nature.geoscience

8/10/2019 Nature.geoscience

1/80


2/80


3/80

COVER IMAGESeismic anisotropy data for

the Great Basin region of the

western United States, coupledwith tomographic images, help

delineate a northeast-dipping

lithospheric drip. Numerical

experiments suggest that the

drip could have formed owing to

gravitational instability triggered

by a density increase of as little

as 1% and a temperature increase

of about 10%. The image shows

Wheeler Peak in Great Basin

National Park, Nevada, USA.

Photo by J. D. West, Silverheels

Photography (www.fodoze.com).

Article p439; News & Views p381;

Backstory p446

EDITORIAL

371 Complex communication

CORRESPONDENCE

372 Unexpected rise in extreme precipitation caused by a shift in rain type?

COMMENTARY

374 Securing the legacy of the IPY

Bob Dickson

BOOKS & ARTS

377 Encounters at the End of the World by Werner Herzog

Reviewed by Anna Armstrong

RESEARCH HIGHLIGHTS

378 Our choice from the recent literature

NEWS & VIEWS

379 Geophysics: Tectonics in the Earths core

Peter Olson

380 Palaeoclimate: Delayed Holocene warming

Martin Widmann

381 Tectonics: Draining Nevada

Vera Schulte-Pelkum

383 Environmental science: Rising arsenic risk?

David Polya and Laurent Charlet

384 Palaeontology: Aging well

Alicia Newton

385 Atmospheric science: Biological ice formation

Corinna Hoose

386 Palaeontology: Extinction before the snowball

Frank A. Corsetti

387 Atmospheric pollution: Brief relief Anna Armstrong

REVIEW ARTICLE

389 Volcanism in the Solar System Lionel Wilson

LETTERS

398 In situdetection of biological particles in cloud ice-crystals

Kerri A. Pratt, Paul J. DeMott, Jeffrey R. French, Zhien Wang, Douglas L. Westphal,

Andrew J. Heymsfield, Cynthia H. Twohy, Anthony J. Prenni and Kimberly A. Prather

N&V p385, online Backstory

ON THE COVER

Biotic turnover

Before global glaciation

Letter p415; News & Views p386;

online Backstory

Biogenic core

Cloud ice-crystals

Letter p398; News & Views p385;

online Backstory

Nordic overflows

Driven from the Atlantic

Letter p406

NATURE GEOSCIENCE| VOL 2 | JUNE 2009 | www.nature.com/naturegeoscience

JUNE 2009 VOL 2 ISSUE 6

Nature Geoscienceis printed on paperrecycled from post-consumer waste.

2009 Macmillan Publishers Limited. All rights reserved


4/80

402 Relative roles of biogenic emissions and Saharan dust as ice nuclei in the

Amazon basin

Anthony J. Prenni, Markus D. Petters, Sonia M. Kreidenweis, Colette L. Heald,

Scot T. Martin, Paulo Artaxo, Rebecca M. Garland, Adam G. Wollny and

Ulrich Pschl

N&V p385

406 Observed sources and variability of Nordic seas overflow

Tor Eldevik, Jan Even . Nilsen, Doroteaciro Iovino, K. Anders Olsson,

Anne Britt Sand and Helge Drange

411 The spatial and temporal complexity of the Holocene thermal maximum

H. Renssen, H. Sepp, O. Heiri, D. M. Roche, H. Goosse and T. Fichefet

N&V p380

415 Biotic turnover driven by eutrophication before the Sturtian

low-latitude glaciation

Robin M. Nagy, Susannah M. Porter, Carol M. Dehler and Yanan Shen

N&V p386, online Backstory

419 Tectonic history of the Earths inner core preserved in its seismic structure

Renaud Deguen and Philippe Cardin

N&V p379

423 Structural reactivation in plate tectonics controlled by olivine

crystal anisotropy

Andra Tommasi, Mickael Knoll, Alain Vauchez, Javier W. Signorelli,

Catherine Thoraval and Roland Log

ARTICLES

428 Oceanic forcing of the Marine Isotope Stage 11 interglacial

Alexander J. Dickson, Christopher J. Beer, Ciara Dempsey, Mark A. Maslin,

James A. Bendle, Erin L. McClymont and Richard D. Pancost

434 Mid-Pliocene climate change amplified by a switch in Indonesian

subsurface throughflow

Cyrus Karas, Dirk Nrnberg, Anil K. Gupta, Ralf Tiedemann, Kuppusamy Mohanand Torsten Bickert

439 Vertical mantle flow associated with a lithospheric drip beneath the

Great Basin

John D. West, Matthew J. Fouch, Jeffrey B. Roth and Linda T. Elkins-Tanton

N&V p381, Backstory p446

444 Erratum

BACKSTORY

446 A hidden drip

John D. West

CLASSIFIEDS

See the back pages

The period of relatively

warm climate from 11,000 to

5,000 years ago was marked

by considerable temporal

and spatial variability. Model

simulations relate this

complexity to the influence

of the waning Laurentide

ice sheet.

Letter p411;

News & Views p380

The myriad bodies that occur

in the Solar System show

a wide range of physical

properties. Exploration by

spacecraft during the past

four decades has shown

that volcanism a major

mechanism by which internal

heat is transported to the

surface is common on many

of these bodies. Image credit:

J. D. Griggs/USGS.

Review Article p389

NATURE GEOSCIENCE| VOL 2 | JUNE 2009 | www.nature.com/naturegeoscience

Nature Geoscience(ISSN 1752-0894) is published monthly by Nature Publishing Group (Porters South, 4 Crinan Street, London N1 9XW, UK). Editorial Office: Porters South, 4 Crinan Street, London N1 9XW, UK. Telephone:

+44 (0)20 7833 4000. Fax: +44 (0)20 7843 4563. Email: [email protected]. North American Advertising: Nature Geoscience, 75 Varick Street, Fl 9, New York, NY, 10013-1917, US. Telephone: +1(212) 726-9200.

Fax: +1(212) 696-9006. European Advertising: Nature Geoscience, Porters South, 4 Crinan Street, London N1 9XW, UK. Telephone: +44 (0)20 7833 4000. Fax: +44 (0)20 7843 4749. New subscriptions/renewals/changes

of address/back issues and all other customer service questions should be addressed to - North America: Nature Geoscience, Subscription Dept, PO Box 5054, Brentwood, TN 37024-5054, USA. Outside North America:

Subscriptions Department, Brunel Road, Basingstoke, Hants. RG21 6XS, UK. Telephone: +44 (0)1256 329242; Fax: +44 (0)1256 812358. Annual subscription rates: US/Canada US$3060, Canada add 5% GST (institutional/

corporate), US$152, Canada add 5% GST (individual making personal payment); UK/Rest of World (excluding Europe and Japan) 1570 (institutional/corporate), 78 (individual making personal payment); Europe 2430

(institutional/corporate),121 (individual making personal payment). Japan: contact Nature Asia-Pacific, Chiyoda Building, 2-37 Ichigayatamachi, Shinjuku-Ku, Tokyo, 162-0843, Japan. Telephone +81 3 3267 8751. For single back

issue prices contact the publisher. Nature Geoscience(ISSN 1752-0894) is published monthly by Nature Publishing Group, c/o Mercury Airfreight International Ltd, 365 Blair Road, Avenel, NJ 07001, USA. Periodicals postage is

paid at Rahway NJ. Postmaster: send address changes to Nature Geoscience, c/o Mercury Airfreight International, 365 Blair Road, Avenel, NJ, USA. Reprints: Nature GeoscienceReprints Department, Porters South, 4 Crinan Street,

London N1 9XW, UK. Subscription information is available at the Nature Geosciencehomepage at http://www.nature.com/naturegeoscience. Postmas ter: send address changes to Nature Geoscience Subscriptions Department,

Brunel Road, Basingstoke, Hants. RG21 6XS, UK or Nature GeoscienceSubscriptions Department PO Box 5054, Brentwood, TN 37024-5054, USA. 2009 Macmillan Publishers Limited. All rights reserved.

http://www.nature.com/naturegeosciencehttp://www.nature.com/naturegeoscience


5/80

NATURE GEOSCIENCE| VOL 2 | JUNE 2009 | www.nature.com/naturegeoscience 371

editorial

Science communication will never bethe same again. Te traditional path ofscientific discoveries from bench to breakfasttable through scientific journals andnewspapers is diversifying at a staggeringrate. Blogs by scientists and non-scientistsalike are jostling for public attention; researchinstitutions provide websites, films andpress releases; and scientific information isFlickring and wittering away in Web 2.0,accessible to anyone who is interested.

Given this variety of competition,

researchers work hard to make theirwork stand out from the crowd. Asession at the General Assembly of theEuropean Geophysical Union (EGU) inApril entitled Te Significance of Marineechnology in Science Communication Challenges and Opportunities highlightedan increasingly important selling point forscience in the public arena: the volumeof publicity that can be generated bystunning images.

Planetary scientists have long exploitedthe power of unique pictures that capture theattention of researchers and the public alike.

Te Cassini/Huygens mission to the Saturnsystem led to a veritable itan-fever amongamateurs, not least because all images wereimmediately posted on a freely accessiblewebsite. Te computer-literate enthusiastseven outcompeted the space agencies when itcame to producing animations of the surfaceof Saturns moon (Naturedoi:10.1038/news050117-7; 2005).

Stimulating the imagination of thenext generation of potential scientists withbeautiful or intriguing pictures shouldbe applauded. Many a good scientist hasbeen drawn into his or her field by an early

fascination with its imagery. And as publicfunding bodies require tangible societal

benefits in return for their generosity,widespread interest in a research project is awelcome boost for the next proposal.

Presentations at the EGU session madeit clear that marine researchers have nowcaught on. Remotely operated vehicles thatdelve into the deep ocean deliver fascinatingphotographs and films from a world whereno human has ever set foot. Te footageof bizarre deep-sea creatures and ancientshipwrecks is lapped up by journalists, whoknow that images sell stories. Now public

attention is being bestowed on deep-searesearch a field not previously known forits success with the general public.

Yet photographs gained throughtechnological advances are usuallytransmitted to the public throughstakeholders. Te aim of any press officeis to gain as much attention as possiblefor their research institution. Mediaoutreach personnel are not employed tobe disinterested chroniclers of scientificprogress, but to push the findings of thescientists at their university or lab to the topof the news agenda.

As the science sections in qualitynewspapers are shrinking, press offices arebeing expanded (Nature458,274277;2009). If this trend continues, reporters willhave less time to research their stories, butmore (and better) press releases to workfrom. As a result, general readers who wantto keep up with scientific progress mayno longer be able to rely on their morningpaper to deliver a fair overview of the bestresearch (as opposed to the work done atthe institutions with the most efficient pressoffices). Already at least, young scienceenthusiasts prefer to seek out their sources

from an almost infinite online choice, ratherthan from a limited number of newspapers

delivering the same news package to alltheir readers.

But unlimited choice is not always a goodthing. In a world where everyone solicits theirown sources of information, it will be muchharder to find a common basis for discussionwithin society. It is difficult for someoneoutside a field to judge which blog provides alevel-headed assessment, and which distortsthe facts to fit an agenda. Furthermore, thereis the danger that people will choose to buildsupport for their prejudices and preconceived

ideas, rather than allowing a trustednewspaper to challenge their establishedpatterns of thinking.

Te difficulty lies with communicatingthe more complicated aspects of science tomore than a niche of dedicated followers.Explaining progress in the context of earlierwork requires background knowledgeand research, as does writing about theuncertainties, contradictions and pitfalls thatare an integral part of science. o providea fair overview of the wealth of publishedpapers requires a disinterested observer.

Researchers blogs can provide a valuable

insider perspective on the progress ofscience. Press officers can (and should) bringthe work of the scientists they represent tothe attention of the media, and stunningimages are an excellent means of achievingthis goal. But it takes someone outside theresearch institutions to assemble all theavailable information into a coherent picture.

As the advertising-based business modelof the print media is crumbling in the faceof cheaper offers online, the need for full-time science reporters (writing online or inprint) remains. A society in which scienceand technology are central must provide the

funding to bring the narrative of scientificprogress to the public.

In the world of Web 2.0, the variety of channels for communicating science is exploding. Technology can help

to generate images that attract attention, but there is much more to reaching the public than pretty pictures.

Complex communication

N

OCS/JC10

MARUM,

UNIVERSITYOFBREMEN

MARUM,U

NIVERSITYOFBREMEN



6/80

372 NATURE GEOSCIENCE| VOL 2 | JUNE 2009 | www.nature.com/naturegeoscience

correspondence

To the Editor In their letter (NatureGeosci. 1,511514; 2008), Geert Lend-erink and Erik van Meijgaard study anhourly time series o precipitation dataobtained at De Bilt, Te Netherlands. Teyobserved an exponential increase o heavyprecipitation with temperature, with acoefficient close to that o the ClausiusClapeyron relation or lower temperatures,whereas or higher temperatures they oundsuper-ClausiusClapeyron behaviour.

We argue that the super-ClausiusClapeyron scaling or hourly, but notor daily, precipitation arises because othe superposition o two acts: (1) thedramatically different timescales betweenlarge-scale and convective precipitationand (2) the dominance o convectiveprecipitation or high temperatures andthe dominance o large-scale precipitationor cool temperatures. Our theory alsoexplains the unusual transition betweenClausiusClapeyron and super-ClausiusClapeyron scaling or hourly precipitationas the temperature changes.

It is o undamental interest tounderstand the origin o the changesin precipitation extremes reported byLenderink and van Meijgaard, considering,

or example, flood risk. o this end weconsider the wet-day probability densityunction o total daily precipitation(Ptot), which depends on precipitationintensity and temperature. Tis is thesum o the probability density unctionso convective precipitation (Pc) andlarge-scale precipitation (Pls). Byconvective precipitation we mean showeryrain that alls over a certain area or arelatively short time, or example, during

mid-latitude thunderstorm events insummer. Large-scale precipitation occurs,or example, due to slow ascent o air insynoptic systems.

Now we consider a given temperaturerange (T, T+ T) and reer to thetotal amount o precipitation withinthis range as its weighting. Generally,higher intensities are less likely thanlower intensities or both large-scale andconvective precipitation events. Dailyprecipitation is accumulated over 24 h, andits intensity is thereore an average overthis time interval.

When the transition is made tosub-daily temporal resolution (such ashourly in Lenderink and van Meijgaard)the different nature o large-scale and

convective precipitation may emerge romthe statistics. Whereas large-scale eventstake place at a lower rate during a largerpart o the day, convective events are likelyto occur as bursts o heavy rain during asmaller raction o the day. Te transitionto sub-daily precipitation then leads to amore pronounced stretching o Pctowardshigher intensities than is the case or Pls.Note that the rescaling leaves the weightingo (T, T+ T) unchanged. Hence, or

Ptotdifferent statistical behaviour andtemperature dependence will emerge ordaily, compared with sub-daily, resolution.In particular, any given percentile o large-scale and convective precipitation will shifby different offsets when the transition tosub-daily resolution is made. As large-scale(convective) precipitation dominates at low(high) temperatures, the correspondingpercentile o total precipitation willollow that o large-scale (convective)precipitation there. In the intermediatetemperature range an increase with anunexpected temperature dependence may

occur (see Supplementary Inormation ormathematical details).We illustrate our analysis in a simple

example where the daily amount o

Unexpected rise in extreme precipitation

caused by a shift in rain type?

0 2 4 6 8 10 12 14

Intensity (mm d1)

Intensity(mmd

1)

Intensity(mmd

1)

Daily

Sub-daily

T = 23 C

T= 13 C

0.1

0

0

0.05

0.02

0.04

0.06

0.080.1

Ptot

Prob.

density

Large-scale

TIs Tc

Convective

10

1

10

1

Sub-dailyDaily

99.9th99.9th

75th 75th

Convectiv

e Large-s

cale

0 5 10 15 20

T (C)

0 5 10 15 20

T (C)

0 5 10 15 20

T (C)

a

b

c d

Figure 1 |Daily and sub-daily precipitation intensity distributions, weighting functions and precipitation percentiles. a, Probability density function of total

precipitationat T= 13 C andT = 23 C. Solid lines are for daily values of total precipitation intensity, dashed lines are for sub-daily rescaling, orange arrows

indicate 75th percentile. b, Weighting functions for large-scale (red) and convective (blue) precipitation as function of temperature: Tls= 13 C and Tc= 23 C

are indicated by vertical lines. c, 99.9th (upper curves) and 75th (lower curves) sub-daily precipitation intensity percentile of large-scale (red), convective (blue)

and total precipitation (black), and double ClausiusClapeyron increase (dotted grey); arrows indicate onset of super-ClausiusClapeyron behaviour. d, Same as

cbut for daily averaged precipitation intensity: all curves collapse on one, slope does not change as function of temperature. Note the logarithmic vertical scale

in c and d.



7/80


correspondence

convective (large-scale) precipitationoccurs constantly during 12 (24) hours othe day. For the two weighting unctions wechose equal Gaussian distributions centredat different temperatures Tls< Tc(Fig. 1b).We use identical daily distributionunctions (p, T) exp(p/(T)) where

pis precipitation intensity, Tis temperatureand is the scale parameter with equalClausiusClapeyron-like exponentialtemperature dependence(T) = exp(bT)where the coefficient o temperaturedependence b= 0.07 K1. Hence, theonly variations result rom the differencebetween Tlsand Tc, and the transition tosub-daily statistics (as shown in Fig. 1aand b). Te percentiles o the transormedPls, Pcand Ptotare depicted in Fig. 1c.Whereas the distribution unctions othe two types o precipitation have aClausiusClapeyron-like exponential

increase o a given percentile intensity,the joint distribution may deviate romthe ClausiusClapeyron-like behaviourin the transition temperature regionwhere the two weighting unctionsoverlap. For the limit o low and hightemperatures, ClausiusClapeyronbehaviour is approached. Tis eature can

be observed in sub-daily data, whereasit disappears on a daily scale (compareFig. 1c and d). Tis is precisely whatLenderink and van Meijgaard ound,both in observational and model results(compare Fig. 1a and c in their paper).Additionally, their Fig. 1a shows a

transition rom ClausiusClapeyron tosuper-ClausiusClapeyron behaviour withan onset moving to lower temperaturesor higher percentiles, which can alsobe seen in our Fig. 1c. We tested theanalysis presented here on model dataand ound generally consistent results(see Supplementary Inormation).

In conclusion, we have provideda general argument which makes theorigin o the unexpected increases inextremes obvious by simple histogramre-weighting or sub-daily values.Super-ClausiusClapeyron behaviour

emerges as a consequence o simultaneousClausiusClapeyron behaviour o both thelarge-scale and convective precipitationin the temperature regime where the twotypes coexist. Outside o this regime theClausiusClapeyron behaviour againemerges. Concerning precipitation in achanging climate, we suggest investigating

whether seasons and regions with mainlylarge-scale or convective precipitationsee only a ClausiusClapeyron increasein extremes with temperature. Forseasons and regions with coexistenceo the two types, our analysis stressesthe importance o studying whether the

temperature o their transition shifs withchanging climate.

AcknowledgementsTis study was partly unded by theEuropean Union FP6 project WACH(contract number 036946). We acknowledgethe HadRM group or providing dataunder the ENSEMBLES project. We thankA. Haensler, S. Hagemann, D. Jacob and. Stacke or ruitul discussions.

Additional informationSupplementary inormation accompanies thispaper on www.nature.com/naturegeoscience.

J. O. Haerter1*and P. Berg2

1Max Planck Institute for Meteorology, 20146

Hamburg, Germany, 2Institute for Meteorology

and Climate Research, University of Karlsruhe

and Forschungszentrum Karlsruhe, 76344

Karlsruhe, Germany.

*e-mail: [email protected]

Lenderink and van Meijgaard reply Intheir correspondence, Haerter and Berg pro-pose an interesting explanation o the super-

ClausiusClapeyron scaling o precipitationextremes we ound in hourly observationsat De Bilt, Te Netherlands and reported inour letter (Nature Geosci. 1,511514; 2008).Tey argue that because convective pre-cipitation events are by their nature moreintense than large-scale events, a changein relative requency o occurrence o bothprecipitation types (histogram re-weighting)influences the scaling o precipitation ex-tremes with temperature. Tey show that inan intermediate temperature range, linkingthe two precipitation regimes, the statisticaleffect o histogram re-weighting may give

rise to a super-ClausiusClapeyron scaling,even when the scaling o the large-scale andthe convective events separately both satisythe ClausiusClapeyron relation.

For the most extreme precipitationevents histogram re-weighting is onlyrelevant when the number o convectiveevents is relatively small comparedwith the number o large-scale events.For temperatures at which the numbero convective events is larger than (orequal to) the number o large-scaleevents, the extreme 99.9th percentileis dominated by the scaling o theconvective events (compare Fig. 1b and

c in the correspondence rom Haerterand Berg). Te temperature range wherethe super-ClausiusClapeyron relation is

obtained is thereore primarily determinedby the ratio between the number oconvective and large-scale events asa unction o temperature. Tis ratioollows rom an arbitrary choice in theconceptual model o Haerter and Berg,not supported by observations and alsoconsiderably different rom the climatemodel results (compare their Fig. 1b andSupplementary Fig. 1b).

We think that the maniestation o thesuper-ClausiusClapeyron scaling has aphysical origin, rather than the statisticalorigin proposed by Haerter and Berg;

in our opinion it is a property o theconvective regime itsel, and results romthe dynamics o convective clouds withstronger updrafs due to increased latentheat release as the temperature rises. Tesuper-ClausiusClapeyron scaling osub-daily precipitation extremes is a robusteature in the observations at De Bilt. Itis consistently obtained using differentmeasures o the temperature (daily meanand daytime maximum), different timeperiods (all year, summer months, summerhal-year), and different measures o thesub-daily precipitation intensity (meandaily intensity at wet hours, hourly

intensity, and daily maximum o the hourlyintensity). Considering the mean dailyintensity at wet hours (daily sum divided

by rainall duration) as in Haerter andBerg, a dependency exceeding two timesthe Clausius-Clapeyron relation is oundin the observations, without any clear signo levelling off in the high temperaturerange (as would be expected romhistogram re-weighting).

Te atmospheric conditions mayvary considerably across the range otemperatures or which the scaling has beenobtained. Tis does not only determine thenumber o rainall events(Nature Geosci.1,511514; 2008), whether precipitation islarge-scale or convective (as mentioned in

Haerter and Berg), but could also influenceother aspects o shower complexes, suchas their level o mesoscale organization ortravel speed. Tese actors potentially affectthe scaling relations ound or the present-day climate and their interpretation inthe context o climate change althoughwe think not in a crucial way and needurther investigation.

Geert Lenderink* and Erik van Meijgaard

Royal Netherlands Meteorological Institute

(KNMI), 3730 AE De Bilt, PO Box 201,

The Netherlands.

*e-mail: [email protected]



8/80


commentary

Securing the legacy of the IPYBob Dickson

Paradoxically, as the International Polar Year ends we enter its most important phase. Now we must

decide and quickly which mix of observations to sustain, based on what we have learnt.

Acouple o months ago, it did seem alittle bizarre to be asked to addressan international workshop1on the

climate o Spain with a talk about observingthe Arctic during the International PolarYear (IPY). But in act I neednt haveworried. Nobody lef. Te delegates got thepoint. Not just that extreme change was

passing through the oceanatmospherecryosphere system o our northern seas,interesting though that was. Te interesto the organizers was ocused much moreon a whole complex o physical processesthat were in some way implicated in drivingthese extreme changes through polarseas: processes that the IPY had set out toobserve and understand, processes that arenot yet represented realistically (or at all) inclimate models.

In act, climate models are inherentlyweak in quite a long list o processesimportant to our understanding o how

change takes place in northern seas andhow it might affect climate. Tese include,as just one example, most aspects o oceancirculation, mixing and exchange on thebroad circum-Arctic shelves (climatemodels have no continental shelves!). Wecould list many more2. Although theseprocesses may range down to relativelysmall scales in space and time, they allhave at least a potential importance orthe prediction o Arctic change. Andthe hope was that many or most othem would have been drawn into theobservational net o the IPY, thus (one

day!) materially improving our ability tosimulate climate.Te environment during the IPY

(March 2007 to March 2009), was in astate o spectacular change. Even thoughconditions over the Arctic were verydifferent during 20002007 compared withmost o the twentieth century describedas unique and given its own special label,the Arctic Warm Period3 the eventsduring the first year o the IPY (2007) weresomething else again, with a persistent andstrong meridional airflow directed acrossthe North Pole rom the Bering Strait ormuch o that summer (locations shown

in Fig. 1). Entering stage lef, and partlyin response to this airflow, the oceanicheat flux passing into the Arctic Oceanthrough Bering Strait in 2007 was at amaximum since records began in 1990,only slightly warmer than normal but withincreased transport (R. Woodgate, personalcommunication). Entering stage right,

our longest hydrographic series confirmthat the warm saline Atlantic currentpassing into the Nordic seas through theFaroeShetland Channel and continuingnorth to enter the Arctic Ocean throughFram Strait and the Barents Sea was at itswarmest or more than 100 years.

Te observing effort that was deployedto meet these changes was in many waysequally spectacular4. But it takes nogreat thought to realise that rather thanthe two-year project itsel, it will be thelegacy phase o the IPY, sustained overyears to decades, that will develop our

understanding o these processes, theirchanges, their eedbacks and their likelyclimatic impacts to the point where theycan be o use to climate models. Plainly,we cant continue everything. What havewe have learnt in the IPY that can helpus design its legacy phase? At the ArcticScience Summit Week in Bergen in March5,the Arctic Ocean Sciences Board setitsel the task o developing a ully costedproposal by the time o the post-IPYconerence in Oslo6in June 2010.

Tere has been little-enough time asyet to assess the data, but hal a dozen

examples illustrative o the ways ourideas have changed or developed duringthe IPY will help to define the scope otheir task.

Six new ideas from IPYFirst, the inputs and a new capability inmeasuring them. Te Norwegian Atlanticcurrent is the principal oceanic transportero heat, salt and mass to the Arctic Ocean.Where it passes west o Norway, we haveknown or some time that it consists otwo branches, but whereas one branchhas been measured or >12 years, theother had not been measurable beore

the IPY. Now the ingenious combiningo hydrography (water sampling),satellite altimetry, conventionalmoored current measurements,seaglider transects and modellingin the Norwegian flagship projectiAOOS-Norway part o the integratedArctic Ocean Observing System has

or the first time provided a completemeasure o both branches (K. A. Mork &. Skagseth, personal communication;computer algorithms to calculatevolume transports based on seagliderand other data are available rom re. 7).As a result our estimates o these keyocean fluxes have effectively doubled.Furthermore, as the Norwegian projecthad intended, the main measurementline off Sviny has been developed intoa complete, sustainable, simple androbust upstream reerence system ormonitoring Atlantic inflow towards the

Arctic Ocean.Second, a change in our conceptualrole or these inputs. When the IPY beganin March 2007, it would probably havebeen the consensus view that a 100-yearmaximum in the warmth o the inflow tothe Arctic must in some way be bound upwith an increased melting o sea ice. Sincethen our ideas have altered in responseto new simulations by a group rom theAlred Wegener Institute (M. Karcher,personal communication), which suggestthat as the warm Atlantic-derived layerspread at subsurace depths through

the Arctic deep basins it did so at asignificantly greater depth than normal.Tough the increased warmth may thusbe too deep to have much effect on the seaice, the intriguing suggestion is made that,as and when this layer circuits the Arcticand drains south again into the Nordicseas, its changed depth and density nowseem capable o slowing the overflow odense water through the Denmark Strait,hitherto regarded as largely unchanging.Tus, in the Atlantic sector at any rate,the climatic impact o the recent inflow owarmth to the Arctic may have less to dowith local effects on sea ice than on the



9/80


commentary

Atlantics thermohaline conveyor, yearslater and ar to the south. As a candidateor the IPY legacy phase, the importanceo this result seems clear. Maintainingsurveillance on an evolving change takingplace throughout the length and breadtho our Arctic and subarctic seas on a

timescale o decades will probably provehighly instructive to our understanding othe role o our northern seas in climate,although detecting and ollowing suchdecadal transient signals is likely to imposea need or new tools in observationalnetwork design.

New thinking has also illuminatedthe old issue o how the signals o oceanclimate change may be carried into andthrough the Arctic deep basins romsubarctic seas. Warm, salty Atlantic-derived waters enter the Arctic Oceanin two branches, the relatively cool

and resh Barents Sea branch and thewarmer, saltier Fram Strait branch. Andthe greatest change in our thinkingconcerns the new idea put orwardby Bert Rudels (Univ. Helsinki) at theArctic Science Summit Week5that theBarents Sea branch may be the one thatdominates the Arctic Ocean beyondthe Nansen basin, with the Fram Straitbranch seldom penetrating beyond theLomonosov ridge. I so, the source o therecent warming graphically describedalong the boundary o the Laptev Sea andCanadian basin8,9 has effectively been

reassigned. esting Rudels idea will bean important task or the legacy phaseto resolve, but the tools to do so are wellproven: detailed ship-borne hydrography,sustained flux measurements throughthe northeast Barents Sea, and continuedor intensified coverage o the boundarycurrents along the Eurasian margin othe Nansen basin rom the point whereboth branches first flow together to theirsupposed points o separation at theLomonosov ridge.

Our ourth example highlights howtechnical advances in the use o untended

instruments have led to a huge growthin the ocean data set. As one example oan increasingly elaborate array o theseinstruments (see Fig. 2), the expanded useo conductivitytemperaturedepth (CD)profiler systems10,11moored to the drifingice has, since 2004, contributed around18,000 high quality temperaturesalinityprofiles to the data set o the Arctic Ocean(J. oole, personal communication),transorming this ormer data desert intoone o the most densely observed oceanson Earth. Te continuing analysis o thesedata (B. Rabe, personal communication)has mapped out the pan-Arctic distribution

o reshwater content, confirming earlierconclusions12that the Beauort gyre is

the largest marine reservoir o reshwateron Earth. Whats more, the analysiso the Woods Hole OceanographicInstitution Beauort Gyre ExplorationProject13has revealed that this centre is asystem in rapid transition, with stronglyincreasing trends in its reshwater contentbetween 2003 and 2008. As the efflux oreshwater rom the Arctic is expected tobe one way in which Arctic change mayreach south to affect the Atlantic conveyor,these are findings o direct relevance toclimate. And with such a key parameter insuch rapid transition, it would be barelyconceivable to enter the legacy phase

without ice-tethered profilers at the core oits observing system.

Fifh, we have learnt some o the keyobservational requirements or improvedsea-ice prediction. Te 22 ice-predictiongroups that participated in the SEARCH-or-DAMOCLES Sea Ice Outlook exercise14concluded that an improved measureo ice thickness in spring was the primerequirement or improved predictiono ice extent at the time o the latesummer minimum. Such an unequivocalrequirement must surely commend the use,in the legacy phase, o the large numbero new techniques (above, on and beneaththe ice) that are now available to assess icethickness. Examples o such techniques

Bathymetric and topographic tints

5,500 4,500 3,500 2,500 1,500 500 300 200 100 20 0

Barents

Sea

Lom

onos

ovridg

e

Canadian

basin

Denmark Strait

Nan

senba

sin

Beaufort

Sea

Laptev

Sea

Nordic

seas

Bering

Strait

Atlantic

current

Fram

Strait

(m)

Figure 1 |The Arctic-subarctic domain showing the system of ridges, shelves and deep basins. The Bering

Strait, Fram Strait and Barents Sea carry the main oceanic inflows to the Arctic Ocean; the gyre of the

Beaufort Sea is the largest marine reservoir of freshwater on Earth; and the Denmark Strait carries both a

main outflow of freshwater from the Arctic and the principal overflow of cold dense water by which the

deep North Atlantic Ocean is renewed. Because of their known or suspected importance to climate, all of

these features are monitored and are strong candidates for continued or expanded monitoring in the IPY

legacy phase. Base map courtesy of Igor Polyakov, Univ. Alaska.



10/80


commentary

include laser and radar altimetry onICESA and ENVISA, tiltmeter buoysdeployed on the ice surace, and thesub-ice floats fitted with upward-lookingsonar that are coming to ruition in

EC-DAMOCLES.Finally, in what may be the big dealo the IPY thus ar, we are beginning torecognize effects o the ice-ree polarocean on the regional and hemisphericatmospheric circulation. As the ice inthe Pacific sector o the Arctic meltedback to its record minimum in summer2007, and the heat storage o theunderlying ocean increased, the release othis heat in autumn was ound to erodethe stratification o the atmosphere toprogressively higher levels, leading to aclear change in the regional atmosphericcirculation15. A similar association o

events also seems to be implied in thereport16that winter Northern Hemisphereweather patterns seem to remembersummer Arctic sea-ice extent. With suchdirect apparent links to polar climate, it

makes sense to enquire, in planning anobservational legacy phase or the IPY,what continued coverage o the upperwater column would be needed to keeptrack o oceanatmosphere heat exchangeas the sea ice dwindles away.

aken together, these ew examplesmake a urther insistent point: that thesubarctic seas are intimately involved; thatyou cant understand Arctic change just bystudying the Arctic.

The road to OsloFrom the point o view o understandingthe role o our northern seas in climate,

these hal-dozen insights alone might havejustified the IPY; but they are intendedas examples only, not an exhaustive list.No doubt with time some o these ideasmay ounder, others will take their place.Te basic need is or a plan capable ointegrating national efforts to provide

a reely shared, flexible, reliable andreasonably complete ocean data set. Itwould detract rom the effectiveness othis pan-Arctic effort i national issues osovereignty, access and logistics promotea dataset whose coverage, quality, cost andaccessibility would be a variable unctiono the country responsible. Needless to say,even when the science o the legacy phaseis agreed, a range o external actors willbe on hand to constrain it, including thickice, orbidden or costly access, icebreakeravailability and technical challenges.Whether top-down or bottom-up solutions

are more appropriate will be defined bythese external constraints, as will thedebate on the type o international body,long-term international coordination andlogistics planning that will be required tominimize them.

I we are to meet the plan o theArctic Ocean Sciences Board to present adetailed and ully costed proposal or theIPYs legacy phase at the 2010 post-IPYconerence in Oslo, we have roughly oneyear to decide.

Bob Dickson is Emeritus Research Fellow at the

Centre or Environment, Fisheries and AquacultureScience (CEFAS), Lowestof NR33 0H, UK.

e-mail: [email protected]

References1. http://clivar.iim.csic.es/?q=es/node/269

2. Dickson, R. R., Meincke, J. & Rhines, P. (eds)ArcticSubarctic

Ocean Fluxes: Defining the Role o the Northern Seas in Climate

(Springer, 2008).

3. Overland, J. E., Wang, M. & Salo, S. ellus

60A,589597 (2008).

4. Dickson, R. R. Te integrated Arctic Ocean Observing System in

2007 and 2008; available rom .

5. http://www.aosb.org/assw.html

6. http://www.ipy-osc.no/

7. http://dokipy.met.no/projects/iaoos-norway/owsm-glider.html.

8. Polyakov, I. et al.Eos88,398399 (2007).9. Dmitrenko, I. A. et al. J. Geophys. Res.

113,C05023 (2008).

10. Krishfield, R. et al.J. Atmos. Ocean. ech.25,20912105 (2008).

11. Kikuchi, ., Inoue, J. & Langevin, D. Deep-Sea Res. I

54,16751686 (2007).

12. Carmack, E. et al.inArctic-Subarctic Ocean Fluxes:

Defining the Role o the Northern Seas in Climate

(eds Dickson, R. R., Meincke, J. & Rhines, P.)

145170 (Springer, 2008).

13. Proshutinsky, A. R. et al. J. Geophys. Res.(in the press).

14. Search Sea Ice Outlook: Summary Report Overview and

Highlights (2008); available at .

15. Overland, J. E. & Wang, M. ellus(in the press).

16. Francis, J. A., Chan, W., Leathers, D. J., Miller, J. R. &

Veron, D. E. Geophys. Res. Lett.36,L07503 (2009).

Figure 2 |The Super Buoy Array at 86 34 N, 134 39 E on the Lomonosov ridge of the central

Arctic Ocean, after deployment by F/S Polarsternin September 2007 perhaps the best example

of the elaboration and integration of Arctic observing systems that was such a strong feature of the

IPY. This particular ice-top observatory combines: an autonomous CTD profiler (ITP; the Woods Hole

Oceanographic Institution Ice-Tethered Profiler10), which has rapidly transformed the Arctic Ocean from

a virtual data desert to one of the best-described oceans on Earth; an Ice Mass Balance Buoy, (IMB;

Don Perovich and Jackie Richter-Menge, CRREL) designed to measure variations in surface and basal

ice-melt during their drift of up to two years; an Autonomous Ocean Flux Buoy (AOFB; Tim Stanton,NPS Monterey) designed to provide accurate measurements of vertical heat fluxes through the upper

ocean; and a prototype Ice-Tethered Acoustic Current profiler (ITAC; Alfred Wegener Institute) designed

to provide regular profiles of ocean current velocity to 500 m as the system drifts through the polar

sea. Photos reproduced with permission from Lasse Rabenstein, Lars Gremlowski, Benjamin Rabe and

Ursula Schauer, Alfred Wegener Institute for Polar and Marine Research.

ITAC

AOFB

ITP

IMB

mailto:[email protected]:[email protected]


11/80


books & arts

Werner Herzog has always

been interested in extremeenvironments, and in therelationship between manand nature. In Encounters atthe End of the World, Herzogfinds a new way to lookat the Antarctic, findinginsights in the stories othe people who live there,and showing us beauty in

unexpected places.With equal measures o wit and curiosity,

Herzog interviews some o the peopleworking in this remote corner o the globe.

Among the glaciologists and biologists, wemeet a philosopher-turned-orklif-driver, alinguist-turned-gardener and a plumber oroyal descent.

First stop is the town o McMurdo, theheart o the US Antarctic programme, andhome to a thousand or so people. Full oconstruction sites and orklif trucks, thisrather bleak and muddy town is a ar cryrom the images o Antarctica we are usedto seeing on television. But the drabness othe place belies the extreme likeability o thecharacters that live there.

One such inhabitant a part-Apache

plumber has immense pride in his fingers.Once told by an anthropologist that theseare the hands o Aztec royalty, he holds themup again and again or the camera tosee. Unortunately, he doesnt explain what isunique about them.

Another resident, an East Europeanmechanic, lived most o his lie like aprisoner behind the Iron Curtain. Chokedback with emotion, he is unable to talkabout his lie then. Instead, he shows us thecontents o his rucksack, which he keepspacked at all times, always ready to escapeor to go on new adventures and explore

new horizons.Outside McMurdo we meet the scientists.Particularly moving is an encounter with acell biologist on mainland Antarctica. Tebiologist, who is about to embark on his lastdive into the Southern Ocean, sits pensively,preparing himsel or his final encounterwith this underwater world. When drawninto conversation by Herzog, he muses onthe unexpected violence and horrors o thathidden world. Te creatures he describesare like something out o a science fictionnovel. One uses its tentacles to ensnareits prey the more the prey struggles,the tighter the tentacles grow. Eventually

the victim exhausts itsel and the creaturemoves in or the kill. Te camera ollowshim on his dive; rather than terriying, it ishauntingly beautiul.

Another touching moment is Herzogsmeeting with a penguin specialist; a reclusiveman who has been studying these creaturesor 20 years. Although not orthcomingwhen it comes to human interaction, hebreaks into a smile when Herzog probeshim on the subject o gay penguins, andexpounds with great amusement on thedynamics o penguin prostitution.

Watching the film, it becomes clearthat there must be something very specialabout Herzogs manner that allows peopleto interact with him in an achingly honestand open way. But he doesnt pander to

the rivolous or sel-indulgent. So when atraveller talks in painstaking detail abouther experiences in a garbage truck inArica, her voice ades out and we hearHerzog in his sof, affable way herstory goes on orever.

But its not just the people that captureyour attention. Although Herzog didntwant to make a film about fluffy penguins,one o the most heart-breaking scenes iswhen a penguin aced with the decisiono whether to ollow his companions to thewaters edge or back to the colony shufflesround in a circle and then heads straightor the mountains and the interior o

Antarctica. Watching the penguin waddleto its death, you will them to interject. But,as our narrator explains, it wouldnt make adifference its mind is made up.

Herzog uses these small stories as a wayin to the majesty, epic scale and strangenesso Antarctica. As he intersperses personaltales with ootage o towering ice cliffs, andicy underwater scenes o colossal beauty,you understand the tremendous andcontagious reverence Herzog has orthis place.

Te topic o climate change is scarcelytouched upon directly, but the theme ohuman ragility and destruction runs rightthrough. From the apocalyptic films theresearchers watch in their spare time toa volcanologists words o warning about

the planets catastrophic past, it is hardto escape. One researcher says that in hisdreams he can hear the iceberg screeching,this iceberg is coming north.

In the final scene, the philosopher-turned-orklif-driver quotes Alan Watts,we are the witnesses through which theuniverse becomes conscious o its glory.Perhaps Herzog shows us a touch o thisglory here.

ANNA ARMSTRONG

Encounters at the End of the World

by Werner Herzog, Discovery Films: 2007.

UK release date: 24 April 2009.

FILM

Tales from Antarctica

REVOLVERENTERTAINM

ENT



12/80


research highlights

Feeling the heatLithos doi: 10.1016/j.lithos.2009.04.002 (2009)

Geochemical and textural evidence suggeststhat pieces of the Earths crust picked up bymagma may at least partially melt in a matter

of hours. These crustal xenoliths, as theyare called, are then carried to the surface byvolcanic eruptions, where they are prized forthe insights they provide into the processesthat alter magma composition.

Cliff Shaw of the University ofNew Brunswick in Fredericton, Canadacompared the mineralogy, texture andchemical composition of crustal xenolithsand host lava from the Rockeskyllerkopfvolcanic complex of the West Eifelvolcanicfield in Germany. The result indicated thatthe glassy rinds that coated many of thexenoliths formed in just 12 hours or so, as the

silica-rich fragments reacted with the hot,silica-poor magma.

Xenoliths from this particular region areunique in that they are almost completelyseparated from the surrounding rock. Shawsuggests that the creation of gas bubblesduring rind formation separated thexenoliths from the enveloping lava.

Restless MoonJ. Geophys. Res. 114,E05001 (2009)

Clusters of quakes in the Moons deep

interior called moonquakes aregenerally attributed to deformationfrom the Earths gravity. Although this isprobably the case for some quakes, a seismicreanalysis shows that others may be causedby mineralogical changes deep within theMoons interior.

Renee Weber, of the US GeologicalSurvey, Flagstaff, and colleagues reanalysed39 moonquake clusters that were recorded bylunar-based seismometers during the ApolloPassive Seismic Experiment, from 1969 to1972. About a third of the quakes could beattributed to failure along discrete planes ofweakness, triggered by tidal deformation

arising from Earths gravitational pull.However, this mechanism could not explainthe remaining moonquakes.

The structure of minerals within theMoons deep interior may evolve in responseto increasing pressure with depth, as on

Earth. The team speculates that that thesestructural changes could generate stressesthat would give rise to some moonquakes.

A tough shellGeol. Soc. Amer. Bull.121,688697 (2009)

Desert pavements are hard surfacesconsisting of a mosaic of pebbles, and arecommon in very arid regions. Dating of suchpavements from the Negev Desert of Israelshows that these particular pebbles haveresisted the forces of erosion for well over amillion years.

Ari Matmon of the Hebrew Universityof Jerusalem and colleagues measured

exposure ages from multiple samples ofdesert pavement at two sites separated byover eight kilometres in the Paran Plainsregion. Whereas individual samples ofpavement from other deserts have suggestedlong exposure times at single locations, thesamples of Negev pavement indicate thatmost of the surface has been exposed for upto 1.8 million years.

The ages of the Negev pavement pointto a landscape that has been exceptionallystable for a long period of time, probablybecause of a combination of environmentalconditions such as extreme dryness, surfaceflatness and a lack of significant tectonic

disturbance, as well as the armouring effectof the pavement.

Carbon consumptionProc. Natl Acad. Sci. USA106,70677072 (2009)

The Intergovermental Panel on ClimateChange projections show that the upperocean is likely to warm by 16 C by theend of this century. Experiments show thatthis warming could decrease the amount ofcarbon transported to the deep ocean.

Julia Wohlers, of the Leibniz Instituteof Marine Sciences, Germany, and

colleagues exposed natural planktoncommunities to temperature increasesof up to 6 C in an indoor mesocosmexperiment. Under the maximum warmingscenario, net consumption of dissolvedinorganic carbon fell by 31%. Theirmeasurements indicate that this was notdue to a reduction in photosynthetic uptake,but rather to an increase in respiratoryconsumption of the photosyntheticallyderived organic carbon, which re-releasesdissolved inorganic carbon. Furthermore,warming increased the concentrationof dissolved organic carbon relative to

particulate organic carbon, making theaccumulating material less susceptibleto sinking.

The researchers warn that an increasedconsumption of organic matter in thesurface ocean, combined with a reductionin the amount of carbon sinking to depth,could increase the atmospheric load ofcarbon dioxide.

2

009GSA

The last glacial period was punctuated by abrupt transitions to interstadial (warm)

conditions. An analysis of an event 38,000 years ago as recorded in the ice core from

the North Greenland Ice Core Project reveals that mid-latitude climate change preceded

Greenland warming by several years.

Elizabeth Thomas of the British Antarctic Survey and colleagues used multiple proxies

to reconstruct climatic conditions during this abrupt warming, one of the most prominent

of the last glacial period. The ice cores annual layers showed that the approximately 11 Cwarming over Greenland occurred over about 26 years. However, the team also found that

a few years before the warming kicked in, the dust supply from Asia declined, which they

relate to a strengthening of the summer monsoon. At about the same time, there was a shift

in the hydrogen and oxygen isotopes of the ice, suggesting a northward migration of the

polar front.

The lag between the strengthening of the Asian monsoon and Greenland warming

could point to a trigger for glacial abrupt climate change in the tropics or the Southern

Hemisphere, rather than the north.

Interstadial timing J. Geophys. Res. 114,D08102 (2009)ISTOCKPHOTO/CAROLINASMITH



13/80


news & views

The solid inner core, with an averageradius of 1,220 km, is the mostremote part of the Earth and would

seem to be an unlikely place to find arecord of the evolution of our planet. Yet,among all of the primary subdivisions ofthe Earth, the inner core seems to be the

youngest and possibly the most rapidlygrowing. Te inner core has grown to itspresent size in less than 2 billion years1,and possibly within 600 million years2.Calculations by Deguen and Cardin,discussed on page 419 of this issue,show that the structure of the inner coreprovides clues about the later stages of thechemical differentiation of the Earth3. Moresurprisingly, it preserves evidence of active,gravity-driven tectonics.

Considerations of the rate at which heatis being lost from the planets deep interiorindicate that the inner core is crystallizing

from the iron-rich, fluid outer core at arate of 45 million kilograms per second2.Te rapid growth traps heat and chemicalheterogeneity, and the inner core hasacquired a remarkably complex structure.For example, recent seismic imaging hasrevealed differences between the deepestparts of the inner core4,5and regions closeto the inner-core boundary the contactbetween the inner and outer cores. Whereasthe innermost parts seem to be anisotropic,the region immediately below the inner-coreboundary shows an isotropic texture aswell as differences between the eastern and

western hemispheres

6

.As the inner core solidifies throughcrystallization, it develops a radial thermalgradient due to the release of latentheat: that is, it is hottest in the centralpart. Tis gradient tends to destabilizethe inner core by inducing convectiveoverturning. But the inner core also hasradial gradients in the concentrations oflight elements such as oxygen, sulphurand silicon, which are enriched in theouter parts of the inner core during thesolidification process. Tese gradients leadto a stabilization of the density distributionthroughout the inner core and thereby

act to prevent convective instabilities. Tetwo processes act in competition witheach other, and their relative importancedepends on the particular stage in the innercores growth.

Another complication in the growth

process is introduced by the non-uniformrate of solidification at the inner-coreboundary. Owing to the influence of fluidmotions in the outer core, the removalof heat from the inner-core boundary isexpected to be greater near the inner-core equator than at other latitudes. Teinner core therefore cools and crystallizespreferentially in the equatorial regions;as a result, these regions are expected tobe topographically higher7. Preferentialheat loss from the equatorial regions isrelated to the northsouth alignment of thegeomagnetic dipole, although the highertopography at the inner-core equator

produced by the more rapid crystallizationinduced by such heat loss has not beendirectly observed.

Deguen and Cardin use numericalsimulations to determine how these variousprocesses, including the thermal gradientsand light element distribution as well as

preferential crystallization at the equator,conspire to produce the heterogeneousinner core that seismology reveals3. Teymodel the evolution of the inner corefrom its inception to its present state.Te calculations suggest that when theinner core was very small, preferentialsolidification at the equator and thermalconvective overturning produced a nearlyuniform solid texture. At this stage, noperceptible compositional stratificationdue to enrichment of light elements in theuppermost regions had developed.

Te situation seems to have changed,

however, as the inner core approached itspresent-day size. Te researchers suggestthat as the growth rate decreases, thereis enough time available for a gradient inlight element concentration to develop, thestabilizing effects of which now come todominate inner-core evolution. Instead ofthe thorough convective churning typicalof the early stages, thin sheets slide off theequatorial topographic high and are thrustover one another, just below the inner-coreboundary (Fig. 1). Tis deformation isloosely analogous to the style of tectonicsseen in mountain belts, where thin slabs

of the continental upper crust are stackedupon each other.Te evolutionary model proposed by

Deguen and Cardin3may help to explainthe structural variations in the fabric ofthe inner core. For example, the shearinginduced by the tectonic activity in theoutermost parts of the inner core over thepast 100200 million years may help tohomogenize the material just beneath theinner-core boundary, thereby providinga possible mechanism for the isotropyinferred from seismic data. Te model alsoimplies that the innermost regions wouldnot be affected by deformation during

GEOPHYSICS

Tectonics in the Earths coreThe complex three-dimensional structure of the Earths solid inner core reveals how it has grown through time.Numerical simulations of the solidification process suggest that part of this structure has resulted from recent

tectonic activity.

Peter Olson

Innermost inner core

Inner-core boundary

Figure 1 |Inner core tectonics. The schematic

sketch shows gravity-driven overturning in a

shallow layer below the inner-core boundary.

Deguen and Cardin suggest that late in the

history of inner-core growth, this activity becamedecoupled from the deeper parts owing to

stable compositional stratification imparted

by the distribution of light elements3, which is

indicated by shading. The white curve denotes the

topographically elevated inner-core equator.



14/80


news & views

the later stages of inner-core growth. It istherefore plausible that the fabric in thedeepest parts of the inner core is an ancientone, potentially providing informationabout ancient processes.

Of course, the simulations do notcement the case for active tectonics on

the top of the inner core. Te researchersexplanation is intriguing, but complicatingfactors remain. For one, the seismicexploration of the inner core is stillincomplete and the proposed model cannotyet be fully tested. And almost certainly,tectonic deformation is not the only

process that affects the structure and fabricof the inner core8.

Interpreting the three-dimensionalstructure of the Earths inner core and itsevolution is a difficult problem, especiallybecause multiple physical processes forexample, convection, solidification and

grain deformation operate simultaneously.Deguen and Cardins modelling work3not only provides new insights intothe history of Earths deep interior, butmay also be a first step on the way tounderstanding the inner cores of otherplanets in the Solar System.

Peter Olson is in the Department of Earth and

Planetary Sciences, Johns Hopkins University,

Baltimore, Maryland 21218, USA.


References

1. Buffett, B. et al.J. Geophys. Res. 101,79898006 (1996).

2. Nimmo, F. in Treatise on GeophysicsVol. 8 (ed. Schubert, G.)

Ch. 2, 3165 (Elsevier, 2007).

3. Deguen, R. & Cardin, P. Nature Geosci. 2, 419422 (2009).

4. Cao, A. & Romanowicz, B. Geophys. Res. Lett.

34,L08303 (2007).

5. Sun, X. & Song, X. D. Earth Planet. Sci. Lett. 269,5665 (2008).

6. Niu, F. L. & Wen, L. X. Nature410,10811084 (2001).

7. Yoshida, S. et al.J. Geophys. Res. 101,2808528103 (1996).

8. romp, J.Ann. Rev. Earth Planet. Sci. 29,4769 (2001).

Varying incoming solar radiation, drivenby changes in the Earths orbit aroundthe Sun and in the tilt of its rotation

axes, is the primary natural climate forcingon timescales of millennia to hundreds ofthousands of years. Climate proxy data,

however, show that temperature changes donot simply follow the orbital forcing, so theclimate response to the forcing must also beaffected by internal feedbacks in the climatesystem related to slow components suchas ice sheets and biogeochemical cycles1,2.Tese feedbacks seem to be relevant for the100-thousand-year cycle of the recent glacialto interglacial transitions, which does notcoincide with the frequencies of the strongestorbital forcing. Additionally, millennial-scaletemperature variability across the NorthernHemisphere in the early part of the Holocenecannot be explained solely by orbital forcing.

On page 411 of this issue, Renssen andcolleagues show that even while decaying,the Laurentide ice sheet kept large parts ofNorth America and western Europe coolthroughout the early Holocene, despite therising incoming solar radiation in mid andhigh northern latitudes during summer3.

Te transition from the cold climate of thelast glacial maximum to the warm Holoceneinterglacial began about 15,000 years ago,stabilizing at warmer interglacial conditionsabout 11,700 years ago. Tis warmingwas associated with rising incoming solarradiation (insolation) in the NorthernHemisphere summer months, which peaked

between 11,000 and 9,000 years ago. Itled to the complete disappearance of theFennoscandian ice sheet, which covered

much of Europe, by about 9,000 yearsago. But on the other side of the AtlanticOcean, despite the warming climate, the lastremains of the Laurentide ice sheet coveredmuch of northeastern Canada until about7,000 years ago3(Fig. 1).

Proxy-based reconstructions oftemperatures throughout the NorthernHemisphere show that temperature evolutionduring the early Holocene was not uniform47.Maximum temperatures occurred slightlymore than 9,000 years ago in Alaska andnorthwest Canada, in accordance with theorbital forcing. emperatures in northeasternCanada, Greenland and Europe, however,peaked between 8,000 and 5,000 years ago,and it had been suggested that the delays inwarming were caused by the remains of theLaurentide ice sheet8.

Renssen and colleagues3test this

hypothesis through a set of climatesimulations tailored to identify how theLaurentide ice sheet affected early Holocenetemperatures. Tey use an Earth systemmodel of intermediate complexity (EMIC)that simulates atmosphere, ocean andvegetation processes. Compared withthe more complex general circulationmodels (GCMs), EMICs use lower spatialresolution and simplified dynamics forthe atmosphere and ocean, which makesthem computationally fast and thus ideallysuited for the relatively long simulationsneeded in palaeoclimatology. Te teamusedproxy-based temperature reconstructions

PALAEOCLIMATE

Delayed Holocene warmingRemnants of the Laurentide ice sheet lasted until about 7,000 years ago. Climate simulations show that theycaused the multimillennial delay between maximum early Holocene solar radiation and temperatures evident in

Northern Hemisphere proxy records.

Martin Widmann

Laurentide

ice sheet

Greenland

ice sheet

Figure 1 |The extent of the Laurentide and

Greenland ice sheets about 9,000 years ago.

At the last glacial maximum, the Laurentide ice

sheet covered most of Canada and the northern

United States (its extent is shown by the blue

line). By 9,000 years ago, the ice sheet was limited

to northeastern Canada. Model simulations

from Renssen and colleagues3show that these

remnants delayed Holocene warming across

eastern North America, the Labrador Sea and

western Europe.



15/80


news & views

rom several locations in North America,Greenland and Europe to assess whether theEMIC can realistically simulate the climaticeffect o the early Holocene ice cover, despitethe considerable simplifications.

Renssen and colleagues perormed severalsimulations designed to investigate the effects

that the decaying Laurentide ice sheet wouldhave on climate across North America andEurope. One model run prescribed thedecline o the ice sheet as reconstructedrom data collected rom its ormer margins.Another simulation also included themeltwater rom the ice sheet entering theLabrador Sea. In addition, the effects o orbitaland greenhouse-gas orcing were simulated.

Te simulations showed that, althoughorbital orcing was the dominant controlon the evolution o average NorthernHemisphere temperatures, the ice sheetstrongly influenced temperature patterns.

Locally, the higher albedo o the ice sheetkept air temperatures low, and the cold airwas advected downwind. But the most ar-reaching effects arose rom the meltwaterflowing into the Labrador Sea. Tisreshwater influx promoted sea-ice ormationand inhibited the deep convective circulationound today, and its attendant northward heattransport. ogether these processes delayedHolocene warming throughout easternNorth America and western Europe, in aregional pattern broadly consistent with theproxy records.

Although cooling effects rom the

Laurentide ice sheet in the early Holocenehave been suggested previously, the study

is highly relevant as it shows that i thehistorical evolution o the Laurentide icesheet is prescribed, a climate model is ableto capture many aspects o the temperaturepatterns during this period and thatprocesses other than local cooling areinvolved. However, the agreement with the

empirical temperature reconstructions is notperect. Potential reasons or discrepanciesinclude uncertainties in the proxy-basedreconstructions and biases in the simulatedatmospheric and ocean circulation, whichwould in turn affect the models propagationo temperature and circulation anomaliesaway rom the ice sheet. GCMs couldprovide an avenue to assessthe influences omodel biases inherent to EMIC studies, assimulations over several thousand years arenow easible9.

Simulations using prescribed ice sheets,such as those o Renssen and colleagues,

are necessary to isolate the influence othe cryosphere on climate, and provideprocess understanding that cannot directlybe obtained rom coupled simulations o theatmosphereoceancryosphere system. Butonly coupled simulations provide the meansto estimate millennial and longer variabilityin both past and uture climates rom theknowledge o external orcing actors,such as orbital parameters or greenhouse-gas concentrations.

EMICs coupled to ice-sheet models havebeen used to simulate past climates10, whereasGCMs or the past have only been used to

drive ice-sheet models without the changingice sheets eeding back on the climate11.

Simulations o uture climate have beenconducted with ice-sheet models coupled toboth model types12,13, but the dynamic ice-sheet model component, and its orcing bycoarsely resolved climate data, is a challenge,and improvements are needed to increasethe reliability o the simulations14. A realistic

representation o ice sheets in climate modelsis not only crucial or estimating sea-levelchange, but, as demonstrated by Renssen andcolleagues3, also or an accurate simulation oregional temperature changes.

Martin Widmann is at the School of Geography,

Earth and Environmental Sciences, University of

Birmingham, Birmingham B15 2, UK.


References1. Jansen, E. et al. in IPCC Climate Change 2007: Te Physical

Science Basis(eds Solomon, S. et al.) 433498

(Cambridge Univ. Press, 2007).

2. Claussen, M., Berger, A. & Held, H. in Te Climate of PastInterglacials (eds Sirocko, F., Claussen, M., Sanchez Goni, M. F.

& Litt, .) 2935 (Elsevier, 2007).

3. Renssen, H. et al.Nature Geosci. 2, 411414 (2009).

4. MacDonald, G. M. et al. Quat. Res. 53,302311 (2000).

5. Davis, B. A. S. et al.Quat. Sci. Rev. 22,17011716 (2003).

6. Kaumann, D. S. et al.Quat. Sci. Rev. 23,529560 (2004).

7. Jansen, E. et al.in Natural Climate Variability and Global

Warming: A Holocene Perspective(eds Battarbee, R. W. &

Binney, H. A.) 123137 (Wiley-Blackwell, 2008)

8. Kutzbach, J. E. & Webb, . III in Global Climates Since the Last

Glacial Maximum(eds Wright, H. E. Jr et al.)

511 (Univ. Minnesota Press, 1993).

9. Wanner, H. et al.Quat. Sci. Rev.27,17911828 (2008).

10. Calov, R., Ganopolski, A., Claussen, M., Petoukov, V. & Greve, R.

Clim. Dynam. 24,545561 (2005).

11. Otto-Bliesner, B. et al.Science311,17511753 (2006).

12. Driesschaert, E. et al.Geophys. Res. Lett.34,L10707 (2007).

13. Vizcano, M. et al.Clim. Dynam. 31,665690 (2008).14. Alley, R. et al.Science310,456460 (2005).

Most travellers crossing the vastand apparently empty expanseo the central Nevada desert will

have little inkling o the puzzlement andcontroversy that the region has causedamong geophysicists or decades. Not onlyhas Nevada somehow managed to extendand triple its surace area without thinningits crust, but it also boasts an enigmatictopography and gravity structure1. On topo that, seismic shear waves in most o thewestern US show considerable splitting, butthis is not the case or central Nevada. Onpage 439 o this issue, West and colleagues

characterize a cylindrical drip o lithospherethat is sinking beneath Nevada2. Teresulting vertical flow in the uppermostmantle differs rom the surrounding largelyhorizontal motion, which may help explainthe distinct shear-wave splitting pattern.

Away rom the boundaries o tectonicplates, the mantles uppermost cool anddense layer the lithosphere generallysits more or less inertly atop the convectingasthenospheric mantle. Under someconditions though, the lower portions o thelithosphere can become unstable and sink ordrip off3. Sinking lithosphere can be detected

by seismic tomography, which makes useo the act that seismic waves tend to travelaster through cooler and denser material.

Another, more direct seismic method todetect mantle flow is the measurement oshear-wave splitting. When a portion o themantle flows, it develops a abric owing tothe preerred alignment o its constituentminerals. Tis causes shear waves passingthrough it to split into two perpendicularpulses, each o which travels with itsown velocity. Te most robust splittingmeasurements are perormed on shear wavesthat travel almost vertically through the

TECTONICS

Draining NevadaThe lack of strong splitting of seismic shear waves below central Nevada is in marked contrast to the surrounding

region. Seismic data and numerical experiments suggest that a skinny, cylindrical drip of lithosphere may be to blame.

Vera Schulte-Pelkum



16/80


news & views

mantle. Such waves show strong splitting orhorizontal mantle flow, but not or verticalflow (Fig. 1).

In most parts o the western US,substantial shear-wave splitting is observed.Tis pattern has generally been consideredto arise because o the horizontal motiono the North American plate relative to theunderlying mantle. A notable exception is aregion o reduced splitting in Nevada, whichhas been interpreted as upwelling mantle4or as toroidal flow around a sinking slab ooceanic lithosphere5.

West and colleagues use tomographicas well as shear-wave splitting techniquesto analyse a new seismic data set or thewestern US and to explore the mantle below

central Nevada

2

. Teir tomographic imagereveals a cylindrical body o cold lithosphericmaterial through which seismic wavestravel aster than the surrounding warmasthenospheric mantle. Te body extendsrom about 75 km below the surace to atleast 500 km below the surace, well into theasthenospheric mantle.

Tis high-velocity cylinder underliesthe region where there is little or no shear-wave splitting. Tis suggests that mantleflow beneath this region is vertical becauseo parts o the lithosphere dripping intothe underlying asthenosphere. Te steepnortheasterly tilt o the drip seen in the

tomographic image is probably due to a pushimparted by asthenospheric flow relative tothe continent. Te direction o asthenospheric

flow in the region has been a subject o debatein the past, and the geometry o this dripprovides a new constraint.

West and colleagues find that numericalexperiments using parameters typical ormodelling lithospheric drips are able tomatch the observed geometry. Te resultssuggest that a density increase o only about1% along with a local temperature increaseo about 10% could have initiated sinking.Increased density is ofen related to coolertemperatures; however, accumulation omaterial lef in place afer melt extraction,or example, could contribute the warm

yet dense material that triggers the drip.Te researchers find that the developmento a lithospheric drip in this setting couldhave taken between one million years andtens o millions o years, depending on theboundary conditions.

Detachment o dense lithospheric mantlerom the overlying crust causes the latterto bob higher, causing uplif at the Earthssurace. Such uplif has been observed inmany other regions o the world, but it isnoteworthy that there is no evidence or it incentral Nevada. Te regions unusually thinlithosphere may cause insignificant uplifrom detachment, or a decoupling layer in

the crust or mantle may prevent uplif byallowing lateral movement.

Te lithospheric drip inerred byWest and colleagues provides a unifiedexplanation or the ast velocities revealed bytomographic analysis and the reduction inshear-wave splitting below central Nevada.

Te drip was not observed in anothertomographic model using the same dataset, with nominally sufficient resolution6.Interestingly, however, a study that usedsurace waves instead o the body waves usedby West and colleagues, recorded at the samestations7, does hint at the existence o thiseature. Tere is an art to tomography, andthe picture in the western US is likely to berefined urther in the near uture.

Te shear-wave splitting pattern inthis region was known previously andwas explained in other ways4,5.West andcolleagues discuss some o the problems with

these previous interpretations, but their ownhypothesis o a lithospheric drip does notexplain some eatures o the pattern either.Tis includes, or example, the wide, bow-wave-shaped pattern o ast orientations thatsurrounds the region o very low splitting,which the previous studies managed toaddress to some extent.

In addition, an intriguing contrast ingravity and topography coinciding withthe southern edge o the proposed drip hasgenerally been interpreted as evidence orhot and buoyant upwelling mantle undercentral and northern Nevada8, including

the region o the proposed drip. A cold,dense downwelling as proposed by Westand colleagues is thus the opposite o whatis needed to explain gravity and suracetopography observations, and increasesthe conundrum rather than putting it torest. It seems that Nevada has not given upall o its secrets quite yet, but the resultspresented by West and colleagues2suggestthat we may be slowly on our way tounderstanding the tectonic evolution o thisenigmatic region.

Vera Schulte-Pelkum is at the Cooperative Institute

for Research in Environmental Sciences and at the

Department of Geological Sciences, University of

Colorado at Boulder, 399 UCB, Boulder, Colorado

80309-0399, USA.


References1. Jones, C. H. et al.ectonophysics213,5796 (1992).

2. West, J. D., Fouch, M. J., Roth, J. B. & Elkins-anton, L. .

Nature Geosci.2,439444 (2009).

3. Houseman, G. A. & Molnar, P. Geophys. J. Int. 128,125150 (1997).

4. Savage, M. K. & Sheehan, A. F.J. Geophys. Res.

105,1371513734 (2000).

5. Zandt, G. & Humphreys, E. Geology36,295298 (2008).

6. Burdick, S. et al.Seismol. Res. Lett. 79,384392 (2008).

7. Yang, Y. & Ritzwoller, M. H. Geophys. Res. Lett. 35,L04308 (2008).

8. Saltus, R. W. & Tompson, G. A. ectonics14,12351244 (1995).

Drip

Crust

Lithosphericmantle

Asthenosphericmantle

N

Seismic station

Figure 1 |Mantle flow and shear-wave splitting. In an area with horizontal mantle flow, a vertically

incident shear wave splits into a fast (red) and an orthogonal slow (blue) pulse. When dense lithosphere

becomes unstable, it can drip off. In the presence of such downwelling material, a vertically incident shear

wave (green) travels with a single velocity and little splitting occurs. The drip will sink vertically but the

horizontal asthenospheric flow in the region will cause its base to tilt in the direction of flow. West and

colleagues suggest that such a lithospheric drip is responsible for the enigmatic seismic observations

below central Nevada2.



17/80


news & views

Today, over 100 million peopleworldwide rely on groundwaterthat is contaminated with arsenic1.

Exposure is greatest in southern Asia, wherearsenic trapped in buried sediments leachesinto groundwaters that support heavilypopulated regions. Tere are now growingconcerns that human manipulation ogroundwater supplies may be exacerbating

the problem. At the American GeophysicalUnion Chapman Conerence on Arsenic inGroundwaters o Southern Asia2, held thisMarch in Siem Reap, Cambodia, participantsdiscussed the possibility that groundwaterextraction may be increasing arsenic levelsin shallow aquiers3(C. Harvey, MI, USA)and aiding the movement o arsenic todeeper, hitherto uncontaminated layers(W. Burgess, UCL, UK).

Arsenic is a naturally occurring elementin the Earths crust, and many soils andsediments contain small amounts o thetoxin. I ingested over long periods o time,

even low doses o arsenic can cause cancerand other diseases: a lietime exposure to just50 g l1o arsenic in drinking water may killone in a hundred people prematurely, andjust a ew years o childhood exposure willlead to increased lung cancer risks as an adult(A. Smith, UC Berkeley, USA). Indeed, thereare growing concerns about environmentalarsenic exposure or some groups in Europeand the US (res 5, 6).Te acuteness o theproblem in southern Asia groundwater-induced poisoning o villagers in Bangladeshhas been described as the largest poisoningo a population in history4 relates to

the environmental conditions there, whichare thought to avour arsenic releaserom sediments.

Te mechanisms that govern arsenicrelease into groundwater are only juststarting to be understood. Arsenic buriedbeneath the surace is thought to be largelybound up with iron and sulphur minerals(B. Bostick, Dartmouth College, USA;G. Breit, USGS, USA; F. Larsen, GEUS,Denmark). Subsequent arsenic release romiron-bearing minerals relies on desorptiono arsenic rom mineral suraces (J. Hering,EAWAG, Switzerland; Y. Zheng, QueensCollege, City Univ. NY, USA) and microbial

reduction under anaerobic conditions7.Genetic profiling o arsenic hotspots andcarbon isotope labelling experiments point toanaerobic iron-reducing and arsenic-reducingbacteria as the main culprits (J. Lloyd, Univ.Manchester, UK). Organic matter stimulatesthe activity o these bacteria, with the degreeo stimulation dependent on the type oorganic matter supplied (D. Polya, Univ.

Manchester, UK working with B. Van Dongenand R. Pancost, Univ. Bristol, UK; D. Postma,echnical Univ. Copenhagen, Denmark).

Some have speculated3that the massiverise in groundwater use in southern Asiain recent years could be increasing theamount o arsenic entering groundwatersupplies (C. Harvey, MI, USA). Shallowtube wells, essentially tubes that bore intounderground aquiers (Fig. 1), are designedto draw water up rom depths o typically2040 m, exactly the region where arsenic-rich groundwaters are ofen ound. Teextraction o large volumes o water or

irrigation rom the shallow aquiers drawssurace water that is rich in highly reactive

organic compounds below ground. Tesecompounds uel the microbes implicatedin the reduction o arsenic-bearing ironminerals, and the reduction o arsenic itsel.Both processes result in arsenic transer romsediments to groundwater.

Tis theory o human-induced arsenicintensification is highly plausible. Terequired mechanisms are well known, and

increases in arsenic concentration withgroundwater age o 1030 g l1per yearhave been observed in Bangladesh (M. Stute,SUNY, USA). Tere are, however, nopublished time series data o sufficient length(over 30 years) to demonstrate unequivocallywhether this process is taking place at a highenough rate to be a concern to policymakersover the next 50 to 100 years. Resolvingthis should be a ocus or uture research;this should include the study o aquiers,or example in Vietnam8and Cambodia9,that are at present relatively unaffected byhuman activities (S. Ferndor, M. Polizzotto,

Stanord Univ., USA; S. Benner, Boise StateUniv., USA).

ENVIRONMENTAL SCIENCE

Rising arsenic risk?Millions of people in southern Asia rely on arsenic-contaminated groundwater to live. Massive water withdrawals

through wells may be increasing the problem by drawing arsenic-mobilizing substances into shallow aquifers and

arsenic-contaminated shallow groundwaters into deeper aquifers.

David Polya and Laurent Charlet

Figure 1 |Tube well used to supply groundwater to villagers. At the Chapman conference 2, participants

discussed the possibility that human manipulation of groundwater supplies is increasing groundwater

arsenic levels. Specifically, there is growing concern that tube wells are drawing arsenic-mobilizing

substances into shallow groundwater supplies, and deep tube wells are drawing arsenic-rich waters

to depth.

THOMASROSENBERG



18/80


news & views

Long before the reign of complex

animals, the Earths crust stretched

and thinned, creating a marine basin in

what is now central India. The Vindhyanbasin was flooded with sea water, and

subsequently filled with sediments.

But it is unclear just when these events

happened. Radiometric ages from

zircons and diamonds, as well as the

orientation of magnetic grains, suggest

that the basin began to fill between

1.7 and 1 billion years ago. However,

purported fossils hidden in the lower

rock layers hint at the presence of

simple animals that are consistent with a

much later origin for the layers in the

Ediacaran or Cambrian periods (about

640 to 500 million years ago).

The ensuing debate has seen some

researchers question whether these

fossils are actually of biological origin,

whereas other groups have suggested that

the radiometric ages reflect the source of

the sediments, rather than the creation of

the basins lower layers. Stefan Bengtson

of the Swedish Museum of Natural

History and his colleagues have jumped

into the fray, with new fossil specimens

and new ages for the surrounding rocks

(Proc. Natl Acad. Sci. USAdoi: 10.1073/

pnas.0812460106; 2009).

Aging well

PALAEONTOLOGY

The miniscule fossils are mostly

contained within phosphorous-rich nodes.

The nodes revealed the remains of bacterial

colonies, which generally show up as small

clusters of filaments that form distinctive

shapes. The group also found segmented

tubes, less than 200 m in diameter,which could be the remnants of algae. But

they failed to find the most contentious

microfossils from previous studies: embryos

from primitive multicellular animal life.

Instead, they suggest that the tiny spherical

shapes are air bubbles produced by

bacterial activity, trapped in the sticky film

covering the microbial mats. Overall, they

contend that the fossil assemblage probably

represents a pre-Ediacaran ecosystem.

Direct dating of the phosphate-rich

material surrounding the fossils using lead

isotopes minimized potential sources of

contamination. Their analyses converged on

an age of roughly 1.65 billion years well

in line with previous mineral-based

estimates leading the group to conclude

that the rocks, and the fossils within, are

indeed from the Palaeoproterozoic era.

Although the rocks no longer seem

to reveal

early animals, Bengston and

colleagues are quick to point out that

the exceptional preservation of the

fossils provides a unique window into

life 1.6 billion years ago. Interpretation

of the segmented tubes as algae would

push back the earliest appearance

of multicellular eukaryotes by up to

600 million years, and the bacterial clumps

show a rare glimpse of pre-Cambrian

calcifying cyanobacteria proving that

there is still much to learn from these

controversial rocks.

ALICIA NEWTON

Given that shallow tube wells tap intoarsenic-laden sedimentary layers andthe concentrations o this toxin in watermay increase, it seems sensible to lookelsewhere or clean water supplies. Teheterogeneous nature o arsenic distributionin shallow groundwaters means that

well-switching 10 many villages lie close tolow-arsenic wells is a useul short-termremediation strategy (A. Van Geen, LDEO-Columbia Univ., USA). Furthermore, the olderand deeper aquiers, made up o Pleistocenedeposits (laid down more than 10,000 yearsago) are ofen low in arsenic, possibly becausethey contain a high proportion o mineralsuraces able to sequester the poison. Tusdeep tube wells, which tap into groundwaterwell below the arsenic-affected layers, are apotential source o arsenic-ree water.

But as highlighted repeatedly at themeeting, even water in these deep aquiers

may not remain arsenic-ree. Tere aregrowing concerns that extensive deepwater extraction will draw arsenic-richwater rom the top o the aquier to de

nature.geoscience

Documents