Annual Meeting 2009

Program

Lamont Doherty Earth Observatory of Columbia University
July 8-10, 2009

Opening Remarks: Understanding the past, a test for predictions of the future – Wally Broecker

The Changing Cryosphere
Neogene perspectives on the stability of the West Antarctic ice sheet – Rob DeConto
Understanding the causes of rapid change in ice discharge from the Greenland ice sheet – Mark Fahnestock
Observing and modeling Arctic sea ice loss – Bruno Tremblay

Holocene Climate Change
Holocene Climate Change- Overview – George Denton
Timing of Holocene mountain glaciers events and relation to interhemispheric patterns of climate change – Joerg Schaefer
Reconstructing Holocene Atlantic Ocean climate change – Peter DeMenocal
Reconstructing global climate fields during the Common Era: Motivations and challenges – Jason Smerdon
Tree ring reconstructions of Asian monsoon variations over the last several centuries – Brendan Buckley
Deglacial lake level history in the Mono Basin, CA – Sidney Hemming
Solar variability and climate change over the Holocene – David Rind
Atlantic Ocean-North Africa-Middle East climate linkages through the Holocene – Yochanan Kushnir

Climate Change and the Carbon Cycle
Reorganization of global winds associated with abrupt climate change and its impact on atmospheric CO2 – Bob Anderson
Changing carbon uptake by ocean and terrestrial biosphere – Jorge Sarmiento

Climate Change and Tropical Storms
How can we predict future changes in tropical storm frequency and intensity? – Suzana Camargo
Hurricane Track Variability and Secular Potential Intensity Trends – Jim Kossin
High resolution modeling of Atlantic hurricanes and climate change – Tom Knutson
Geological records of past tropical cyclone activity – Amy Frappier

Past, Present and Future Hydroclimate Change
Global patterns of paleo dust transport and deposition – Gisela Winckler and David McGee
Late Pleistocene and Holocene hydroclimate change – Wally Broecker
Climate-dust interactions in a warming climate – Ben Cook
Is anthropogenic subtropical drying already occurring? – Richard Seager
Projections of African climate change – Michaela Biasutti

The Role of Ocean Circulation in Climate Change
The changing high latitude ocean circulation and interactions with the cryosphere – Peter Schlosser
Holocene changes in North Atlantic Ocean circulation – David Lund

Social Impacts of Climate Change
Climate and disease: an on the ground perspective- Judy Omumbo
Managing transborder water resources under environmental and economic uncertainty – Tobias Siegfried
Climate and societal tipping points – socio-economy of India and Colorado river water management – Balaji Rajagopalan
Climate change and agricultural productivity – David Battisti
Managing emissions from fossil energy resources – Klaus Lackner

Presentation Summaries (some presentations available for download)

Understanding the past, a test for predictions of the future – Wally Broecker

The meeting kicked off with a talk by Wally Broecker (LDEO):  ‘Understanding the past, a test for predictions of the future’.  Wally emphasized the power of combining paleo records and climate models to understand climate change.  Climate models have frequently found it hard to replicate events in the past and, in their attempts to do so, climate modelers have often learned the importance of processes that they had hitherto neglected or which their models simulated poorly.

Wally provided three examples (out of many possibilities).

Broecker concluded by emphasizing NOAA’s responsibility to fund paleo research because paleo records and the most powerful tools with which to test climate model’s predictions of the consequences of the build up of greenhouse gases.

The Changing Cryosphere

Neogene perspectives on the stability of the West Antarctic ice sheet – Rob DeConto

Rob DeConto (U. Mass) ocused on the past and future of the Antarctic ice sheet.  The primary concern is the West Antarctic ice sheet (WAIS) because it terminates below sea level and is buttressed by floating ice sheets that are pinned to subsurface topography.  Melting ice shelves could reduce back-pressure on the WAIS and allow increased ice loss to the ocean.  Since the WAIS bed slopes down into the continent, and ice flow through the grounding line is proportional to the depth of the ice there, this presents the possibility of runaway retreat although the timescale for this (centuries, thousands of years) is not known.  Recently the ANDRILL drilling program has uncovered evidence of open ocean at the current location of the WAIS as recently as one million years ago.  Open water occurred frequently in the Pliocene.  DeConto uses an ice sheet model matched to a floating ice shelf model to simulate this system.  He used some idealized representation of past climates (temperature, precipitation and ocean melt rates).  The model shows three configurations largely controlled by the ocean melt rates  (1) a ‘super-interglacial’ when melt rates are too large to sustain floating ice sheets and open ocean appears, (2) a ‘glacial’ state of an expanded WAIS, and (3)  a ‘modern state’ that is inbetween and akin to today.  The model simulates collapses of the WAIS with the whole WAIS collapsing at once and providing a 7m sea level rise.  However, proxy records suggest that at stage 31 sea level rise was more like 20m and the model cannot produce such large ice loss, regardless of changes in the Earth’s orbit, carbon dioxide, or sea level.  Resolving this requires more work.  For the future, imposed increasing melt rates suggest an e-folding time for WAIS collapse of just over a millennium with the ice loss only partially compensated for by increased ice mass from increased precipitation over the ice sheet. [download]

Understanding the causes of rapid change in ice discharge from the Greenland ice sheet – Mark Fahnestock

Moving from models to data and from Antarctica to Greenland, Mark Fahnestock (U. New Hampshire) presented the latest observations of Greenland ice loss.  Due to acceleration of outlet glaciers in recent years to decades exceeding any surface mass balance gain, the Greenland ice sheet (GIS) is probably currently well into negative mass balance.  For example the Jakobshaven glacier has accelerated from 7km/yr in the early 1990s to 14 km/yr now.  This is related to about 3-4K ocean warming around Greenland that appears as a mid 1990s jump coinciding also with N. Atlantic sea ice loss.  The details of the link are not well understood but the warming has probably caused increased basal melt and/or modulated calving with complex and potentially nonlinear ice flow dynamics then coming into play.

Observing and modeling Arctic sea ice loss – Bruno Tremblay

Bruno Tremblay (McGill U.) followed up with a discussion of Arctic sea ice loss.  Arctic sea ice cover has been steadily declining since satellite data began in the late 1970s and its loss has accelerated in the last few years. This raised the question of whether it will recover.  In 2007 it is thought that winds drove ice across pole towards Greenland exposing open ocean which was then heated by the sun and triggering a positive ice-albedo feedback.  In 2008, although there was a modest increase in sea ice cover, ice age was younger because of loss of multiyear ice and it is possible that ice volume was less than in the previous year.  Climate model simulations also show such transitions with steady thinning until a warm year causes ice loss and onset of positive ice-albedo feedback.  It is noteworthy that the real world has seen subsurface ocean warming but, to date, this heat has not mixed up into the layer below the sea ice which questions the model mechanism for rapid sea ice loss.  Further, climate model simulations also show that years of rapid sea ice loss such as 2007 can occur as a result of natural variability but that afterwards the sea ice recovers.  It is not yet clear if the real world will undergo such a recovery or if the transition to less summer sea ice will continue.

Holocene Climate Change

Holocene Climate Change- Overview – George Denton

George Denton (U. Maine) used glacial records to explore the global history of the Holocene.  Throughout the Holocene there are millennial scale events such as the Roman optimum, the Dark Ages cool period, the Medieval warm period and the Little Ice Age (LIA).  The LIA advance was the largest advance of the Holocene in northern hemisphere glacial records.  Indeed, collapsing plateau glaciers in Greenland are revealing vegetation such as dwarf birches that date to the Medieval period.  This matches records of North Atlantic sea ice, the Alps glacier record and marine sediments north of Iceland in showing the Medieval to LIA circum-Atlantic cooling.  During the 20th Century, Alpine glacier advances and retreats show a relation to the AMO and, hence, it is hypothesized that the millennial scale changes are also related to changes in North Atlantic Ocean heat transport.  The LIA maximum of glaciers is also found in North America, Iceland and Europe. However, in the southern hemisphere glaciers have in general been retreating through the Holocene.  Part of the difference between the hemispheres is explained by the precession cycle and its influence on summer insolation.  Antarctica reveals some interesting Holocene records with glaciers on the peninsula retreating and revealing organic materials and Holocene retreat in the Ross Sea area uncovering remains of elephant seals, who don’t live there now,  dating to about 6000 to 500 years B.P. These remains suggest that throughout the late Holocene ice cover was much less than now.

Timing of Holocene mountain glaciers events and relation to interhemispheric patterns of climate change – Joerg Schaefer

Joerg Schaefer (LDEO) continued with a more detailed analysis of New Zealand glacial records based on surface exposure dating showing high-frequency glacier advances throughout the Holocene and hence stark differences to the classic northern climate pattern.  In New Zealand the glacial maximum occurred about 13K years BP and glaciers retreated through the Younger Dryas (unlike in the northern hemisphere) after which a succession of moraines were laid down until 6k BP after which there were no more until the last millennium.  New Zealand glacier advances became smaller during the Holocene in contrast to the northern hemisphere.  It was suggested that New Zealand glaciers respond to multi-decadal changes of the Indo-Pacific Oceans.  A first precisely dated tropical glacier record from southern Peru indicates that an earliest Holocene and a LIA advances dominated the glacier advances, implying climate links of at least this area of the tropics to the north Atlantic. An effort to enable generation of near global maps of glacier fluctuations is ongoing. [download]

Reconstructing Holocene Atlantic Ocean climate change – Peter DeMenocal

Peter DeMenocal (LDEO) presented some new results on reconstructing Holocene SSTs. This is highly important as ocean changes are typically invoked as explanations for Holocene climate changes over land.  SST records require high sedimentation rates, a reliable calibrated SST proxy and excellent dating.  deMenocal described how he has been examining cores from high sedimentation rates and analyzing both Mg/Ca and oxygen isotopes from forams.  Both of these are impacted by temperature and salinity and a multivariate regression technique he has developed allows reconstruction of both temperature (T) and salinity (S) from the sediment cores.  This methodology replicated observed T and S variations along the mid Atlantic ridge.  deMenocal applied this to a core from the Carolina slope.  This showed warm SSTs there during times of minimum solar irradiance (e.g. the LIA) and colder conditions during the Medieval period which needs some explanation in terms of climate dynamics. [download]

Reconstructing global climate fields during the Common Era: Motivations and challenges – Jason Smerdon

Jason Smerdon (LDEO) pointed out some of the ample evidence of climate impacts on human society over the last millennium, including the Norse in Greenland, the Anasazi and the Maya.  Further the last millennium offers the best chance for developing annually resolved records of climate with good spatial coverage for the pre-instrumental era.  In terms of reconstructions of global temperature variations there is currently a lot of uncertainty with higher amplitude reconstructed variations implying larger climate sensitivities.  Constraining the real sensitivity requires determining past changes with greater accuracy.  Many methods have been used in recent years to reconstruct variations over past centuries including the use of pseudo proxies within climate models.  All methods suffer from variance loss back in time as records become fewer, and all tend to miss extreme events.  This emphasizes the need for more proxy data.  Further there is a need for more reconstructions of climate fields as opposed to climate indices. [download]

Tree ring reconstructions of Asian monsoon variations over the last several centuries – Brendan Buckley

Brendan Buckley (LDEO) provided the latest results from his work on using tree ring records to reconstruct the Southeast Asian monsoon over past centuries.  This is part of a larger effort to develop a drought atlas for Asia analogous to the North American Drought Atlas.  The talk focused on millennium-long records from conifers in Vietnam that provide good cross-dating – probably the best tropical tree ring records ever developed.  Ring width correlates best to March-April-May (MAM) Palmer Drought Severity Index (PDSI), which marks the transition from the winter dry season into the annual monsoon. A robust reconstruction of MAM PDSI shows great droughts in the late 14th and early 15th Centuries as well as during the mid to late 18th Century at a time of widespread collapse of southeast Asian kingdoms and when historical records frequently mention drought.  The earlier drought could well have been implicated in the collapse of Angkor Wat – ‘the hydraulic city’ – although other factors surely played a role, including mismanagement of water storage systems, damage to infrastructure etc.  Buckley showed how drought in Vietnam was linked to El Nino conditions, with an impressive anti-phasing of drought in Vietnam and drought in North America, emphasizing the global nature of climate variations. [download]

Deglacial lake level history in the Mono Basin, CA – Sidney Hemming

Sidney Hemming (LDEO) presented early results for a comparative study of the paleohydrology of Lake Lisan in Israel and lakes in the US Southwest. Lake Lisan (the large lake that formerly filled the basin where the Dead Sea is currently located) underwent large changes, with dry conditions during Heinrich Events and high levels (wet conditions) in the LGM. The level of  Mono Lake in California is known to have been much higher during the last glacial period (roughly between 70,000 and 10,000 years BP) from the presence of thick sediments, now exposed in the channel of Wilson Creek. Unfortunately, the radiocarbon ages are unreliable because the carbonate minerals contain reworked carbon. More recent deposits also shed evidence, including sediments deposited within the deepest part of the lake as well as carbonate minerals that formed along the shore line when the lake stood at its highest level. Carbonate mineral deposits called thinolite fans indicate very  cold conditions when at the highest level.  Thinolite forms by recrystalization of ikaite, a carbonate mineral that only forms under near freezing conditions. It is suggested that the highest lake level and shoreline, with the thinolite fans, occurred during the Younger Dryas period, about 12000 years ago. They aim to collect a core from the deep lake basin long enough to capture a volcanic ash layer dated on land to have been deposited about 11000 years ago, thus providing a good age marker for the lake sediments. [download]

Solar variability and climate change over the Holocene – David Rind

David Rind (GISS) reviewed the state of our understanding of forcing of climate variability and change by solar irradiance variability.  He began by noting that the relatively stable global mean temperature of the last decade is partly caused by reduced solar irradiance as part of the solar cycle and partly by a more La Nina-like phase in the tropical Pacific.  Nonetheless anthropogenic radiative forcing from rising GHGs greatly exceeds solar forcing and will propel further warming in the near future.  We only have observations of total solar irradiance (TSI) for the last few sunspot cycles and, therefore, there is huge uncertainty in determining what TSI was over the Holocene, including during the recent LIA.  Early, larger, estimates of TSI variability were based on observations of Sun-like stars in non-cycling states but these are not believed anymore.  Consequently the issue is whether during the Maunder Minimum stage of the LIA the TSI was less than that during the observed minimum of the sunspot cycle.  If not, the TSI for the Maunder Minimum was reduced by about 0.05% which is very small.  Although solar physics is a developing science, solar models do not support a large TSI reduction for the Maunder Minimum.  Solar models also call into question the existence of longer timescale solar cycles (e.g. Suess, Wolf-Gleissberg) In simulations of late Holocene climates, different groups have used different assumptions for TSI variability and hence the model results show great spread.  Several mechanisms have been proposed for how TSI variability can impact climate.  The first is simply based on the TSI change itself and how, for example, it could impact land-sea temperature contrasts.  More complex proposed mechanisms have involved an impact of TSI on tropical Pacific coupled atmosphere-ocean dynamics, on the Northern Annular Mode (NAM) and on North Atlantic Deep Water formation.  Also changes in UV radiation associated with TSI variability have been proposed to impact stratospheric ozone and, hence, temperatures with a feedback via wave dynamics onto tropospheric climate.  For example, just as warming shifts the climate to a more positive phase of the NAM, so can an increase in TSI.  Or, a solar maximum can cause increased stratospheric ozone and a warming that extends into the upper troposphere stabilizing the atmosphere with potential impacts on circulation and precipitation.  Hence there are many ways that solar variability could have impacted Holocene climate change but in the end it all comes down to just how large the variations in total or spectral irradiance were and that is not known. [download]

Atlantic Ocean-North Africa-Middle East climate linkages through the Holocene – Yochanan Kushnir

Yochanan Kushnir (LDEO) discussed impressive changes in climate over the Holocene in the Mediterranean region and North African.  This region is projected to dry substantially as a consequence of rising GHGs and hence it is imperative to understand the mechanisms that control climate variability and change in this region.  Over the Holocene, as northern summer insolation dropped and the North Atlantic Ocean cooled, the Sahel dried but the Dead Sea levels went up in an antiphased relation.  In modern records the Dead Sea level went up with local precipitation and, also, Jerusalem and Sahel precipitation were out of phase, suggesting similar mechanisms of North Africa-Middle East climate change operating on interannual and millennial timescales.  Jerusalem precipitation tends to be in phase with precipitation in western North America.  All this suggests an AMO influence – with a warm AMO causing less precipitation in Jerusalem and the American west and more in the Sahel.  The mechanisms for the AMO-Jerusalem precipitation link are proposed to involve pressure systems over the Atlantic (high during cold AMO phases) that cause anomalous temperature advection over the Mediterranean (cold advection for the cold AMO phase) and impact regional cyclogenesis. [download]

Climate Change and the Carbon Cycle

Reorganization of global winds associated with abrupt climate change and its impact on atmospheric CO2 – Bob Anderson

Bob Anderson (LDEO) noted that the correlation between carbon dioxide and temperature over the last 800K years remains unexplained.  Organic matter settles into the deep sea, is respired to carbon dioxide and comes back in contact with the atmosphere in regions of upwelling especially in the Southern Ocean.  A few years ago Robbie Toggweiler suggested that a northward shift of southern hemisphere westerlies during the LGM led to reduced ventilation of carbon dioxide to the atmosphere but until recently there has been no proxy for Southern Ocean upwelling.  Anderson has developed the opal flux as an upwelling proxy since the production of opal by diatoms can exceed the rate today only if the silica supply by upwelling increases.  The period of maximum opal burial occurs around 17-13 K BP.  In general, from the opal record, peak upwelling coincides with warming in Antarctica and a rise in carbon dioxide and a drop in 14C (from ice core records).  Anderson proposed a mechanism  driven from the northern hemisphere where extreme cold leads to iceberg discharge and increased sea ice cover which impacts global wind systems including a southward shift of the ITCZ and the southern hemisphere westerlies which in turn increase upwelling and atmospheric carbon dioxide. This same basic pattern, where cooling in Greenland was accompanied warming in Antarctica, increased upwelling in the Southern Ocean, and rising atmospheric CO2, occurred several times throughout the last glacial cycle, indicating that the sequence of processes is fundamental to Earth’s climate system.  Similar features are found in the paleoclimate record soon after the peak of the last interglacial period, indicating that large Northern Hemisphere ice sheets and low sea level are not prerequisite conditions for Northern Hemisphere forcing of processes that release CO2 from the deep sea. [download]

Changing carbon uptake by ocean and terrestrial biosphere – Jorge Sarmiento

Jorge Sarmiento (Princeton U.) continued with a carbon mystery.  So far, 58% of combined fuel and land emissions  remain in the atmosphere and 42% in the ocean.  The growth in atmospheric carbon dioxide is driven by fossil fuel emissions while the variability is controlled by ENSO.  The residual is  seen to shift around October 1989 with a drop then in the atmospheric growth rate.  Possible explanations for this include an over-reporting of emissions since then, a sudden increase in ocean uptake (not supported by models) or an increase in net land uptake from reforestation, fire suppression, climate change.  However there are no clear climate shifts in 1989 and so the cause of this carbon shift remains unknown.

Climate Change and Tropical Storms

How can we predict future changes in tropical storm frequency and intensity? – Suzana Camargo

Suzana Camargo (LDEO) reviewed what needs to be understood in order to project how climate change will impacts tropical cyclones (TCs).  She pointed out that TC formation and growth depend on tropospheric temperature and humidity, SST and the atmospheric circulation. Therefore, TCs are influenced by climate modes such as ENSO and the AMO.  However, given the lack of a comprehensive theory for TC cyclogenesis, there is no clear guidance for how TC frequency will change with climate change.  This has led to the use of empirical indices for tropical cyclonegenesis. In the case of intensity,  Emanuel’s Potential Intensity (PI)  theory addresses the maximum theoretical intensity a TC can reach under certain environmental condtions; it indicates higher PI in a warmer climate.  More recently changes in vertical wind shear has been investigated as another important factor.  Global climate models do not have TCs, only ‘TC-like’ storms, and generally have problems reproducing TC intensity because of inadequate spatial resolution.  Hence many modelers have used regional climate models (RCMs).  Others have used statistical-dynamical downscaling based on global models.  Many of these methods suggest that there could be a reduction in the number of Atlantic TCs under global warming but an increase in the number of the most intense storms. However uncertainties abound. [download]

Hurricane Track Variability and Secular Potential Intensity Trends – Jim Kossin

Jim Kossin (NOAA/NCDC) began with the observed strong relation between TC power dissipation index (PDI) and area averaged Atlantic SST for the past decades.   The potential intensity of storms is controlled by SST, tropopause temperature and convective available potential energy (CAPE)  About  8% of storms reach their PI which is another way of saying that about 8% of storms will be most impacted by a rise in the PI.  The calculated PI has increased since the 1960s for the Atlantic basin.  However when PI is computed along the actual storm tracks no trend is seen.  That is, over the late 20th Century, TCs have shifted to cooler SSTs and lower PI even as the basin mean SST warms and PI increases.  The lack of trend in PI suggests that observed trends in hurricane intensity cannot be attributed to PI trend.  There is no reason to believe, however, that this shift in TC tracks will continue to counteract the influence of basin warming. [download]

High resolution modeling of Atlantic hurricanes and climate change – Tom Knutson

Tom Knutson (GFDL) continued the discussion of hurricanes and climate change by pointing out the dangers of extrapolating empirical indices into the future.  For example the local power dissipation index (PDI)-SST relation for the Atlantic implies a 300% increase in PDI over the current century.  On the other hand the equally strong relation between PDI and Atlantic SST relative to mean tropical SST (which attempts to build in the influence of static stability) leads to a much smaller projected increase.  Similarly the observed increase in the number of Atlantic TCs occurs only in the open basin and could be accounted for by improved detection of TCs.  Indeed when changes in observing systems are accounted for there appears to be no significant trend in TC frequency.  GFDL uses an 18km grid Atlantic regional model that accurately accounts for hurricane counts since 1979 to assess the impact of  climate change on TCs.  Large-scale conditions from IPCC AR4 projections are then imposed on the regional model and it produces a reduction in TC numbers and a small increase in the number of the most intense storms.  But even an 18km grid cannot simulate Category 4 and 5 storms.  Consequently the operational GFDL hurricane prediction model was used to re-simulate the TCs in the regional model.  This again led to an increase over the current century in the number of the strongest storms. [download]

Geological records of past tropical cyclone activity – Amy Frappier

To finish the session on tropical cyclones Amy Frappier (Skidmore College) talked about using geological records to study TCs in the pre-instrumental period.  She began with a list of why TCs are important from replenishing aquifers, ending droughts, replenishing beaches, and through interactions with marine and terrestrial ecosystems.  Many efforts to study TCs in the pre-instrumental period involve coastal sediments and speleothems.  For example a record of outwash sands from the Gulf Coast extends for 4,500 years and reveals a hyperactive TC period from about 3,000 to 1,000 years ago but cannot distinguish whether this was caused by a shift in tracks or intensity.  Another record from Puerto Rico also shows a period of more TCs between 2,500 and 1,000 years ago which may be linked to changes in ENSO (according to the Rodbell et al. record from the Andes).  Hurricane rainfall can also be recorded as oxygen isotope departures in speleothems and Frappier is examining this in a record from Belize.  A multicentennial speleothem record from Australia shows a period around 1700 when storm incidence was several folds greater than in modern times.  Paleotempestology is a new field but these studies reveal the potential for this to yield long records of landfalling TCs to sample past hurricane-climate interaction outcomes, and to test theoretical and modeling predictions relevant to projecting the changing character of TC risk in a warming world. [download]

Past, Present and Future Hydroclimate Change

Global patterns of paleo dust transport and deposition – Gisela Winckler and David McGee

Gisela Winckler (LDEO) examined dust variations over glacial interglacial cycles and pointed out that ice core records indicate 25-50 times more dust during glacials than during interglacials.  However it is not known if the dust is a passive recorder of environmental conditions or an active player, via radiation, cloud physics and nutrient transport, in the glacial-interglacial cycles.  Three new core sites from the equatorial Pacific show maximum dust flux in glacials which is in phase with the rest of the globe.  The glacial-interglacial variations here are a factor of 7 or so.  Southern Ocean cores also show more dust during glacials and coherence  between equatorial Pacific and Antarctic core records even though  the dust sources are different.  Core top dust fluxes are greater in the west equatorial Pacific than in the east equatorial Pacific; this gradient is opposite in sign to that simulated by models indicating the need for improvement in modeling dust emission, transport and deposition. David McGee (LDEO) followed by suggesting that increased glacial dust fluxes were caused by increased gustiness because a very large proportion of dust emissions are caused by the strongest few per cent of wind speeds. This argues against the widely-cited hypothesis that aridity is the primary control of global dustiness and is consistent with the observation of high closed-basin lake levels near major mid-latitude dust source areas during the last glacial.

Late Pleistocene and Holocene hydroclimate change – Wally Broecker

Wally Broecker (LDEO) talked about lake level records from past climates,motivated by the need to test ‘Held’s hypothesis’ that in a warmer climate dry regions get drier and wet regions get wetter.  Support for the theory comes from glacial era lakes in western North America that were higher than now and from evidence of drier conditions during the Medieval period coinciding with evidence for warm conditions.  Broecker also showed that during the Last Glacial Maximum (LGM) Cari Laufguen in Argentina was a very large lake,  as were the Dead Sea, and lakes in China, all adding up to a picture of a moist subtropics during the glacial.  In tropical Africa, by contrast, Lake Victoria was dry.  Broecker has also been studying lake levels in the so-called Mystery Interval that followed the LGM and preceded the Bolling/Allerod. [download]

Climate-dust interactions in a warming climate – Ben Cook

Ben Cook (NOAA Climate and Global Change postdoctoral fellow) examined the role of dust as a forcer of regional climate change in both historical and future periods.  Dust sources can be distinguished as (1) geologically supported and long term (such as paleo-lakes), where emission is a function of atmospheric demand, or (2) short term, intensive and transient, as occurs from degraded or deforested land, where  emission is a function of supply and demand.  Primary long term sources are the dry subtropics (e.g. the Sahara and parts of the Middle East and Asia) with maximum dust loading in the atmosphere during summer months.  The Great Plains during the 1930s, when drought and crop failure following agricultural expansion caused massive soil erosion, are an example of transient dust sources.  Modeling experiments in which dust sources are imposed show that dust loading caused the 1930s Dust Bowl drought (initiated by La Nina SSTs) to intensify and move northward to become centered in the central Plains, whereas purely SST-forced droughts tend to be located in southwest North America.  The mechanism involves reflection of solar radiation by the dust, which thus reduces the energy available for ascent and condensation of water vapor.  Ice cores from off Antarctica show an increase in dust content over the 20th Century with the dust coming from South America.  Since precipitation has been increasing in South America during this period, it must be due to increased land degradation.  For the future, models that allow for carbon dioxide fertilization generally predict a drop in dust sources even as many subtropical areas of the world dry.  However the fertilization effect is questioned and, when the pressure on land from rising populations is accounted for, it is conceivable that dust emissions from rangelands and croplands could increase and initiate further climate-dust-land interactions. [download]

Projections of African climate change – Michaela Biasutti

Michaela Biasutti (LDEO) discussed the recent drying of the Sahel, one of the most notable decadal climate variations seen on the planet and one with enormous social impacts.  Changes in Sahel precipitation have been linked with variations in SST with an increased south to north SST gradient in the Atlantic and overall tropical ocean warming both being conducive to drought.  Coupled models forced by all natural and anthropogenic forcings over the 20th Century do produce a drying of the Sahel but only about a third of the amount observed.  For the 21st Century the Sahel precipitation trends are highly variable in a way that is yet to be explained.  A more robust prediction for the Sahel is that there will be a reduction in precipitation in the early part of the West African monsoon season and an increase in the latter part – a backward shift of the monsoon season but also, because the earlier shift is dominant, a reduction in the length of the rainy season.  Post 1979 precipitation data provides some evidence that this is already occurring. [download]

Is anthropogenic subtropical drying already occurring? – Richard Seager

Richard Seager (LDEO) addressed the question of whether anthropogenic subtropical drying is already occurring as IPCC models suggest it may be.  The mechanisms for subtropical drying in a warming world involve intensification of water vapor transport as the atmosphere holds more moisture, which makes dry regions drier and wet regions wetter, as well a poleward expansion of the Hadley Cell and poleward shift of the mid-latitude storm tracks, which expand the subtropical dry zones poleward.  These changes are robustly projected by IPCC models and impose drying on southwestern North America and the Mediterranean region.  It is easiest to look at planetary scale changes in precipitation using the satellite era which begins in 1979.  Over this period southwestern North America has dried  by an amount that is in excess of IPCC simulations for the same post 1979 period.  However the drying has coincided with a trend towards a La Nina-like state in the tropical Pacific.  The SST trend is argued to be the result of natural decadal variability of the tropical Pacific atmosphere-ocean system and its pattern corresponds quite well to the first EOF of 25 year trends of SST.  It is therefore argued that recent drying of southwestern North America is primarily associated with natural variability.  Indeed it was shown that similar trends in SST and precipitation occurred from the 1920s to the 1950s across a possible decadal regime change in the 1940s.  Models forced by historical SSTs reproduce well these precipitation trends.  This emphasizes the importance of decadal variability for current and near term future precipitation in the subtropical dry zones. [download]

The Role of Ocean Circulation in Climate Change

The changing high latitude ocean circulation and interactions with the cryosphere – Peter Schlosser

Peter Schlosser (LDEO) reviewed changes in recent decades in high latitude North Atlantic and Arctic Ocean climate.  The oceans have warmed to great depth since the 1980s with maximum warming near the surface but below the mixed layer.  There has also been a well known trend to reduced sea ice cover and increased air temperature over most of the Arctic area that has led to melting of permafrost. Arctic subsurface warming has followed from increased import of warm water from the Atlantic Ocean.  However this warm water is capped by the halocline at about 100m depth.  One question in regard to future change is whether loss of Arctic sea ice and exposure of open ocean will lead to an increase in wind mixing that breaks through the Arctic Ocean stratification to tap the anomalously warm water below, potentially leading to rapid further sea ice loss.

Holocene changes in North Atlantic Ocean circulation – David Lund

David Lund (U. Michigan) spoke about whether proxy data support the idea that changes in ocean circulation can explain past centennial to millennial climate changes (e.g. the LIA).  Currently the Gulf Stream transports 31Sv and 1.3PW of heat through the Florida Straits of which 25-40% is wind-driven and part of the subtropical gyre and the remainder due to meridional overturning.  Cores from either side of the Florida Straits, and analysis of oxygen isotopes in the sediments, allows an assessment of the cross-flow density gradient as a function of time.  Using an assumed level of no motion and the thermal wind equations the flow can then be estimated.  The density cross sections suggest the transport was 31.5 Sv for the modern period, 28 for the LIA and 30 for the Medieval period.  The temperature of the flow can also be estimated using standard techniques and suggests that it was warm in the LIA – warmer than in the Medieval period – and that LIA salinity was high too.  However, using the DeMenocal multivariate approach on oxygen isotopes and Mg/Ca ratios gives high salinity for the LIA but cool SSTs.  The weaker LIA flow with higher salinity may be caused by southward shifts in the wind stress pattern and a southward shift of the ITCZ as inferred from the Cariaco record. [download]

Social Impacts of Climate Change

Climate and disease: an on the ground perspective- Judy Omumbo

Judy Omumbo (IRI) discussed the application of climate information in the public health sector where health is a component of development with goals such as reduction of child mortality, improved maternal health etc. and an overarching goal of poverty and hunger eradication.  Recently Margaret Chan, Director General of the WMO, has stated that climate change is a serious problem that will impact health, water and food.  Climate change can impact health via direct routes such as heat waves, indirect routes via water, food etc. and also through causing social and economic disruption as can happen, say, during droughts.  It is estimated that the health impacts of climate change will disproportionately impact Africa with Asia running second.  Climate sensitive diseases include vector-borne diseases such as lyme, malaria, dengue and Rift Valley fever, water-borne diseases like cholera and airborne diseases like meningitis which has been linked to dust storms.  Malaria is a good example of a climate sensitive disease.  Formerly widespread into the mid-latitudes it persists in the tropics and subtropics because temperature is suitable year round for its transmission.  As such, climate change is expected to cause a modest expansion of areas of malaria. Much work is being done on how climate variability impacts the spatial patterns of disease risk, how risk varies by season and from year to year and to assess trends.  This work guides assessments of past disease interventions to attribute the results to the intervention or to climate variations.  The IRI has been developing databases that show maps of climate suitability for malaria transmission that can be used to assess the changing risk and also be used to retrospectively assess climate impacts on past malaria occurrence and how effective health interventions (e.g. spraying were).  For climate change, health professionals are most interested in the expected change out to a decade because the longer time horizon is outside the decision-making and policy environment.  In terms of needs there is a huge need for more health surveillance data as well as climate data and the need to develop a ‘climate-smart’ health sector. To that end the IRI hosts a summer institute that brings together health and climate scientists fostering lasting connection that will enable use of climate information in public health. [download]

Managing transborder water resources under environmental and economic uncertainty – Tobias Siegfried

Tobias Siegfried (Columbia Water Center) addressed the issue of climate and water resources.  Water problems around the world include decreases of low flows, changing seasonal patterns of snow melt, aquifer depletion, surface and groundwater pollution, salinization and drying of wetlands.  Climate change impacts on snow melt can for example reduce water supply in the summer when water is most in demand.  Further there are always conflicts in regard to demand on water pitting irrigation use (in summer) versus water use for power generation (needed in winter).  He discussed the need for combined economic and hydrologic modeling to optimize decision-making in such situations using an example from former Soviet republics of Central Asia. [download]

Climate and societal tipping points – socio-economy of India and Colorado river water management – Balaji Rajagopalan

Balaji Rajagopolan (U. Colorado) focused on the future of the Colorado River which is shared by six states and two countries.  The river flow is about 15 million acre feet (MAF) and the storage on the river, almost all in Lakes Mead and Powell, is about 60 MAF which allows a huge buffer and has ensured water deliveries even during the recent drought.  Colorado River water irrigates 3.5 m acres and supplies water to 30m people.  Almost all of the water in the river comes from precipitation in the headwaters and most of that from the highest elevations. In contrast, most of the river water is used in the lower basin.  Tree rings have been used to asses the river flow over past centuries and these reveal the early 20th Century, when the river compact was designed and signed, to have been a period of very high flow and that there have been very long periods of very low flow in the past.  Climate models do not have the resolution to represent the highlands where the precipitation occurs that sustain the Colorado River flow and, hence, there is considerable uncertainty as to how flow will change in the future although most estimates are that it will decrease.  Rajagopolan estimated that with a 10% reduction of flow there is a 20% chance of running out of water by 2050 and a 50% chance if the flow reduction is 20%.  However the water supply is ensured for the next decade or two, which provides time to plan for a future decline in river flow. [download]

Climate change and agricultural productivity – David Battisti

Turning to food production David Battisti (U. Washington) discussed the impact of climate change on tropical agriculture.  Already 940m people in the world are malnourished of whom 95% are in the tropics and subtropics depending on agriculture for both food and income.  The 1998-2002 drought in the Middle East and central Asia revealed the vulnerability of agriculture: in Iran 80% of livestock was lost and there was a 50% drop of grain yield in Tajikstan, as examples.  However scientists who develop seeds are familiar with precipitation variations and breeding for drought resistance.  Of more concern to them is temperature increase.  For example during the 2003 European heat wave temperature was 3.6 deg C above normal and Italy lost 36% of its maize yield and France lost 30%.  Battisti computed the chance that by the end of the century the mean summer temperature would exceed the warmest year on record.  This number reaches almost 100% in most of Africa, the western U.S. and Asia.  There is observational support for reduced crop yield as temperature increases so this warming could translate into 30-40% reductions in yield in India, Africa and the Middle East as well as reduced nutritional content per yield.  It is also possible that plants will become more vulnerable to disease as a result of warming.  These changes, combined with water stress, pests, rising sea level, etc. will lead to increases in the number of malnourished people in the world. [download]

Managing emissions from fossil energy resources – Klaus Lackner

Klaus Lackner (Earth Engineering Center of Columbia University) presented an important tool in solving the intensifying problem of CO2 climate forcing: carbon capture and storage (CCS). The world’s population has an increasing demand of energy that needs to be satisfied somehow. Klaus made clear that there is plenty of fossil energy left, including ~ 3000-5000 Gt coal, which could be used to produce synthetic fuels.  Three options for future, carbon neutral energy scenarios were presented: 1) solar energy, which is still too expensive; 2) nuclear energy, for which safety and nuclear waste problems remain unsolved; 3) fossil energy, but the CO2/global warming problem would get in the way unless the CO2 produced is captured and stored. Without CCS the use of fossil fuels is not compatible with CO2 stabilization.  Klaus presented filtering of CO2 directly out of the air as an efficient and cost-effective way of carbon capture (due to fast mixing of CO2 in the atmosphere, it does not matter where on earth to filter the CO2 out). Their team, with help from the CSEF, developed a proto-type filter unit reducing the CO2 concentration of air from e.g. 380 ppm to 280 ppm, based on a plastic CO2 absorbent, from which the CO2 can simply be washed off with water and the filter material can be reused after wash-off. The energy cost is about 50 kJ per mol of CO2. A carbon filter unit for one-ton of CO2 per day would fit in a 40 feet shipping container, i.e. is quasi-mobile and could be installed anywhere on earth, e.g. near water sources and/or CO2 storage locations.  If these instruments were to be set up on a large scale, billions of tons of CO2 would need to be stored. The most cited solution is underground storage, e.g. in oil-fields, but underground CO2 storage is problematic over the storage times required (1,000-10,000 year scale) and the volume to store remains challenging. As an alternative, Klaus promotes mineral sequestration and presented recent progress by adding citrate to the Mg-sequestration chemistry, which increased the efficiency dramatically. Finally, estimates to ‘close the carbon cycle’ run at about 1% of the GDP, the estimated 30$/ton CO2 would translate into a 25 cent increase per gallon fuel. [download]

Notes prepared by Richard Seager (LDEO), July 26, 2009