solar interplanetary magnetic
field allowed enhanced input of galactic cosmic
rays (GCRs) into the upper atmosphere. This
kauri-Hulu record provides a precise radio-
car bon calibration curve for this period, per-
mitting a detailed recalibration of wider
environmental changes to test synchrony
between events while also enabling us to in-
vestigate the potential climate drivers during
the Laschamps.
Global chemistry-climate modeling
To explore the impacts of a greatly weakened
geomagnetic field on atmospheric ionization,
chemistry, and dynamics, we undertook a se-
ries of simulations using a global chemistry-
climate model, SOCOL-MPIOM (8)(see
supplementary materials). First, the global
conditions before the Laschamps were mod-
eled by using modern values of the geo-
magnetic dipole moment (M) and average
solar modulation potential (f) of 800 MV
(equivalent to the modern value). After a
398-year spin-up, three 72-year–long sim-
ulations (from which the last 60 years were
used for analysis) were branched off to study
the Laschamps and two additional solar
states likely to influence atmospheric ioniza-
tion: a reference run keeping M = 100% cur-
rent and f= 800 MV (experiment REF); the
Laschamps with weakened geomagnetic field
(M = 0% current, f= 800 MV; experiment
M0P800) (2);andaLaschampsweakenedgeo-
magnetic field plus GSM when the decreased
geomagnetic field and the reduced solar mod-
ulation potential greatly increase the GCR
ionization rate in Earth’s atmosphere (M = 0%
current, f= 0 MV; experiment M0P0).
Although our simulation for the weakened
magnetic field during the Laschamps (M0P800)
showed modest but significant changes in at-
mospheric chemistry and climate (see supple-
mentary materials), the scenario for Laschamps
plus GSM (M0P0) showed greatly amplified
impacts, most notablyduringtheboreal
winter and austral summer (December to
February) (Figs. 2 and 3 and figs. S18 to S30).
Cooper et al., Science 371, 811–818 (2021) 19 February 2021 3of8
Fig. 2. The impact ofa weakened geomagnetic field and GSM on global chemistry.
(Ato C) Simulated anomalies in boreal winter and austral summer (relative to
experiment REF; December to February) atmospheric chemistry [HOx, NOx, and O
3
in
(A), (B), and (C), respectively] for a weakened magnetic field during the Laschamps
plusGSM(M=0,f= 0); colored areas denote 10% significance. (D) Total ozone column
change [Dobson units (DU)]; hatched areas denote 10% significance level.
RESEARCH |RESEARCH ARTICLE
on February 18, 2021 http://science.sciencemag.org/Downloaded from

Our results yield a large increase in atmo-
spheric ionization from GCRs, resulting in
an enhanced production of hydrogen and
nitrogen oxides (HOx and NOx, respectively)
(Fig. 2, A and B) (8) down to very low altitudes.
The increased HOx and NOx concentrations
influenced ozone levels over the entire atmo-
sphere, decreasing the O
3
mixing ratio in the
stratosphere (~5%) while increasing the O
3
mixing ratio in the troposphere, with the great-
est changes observed over Antarctica (~5%)
(Fig. 2, C and D).
We find that decreasing stratospheric O
3
concentrations had climatic impacts over the
mid- to high latitudes in both hemispheres
(Fig. 3). In the Northern Hemisphere, this
changed the equator-to-pole temperature
gradient, weakening the Arctic polar vortex
and leading to a net warming effect in the
lower polar stratosphere. We postulate that
this positive temperature anomaly was fur-
ther amplified by an increase in the Brewer-
Dobson Circulation, which would have, among
other effects, led to adiabatic warming of air
masses sinking from upper to lower strato-
spheric heights. The reason for the increased
Brewer-Dobson Circulation may originate from
a wavier jet stream, reaching higher velocities
at the Northern Hemisphere latitudes with the
most pronounced orographic barriers and in-
creasing gravity-wave production, which sub-
sequently propagated vertically up through the
atmospheric column (26). The lower atmo-
sphere responded to those factors with sea-
level pressure increases over the Arctic and
North America and decreases over Western
Europe, with parallel changes in surface tem-
perature. These changes resemble a negative
phase of the Arctic Oscillation (AO) and North
Atlantic Oscillation (NAO), consistent with
reanalysis studies (27). In the Southern Hem-
isphere, decreasing stratospheric O
3
appears
to be associated with small changes in the
mid-latitude airflow (Fig. 3) and subtropical
precipitation patterns (see supplementary ma-
terial) (8). Because the significance is <10%, an
ensemble of longer model runs is required to
confirm this finding.
Although previous studies have suggested
that snowfall over Greenland is summer dom-
inated during glacial conditions (28), the
model predictions of pronounced boreal win-
ter Arctic surface cooling are potentially im-
portant in the context of the Greenland ice
core records spanning this period. It is notable
that the two weakest phases of geomagnetic
field strength during the Laschamps closely
coincide with the cold Greenland stadials 11
(GS-11) and GS-10 (Fig. 1). Furthermore, GS-10
and the following brief interstadial GI-9
have a number of atypical features that have
led to suggestions that they might represent
the interruption of a single long warm inter-
stadial (composed of GI-10 and GI-9, as seen
in other records such as Cariaco Basin) (Fig. 1)
by an abrupt cold phase, likely related to an
expansion in North Atlantic sea-ice extent,
which changed the climatic gradient between
the mid-latitudes and Greenland (29,30)(sup-
plementary materials). As a result, the climatic
impacts of the Laschamps may have been ob-
scured by the way they are represented in the
Greenland ice core records.
Pacific climate and environmental impacts
In the Northern Hemisphere, it remains difficult
to disentangle the similarly timed impacts of
Laschamps from Greenland stadial-interstadial
events, early glacial advances, and the expansion
of anatomically modern humans (AMHs) (4,31).
Therefore, to isolate the potential impacts of
Cooper et al., Science 371, 811–818 (2021) 19 February 2021 4of8
Fig. 3. The impact of a weakened geomagnetic field and GSM on global climate. (Ato D) Simulated anomalies in boreal winter and austral summer (relative to
experiment REF; December to February) wind speed at 10 m, sea-level pressure, surface temperature, and zonal wind [(A) to (D), respectively] for a weakened
magnetic field during the Laschamps plus GSM (M = 0, f= 0); hatched areas denote 10% significance level.
RESEARCH |RESEARCH ARTICLE
on February 18, 2021 http://science.sciencemag.org/Downloaded from

Laschamps from these confounding factors,
we used the kauri-Hulu radiocarbon calibra-
tion curve to examine a transect of stratigraph-
ically constrained sites in the Pacific (i.e.,
outside the Atlantic Ocean basin) from the
subantarctic to the tropics (Fig. 4).
We used high-resolution radiocarbon dating
to investigate Laschamps-aged sedimentary
sequences at sites that record the behavior
of both the Intertropical Convergence Zone
(ITCZ) (Lake Towuti, Sulawesi) and mid-
latitude Southern Hemisphere westerlies
(subantarctic Auckland Islands). On the
subantarctic Auckland Islands (50°S) (see
supplementary materials), which currently
sit under Southern Hemisphere mid-latitude
westerly airflow (12), a lignite sedimentary
horizon at Pillar Rock records a warm period
Cooper et al., Science 371, 811–818 (2021) 19 February 2021 5of8
Fig. 4. Climatic and environmental
changes across the 42 ka Adams
Event. (A) NGRIP d
18
O(41). HE4
refers to Heinrich Event 4 within GS-9.
(Band C) Neanderthal extinction in
Europe (40.9 to 40.5 ka) recalibrated with
the kauri-Hulu calibration curve (B)
and arrival of AMHs in Europe (C) and
development of Initial Upper Paleolithic
cultures (37). (D) Early cave art and cave
utilization across both Europe and
Southeast Asia (34–36). (E) Rapid
expansion of LIS after a minimum extent
at 42 ka calendar B.P. that commenced
immediately before or during the
Laschamps (17–19). (F) Kauri and Hulu
Cave D
14
C values; colored solid circles
denote individual trees (key in Fig. 1). Error
bars indicate 1s.(G) Migration south of
the Intertropical Convergence Zone (ITCZ)
as recorded by Lake Towuti d
13
C
leaf wax
reported in this study (11). (H) Onset
of burning in tropical Queensland,
Australia (15). (I) Extinction of megafauna
in Australia (and continental-wide
aridification) (13,14) and faunal turnover
in southern Africa (42). SH, Southern
Hemisphere. (J) Local glacial maxima in
the southwest Pacific (7,10) and the
central Andes (7,10). (K) Migration of
mid-latitude westerly airflow equatorward
from Auckland Islands (12) and possible
extinction of subantarctic plant species
(fig. S6). (L) West Antarctic (WAIS)
d
18
O(41). NGRIP and WAIS d
18
O values
reported on the GICC05 (+265 years) time
scale (41). AIM, Antarctic isotopic
maximum events. The Laschamps is
shown in three phases: the transition
phases (dark gray) on either side of the
reversed polarity (light gray). The
Mono Lake Excursion (34.6 ka) is also
shown (light gray column) (1). Asterisks
denote evidence of geomagnetic alteration
associated with the Laschamps,
demonstrating synchronous change
across many parts of the globe.
RESEARCH |RESEARCH ARTICLE
on February 18, 2021 http://science.sciencemag.org/Downloaded from

from 54 to 42 ka within the last glaciation.
Pollen records of Dracophyllum scrub grassland
on this exposed cliff top and long-distance
transport of lowland podocarp forest pollen
from the New Zealand mainland indicate
weaker westerly winds than now and mean
annual temperatures within 1° to 1.5°C of
those today (supplementary materials), inter-
preted to represent a period when the core
of the Southern Hemisphere westerlies lay
relatively poleward, delivering mid-latitude
air masses over this sector of the Southern
Ocean (12). A series of 12 contiguous
14
Cages
reveals the upper stratigraphic boundary, mark-
ing a return to periglacial conditions that
occurred at 42.23 ± 0.2 ka, coincident with
the weakening of the magnetic field during
the transition phase into the Laschamps
(Fig. 1). The periglacial conditions lasted
until the Holocene (12), suggesting perva-
sive and widespread cold conditions (asso-
ciated with a strengthening or northward
shift in the core westerly airflow) across
this sector of the Southern Ocean.
In the equatorial west Pacific, Lake Towuti
currently experiences a wet season from
December to May as the ITCZ migrates south-
ward (11). During the last glacial period, Lake
Towuti preserves a marked and sustained
shift in d
13
C
leaf wax
to more positive values
(interpreted as representing more arid con-
ditions), which persisted until the Pleistocene-
Holocene boundary (Fig. 4 and supplementary
materials) (11). A comprehensive series of 13
new radiocarbon dates and sediment magnetic
intensity minima suggest that the ITCZ shift
occurred at 42.35 ± 0.2 ka, again during the
geomagnetic transition phase into the Laschamps
(fig. S11), precisely aligning to the westerly
airflow shift recorded at Pillar Rock. The high
level of precision on the ages obtained for the
major climatic boundaries recorded in Lake
Towuti and the Auckland Islands is only
possible because of the contiguous series of
radiocarbon dates from each sequence, which
permit accurate alignment against the steep
rise in atmospheric radiocarbon values across
this period (Fig. 1).
The above changes are consistent with a
wealth of observations that indicate major
environmental changes around the Pacific
Cooper et al., Science 371, 811–818 (2021) 19 February 2021 6of8
Fig. 5. Summarized potential impacts of different states of Earth’s
magnetic field and solar events. (Ato D) Size of the magnetosphere,
auroral extent, atmospheric ionization and associated chemistry and climate
impacts for (A) today, (B) the Laschamps, (C) GSM during the Laschamps,
and (D) major SEP events during the Laschamps (see supplementary
materials). Upward arrows represent increase, and downward arrows represent
decrease. max., maximum; min., minimum; SCR, solar cosmic radiation;
GV, gigavolt; MV, megavolt.
RESEARCH |RESEARCH ARTICLE
on February 18, 2021 http://science.sciencemag.org/Downloaded from

Basin at the time of the geomagnetic tran-
sition into the Laschamps. For instance, a
northward movement of the Southern Hemi-
sphere westerlies has been proposed to ex-
plain the local peak glacial advance in the
arid southern-central Andes, sometime before
39 ka (Fig. 4) (7) and tentatively related to
fluxes in cosmic radiation and the Laschamps
(8). Maximum glacial advances are also ob-
served in New Zealand ~42 ka (10), consistent
with the climate modeling of enhanced
southwesterly airflow over the mid-latitudes
(Fig. 3) (8). These broad-scale atmospheric
circulation changes appear to have had far-
reaching consequences.WithinAustralia,the
peak megafaunal extinction phase is dated at
~42.1 ka, both in the mainland and Tasmania
(Fig. 4) (14–16), and has generally been at-
tributed to human action, although well after
their initial arrival at least 50 ka (14,16,32).
Instead, the megafaunal extinctions appear to
be contemporaneous with a pronounced cli-
matic phase shift to arid conditions that re-
sulted in the loss of the large interior lakes
and widespread change in vegetation patterns
(13,15). At Lynch’s Crater in northeast Australia,
the shift in vegetation structure, accompanied
by increased burning (15), has been recalibrated
here at 41.91 ± 0.4 ka, overlapping with the
climatic boundaries observed at Lake Towuti
and the Auckland Islands. Likewise, sedi-
ments at the Lake Mungo site associate the
timing of the loss of Australia’s interior lakes
and megafaunal extinction phase with a re-
ported geomagnetic excursion ~42 ka (locally
called the “Lake Mungo Excursion”)(supple-
mentary materials) (13). Similar signals of
marked floral and faunal change also appear
to exist on New Caledonia and as far afield as
South Africa (see supplementary materials).
Together, these records suggest that both a
mid-latitude climatic shift and major extinction
phases overlap with the geomagnetic transition
leading into the Laschamps, implying an asso-
ciation between these events.
Our model simulations potentially provide
important insights into the global nature of
thechangesobservedaroundthetimeofthe
Laschamps. Although the immediate impacts
associated with the geomagnetic transition were
likely on the order of the duration (800 years),
many of the above synchronous changes per-
sisted for millennia. This implies that a thresh-
old may have been passed in one or more Earth
system components, effectively tipping into a
different persistent state (Fig. 4). One possibility
is that with Earth’s orbital configuration moving
toward a full glacial state and limited global
ocean ventilation (see supplementary mate-
rials), the climate system may have been
sensitive to a relatively short but extreme
forcing around the time of the Laschamps.
For instance, terrestrial and marine sedimen-
tary records have revealed that the LIS ex-
panded rapidly from a local minimum at
42 ka (18,19) in association with a magnetic
reversal (17,19), with geological constraints
and numerical models indicating that some
parts of the ice sheet may have expanded
>1000 km by ~39 to 37 ka (18). Although our
model simulations do not suggest any major
change in airflow over the equatorial and
southern Pacific, we do find a substantially
weakened polar vortex, most notably during
the boreal winter (Fig. 3 and supplementary
materials). The greatly reduced surface tem-
peratures akin to a negative phase of the AO
and NAO could potentially have created a
positive feedback for ice sheet growth, reduc-
ing global sea levels. Recent work has sug-
gested that a greatly expanded LIS would have
reorganized atmospheric circulation across the
wider Pacific Ocean (11). Such a hemisphere-
wide response to abrupt forcing is consistent
with the synchronous movement of the mid-
latitude Southern Hemisphere westerlies deter-
mined from Pillar Rock and implied from glacial
behavior in New Zealand, Australia, and the
central Andes.
The Adams Transitional Geomagnetic Event
and wider implications
Overall, the signals discussed above suggest
that contemporaneous climatic and environ-
mental impacts occurred across the mid- to
lower latitudes ~42 ka, coincident with Earth’s
weakened geomagnetic field immediately pre-
ceding the reversed state of the Laschamps
(Fig. 4). We describe this as the “Adams
Transitional Geomagnetic Event”(hereafter
“Adams Event”), named after the science
writer Douglas Adams because of the timing
(the number “42”) and the associated range of
extinctions (33). Previous studies may have
failed to identify such a relationship between
the Laschamps and climatic impacts because
of the lack of temporal resolution and by
focusing on the period of actual reversed
geomagnetic field (41.5 to 41.1 ka) (5,6)rather
than the preceding extended phase of much
weaker geomagnetic field (42.4 to 41.5 ka).
The lowered geomagnetic field intensity
during the Adams Event, together with major
changes in the intensity of cosmic radiation,
is estimated to have increased levels of at-
mospheric ionization and ultraviolet (UV)
radiation, especially in equatorial and low
latitudes (<40°), because of a 10-fold decrease
in the cosmic ray cut-off rigidity (Fig. 5 and
supplementary materials). During GSM, the
ionization in the middle stratosphere and
surface UV radiation levels are estimated to
have been further heightened (up to 25 to
40% and 10 to 15% above current levels, re-
spectively), after taking into account changes
in the solar spectrum. Furthermore, these
values are likely to be much greater during
the short-lived solar energetic particle (SEP)
events (Fig. 5 and supplementary materials).
Although the relationship between increased
atmospheric ionization and stratospheric and
tropospheric cloudiness through cloud-seeding–
type impacts remains uncertain, such impacts
would be focused toward the lower latitudes
(9), where the potential for lightning strikes
could explain the increased records of charcoal
observed around the Laschamps in Australia
(15) and lack of relationship with archaeological
signs of human activity (see supplementary
materials).
The implications of this study are consider-
able. For instance, the Adams Event is very
close in timing to the globally widespread
appearance and increase in figurative cave
art, red ochre handprints, and changing use of
caves ~40 to 42 ka, e.g., in Europe and Island
Southeast Asia (fig. S34 and supplementary
materials) (34–36). This sudden behavioral
shift in very different parts of the world is
consistent with an increasing or changed use
of caves during the Adams Event, potentially
as shelter from the increase of ultraviolet B,
potentially to harmful levels, during GSM or
SEPs, which might also explain an increased
useofredochresunscreen(4). Rather than the
actual advent of figurative art, early cave art
would therefore appear to represent the pre-
servation of preexisting behaviors on a new
medium (supplementary materials) (36). Other
important archaeological boundaries during
the wider Laschamps include the extinction
of the Neanderthals (recalibrated here at 40.9
to 40.5 ka), along with the disappearance of
some of the first European AMH cultures and
the subsequent widespread appearance of
the Aurignacian technocomplex (Fig. 4 and
supplementary materials) (4,31,37).
The Adams Event appears to represent a
major climatic, environmental, and archaeo-
logical boundary that has previously gone
largely unrecognized. Furthermore, another
well-known geomagnetic excursion in the
recent past, Mono Lake (34 ka) (1), also
appears to be marked by a distinct peak in
the D
14
C levels in the Hulu Cave stalagmite
(Fig. 3) and aligns closely with a further
latitudinal shift in the ITCZ as recorded in
Lake Towuti, as well as a cluster of mega-
faunal extinctions in Eurasia (38)(supple-
mentary materials). Importantly, geomagnetic
transition phases can last substantially longer
than during the Laschamps; for instance, the
most recent full geomagnetic reversal, the
Brunhes-Matuyama (at ~790 ka) (39) has a
transition phase of ~20 ka, some 25 times
longer than the Adams Event, with potentially
far-reaching global climatic and evolutionary
effects. The discovery that geomagnetic excur-
sions can alter latitudinal temperature gra-
dients through drastic increases in cosmic
radiation and decreased ozone concentrations
provides a new model for sudden paleoclimate
Cooper et al., Science 371, 811–818 (2021) 19 February 2021 7of8
RESEARCH |RESEARCH ARTICLE
on February 18, 2021 http://science.sciencemag.org/Downloaded from

shifts. Overall, these findings raise important
questions about the evolutionary impacts of
geomagnetic reversals and excursions through-
out the deeper geological record (4).
REFERENCES AND NOTES
1. C. Laj, H. Guillou, C. Kissel, Earth Planet. Sci. Lett. 387,
184–197 (2014).
2. M. Brown, M. Korte, R. Holme, I. Wardinski, S. Gunnarson,
Proc. Natl. Acad. Sci. U.S.A. 115, 5111–5116 (2018).
3. E. J. Oughton, A. Skelton, R. B. Horne, A. W. P. Thomson,
C. T. Gaunt, Space Weather 15,65–83 (2017).
4. J. E. T. Channell, L. Vigliotti, Rev. Geophys. 57,709–738
(2019).
5. J. Liu, N. R. Nowaczyk, S. Panovska, M. Korte, H. W. Arz,
J. Geophys. Res. 125, e2019JB019225 (2020).
6. G. Wagner, D. M. Livingstone, J. Masarik, R. Muscheler, J. Beer,
J. Geophys. Res. 106, 3381–3387 (2001).
7. R. Zech, J. Zech, C. Kull, P. W. Kubik, H. Veit, Clim. Past 7,
41–46 (2011).
8. I. Suter, R. Zech, J. Anet, T. Peter, Clim. Past 10, 1183–1194
(2014).
9. J. Kirkby et al., Nature 533, 521–526 (2016).
10. P. D. Strand et al., Quat. Sci. Rev. 220, 165–177 (2019).
11. J. M. Russell et al., Proc. Natl. Acad. Sci. U.S.A. 111, 5100–5105
(2014).
12. E. Rainsley et al., Clim. Past 15, 423–448 (2019).
13. J. M. Bowler et al., Nature 421, 837–840 (2003).
14. F. Saltré et al., Nat. Commun. 7, 10511 (2016).
15. S. Rule et al., Science 335,1483–1486 (2012).
16. C. S. M. Turney et al., Proc. Natl. Acad. Sci. U.S.A. 105,
12150–12153 (2008).
17. H. W. Hill, B. P. Flower, T. M. Quinn, D. J. Hollander,
T. P. Guilderson, Paleoceanography 21, PA1006 (2006).
18. A. S. Dalton et al., Geology 47, 111–114 (2019).
19. A. S. Dalton, S. A. Finkelstein, P. J. Barnett, S. L. Forman,
Quat. Sci. Rev. 146, 288–299 (2016).
20. G. M. Raisbeck et al., Clim. Past 13, 217–229 (2017).
21. F. Adolphi et al., Clim. Past 14, 1755–1781 (2018).
22. H. Cheng et al., Science 362, 1293–1297 (2018).
23. R. A. Staff et al., Earth Planet. Sci. Lett. 520,1–9 (2019).
24. A. M. Lorrey et al., Quat. Sci. Rev. 183, 124–139 (2018).
25. C. S. M. Turney et al., Sci. Rep. 8, 3293 (2018).
26. K. Kodera, Y. Kuroda, J. Geophys. Res. 107, ACL 5-1–ACL 5-12
(2002).
27. J. Kidston et al., Nat. Geosci. 8, 433–440 (2015).
28. V. Masson-Delmotte et al., Science 309, 118–121 (2005).
29. S. O. Rasmussen et al., Quat. Sci. Rev. 106,14–28 (2014).
30. G. Deplazes et al., Nat. Geosci. 6, 213–217 (2013).
31. M. Staubwasser et al., Proc. Natl. Acad. Sci. U.S.A. 115,
9116–9121 (2018).
32. J. F. O’Connell et al., Proc. Natl. Acad. Sci. U.S.A. 115,
8482–8490 (2018).
33. D. Adams, M. Carwardine, Last Chance to See (Arrow Books,
2009).
34. A. W. Pike et al., Science 336, 1409–1413 (2012).
35. S. O’Connor, R. Ono, C. Clarkson, Science 334, 1117–1121
(2011).
36. M. Aubert et al., Nature 564, 254–257 (2018).
37. J.-J. Hublin et al., Nature 581, 299–302 (2020).
38. A. Cooper et al., Science 349, 602–606 (2015).
39. B. S. Singer, B. R. Jicha, N. Mochizuki, R. S. Coe, Sci. Adv. 5,
eaaw4621 (2019).
40. K. Hughen et al., Science 303, 202–207 (2004).
41. WAIS Divide Project Members, Nature 520, 661–665 (2015).
42. J. T. Faith, J. Hum. Evol. 65, 715–730 (2013).
ACKNOW LEDGM ENTS
We thank Marble Hill Estate, Adelaide Hills, for hosting a research
meeting in 2016 that helped initiate this project. We are grateful
for the assistance of N. Parker at Nelson’s Kaihu Kauri, Top Energy,
and Ngāpuhi iwi for permission to obtain and analyze samples of
the Ngāwhātree. We would also like to thank B. Finlayson,
R. Finlayson, K. Finlayson, and G. Beckham for providing access
to the subfossil kauri. M. Guzowski kindly helped prepare the
map used in Fig. 5. Many colleagues have provided valuable
support and suggestions; L. Van Gelder, H. Cadd, S. Haberle,
S. Haaland, T. Cooper, N. Golledge, and S. Fry. M. Thuerkow kindly
provided the initial land surface dataset and ocean temperatures
for the model simulations, and G. Chiodo provided support for the
chemistry-climate modeling. We acknowledge ETH Zürich for use
of the Euler Linux cluster and other computing facilities. We thank
three anonymous reviewers whose comments improved the
manuscript considerably. This work is dedicated to Douglas Adams
and Roger Cooper. Funding: We thank the Australian Research
Council for Discovery Grant and Laureate Fellowship support
[DP170104665 and FL140100260 (A.C.) and FL100100195
(C.S.M.T.)], the University of Adelaide Environment Institute, the
Australasian Antarctic Expedition 2013–2014 and The Tiama,
the Royal Society of New Zealand Marsden Fund (contract MFP-
NIW1803), NIWA, the University of Waikato, the Leverhulme Trust
(RF-2019-140\9), the Russian Science Foundation (20-67-46016),
and a Swiss National Science Foundation Ambizione grant
(PZ00P2_180043). Author contributions: C.S.M.T. and A.C.
designed the research; C.S.M.T., A.C., J.P., A.H., M.M., J.W., A.M.L.,
T.J.H., J.M.R., J.G.A., I.S., and P.F. performed the research; A.C.,
C.S.M.T., J.P., A.H., J.G.A., T.P., E.Ro., K.M., M.M., J.W., A.M.L.,
T.J.H., J.M.R., R.M., F.A., A.D., and I.S. analyzed the data; and A.C.
and C.S.M.T. wrote the paper with input from all authors.
Competing interests: The authors declare no competing interests.
Data and materials availability: All data are available in the main
text or the supplementary materials.
SUPPLEMENTARY MATERIALS
science.sciencemag.org/content/371/6531/811/suppl/DC1
Materials and Methods
Figs. S1 to S34
Tables S1 to S8
References (43–175)
24 March 2020; accepted 14 December 2020
10.1126/science.abb8677
Cooper et al., Science 371, 811–818 (2021) 19 February 2021 8of8
RESEARCH |RESEARCH ARTICLE
on February 18, 2021 http://science.sciencemag.org/Downloaded from

A global environmental crisis 42,000 years ago
Souilmi, Janelle Stevenson, Zoë Thomas, Raymond Tobler and Roland Zech
Mathew Lipson, Jiabo Liu, Norbert Nowaczyk, Eleanor Rainsley, Christopher Bronk Ramsey, Paolo Sebastianelli, Yassine
Raimund Muscheler, Florian Adolphi, Anthony Dosseto, J. Tyler Faith, Pavla Fenwick, Christopher J. Fogwill, Konrad Hughen,
J. Heaton, James M. Russell, Ken McCracken, Julien G. Anet, Eugene Rozanov, Marina Friedel, Ivo Suter, Thomas Peter,
Alan Cooper, Chris S. M. Turney, Jonathan Palmer, Alan Hogg, Matt McGlone, Janet Wilmshurst, Andrew M. Lorrey, Timothy
DOI: 10.1126/science.abb8677
(6531), 811-818.371Science
, this issue p. 811Science
changes in atmospheric ozone concentration that drove synchronous global climate and environmental shifts.
authors modeled the consequences of this event and concluded that the geomagnetic field minimum caused substantial
atmosphere culminating during the period of weakening magnetic field strength preceding the polarity switch. The
of New Zealand swamp kauri trees. This record reveals a substantial increase in the carbon-14 content of the ringsdated radiocarbon record around the time of the Laschamps geomagnetic reversal about 41,000 years ago from the
created a preciselyet al.Do terrestrial geomagnetic field reversals have an effect on Earth's climate? Cooper
Reversing the field
ARTICLE TOOLS http://science.sciencemag.org/content/371/6531/811
MATERIALS
SUPPLEMENTARY http://science.sciencemag.org/content/suppl/2021/02/17/371.6531.811.DC1
CONTENT
RELATED http://science.sciencemag.org/content/sci/371/6531/766.full
REFERENCES http://science.sciencemag.org/content/371/6531/811#BIBL
This article cites 163 articles, 25 of which you can access for free
PERMISSIONS http://www.sciencemag.org/help/reprints-and-permissions
Terms of ServiceUse of this article is subject to the
is a registered trademark of AAAS.ScienceScience, 1200 New York Avenue NW, Washington, DC 20005. The title
(print ISSN 0036-8075; online ISSN 1095-9203) is published by the American Association for the Advancement ofScience
Science. No claim to original U.S. Government Works
Copyright © 2020 The Authors, some rights reserved; exclusive licensee American Association for the Advancement of
on February 18, 2021 http://science.sciencemag.org/Downloaded from