CLIVAR

CLIMATE VARIABILITY AND PREDICTABILITY

International CLIVAR Project Office
National Oceanography Centre
European Way
Southampton, SO14 3ZH, UK
Phone: +44-2380 596777
Fax: +44-2380 596204
Email: icpo@noc.soton.ac.uk

WCRP Decadal Prediction

Predicting The Climate Of The Coming Decades

RSMAS, Miami, USA
January 11-14, 2010

Overview | Presentations | Posters | Venue & Logistics | Organizing Committee

Poster Abstracts

Subrahmanyam Bulusu
Department of Earth and Ocean Sciences, University of South Carolina
Influence of Solar Radiation Absorbed by Phytoplankton on Ocean Circulation and Climate

Roque Vinicio Cespedes¹, Greg J. Holland, Linda O Mearns, and Brian J. Soden
UCAR/NCAR/Significant Opportunities in Atmospheric Research and Science(SOARS) 2009
Decrease in the summer rainfall of the southern United States coast and the Caribbean due to climate change

Gokhan Danabasoglu, Steve Yeager, and Joe Tribbia
National Center for Atmospheric Research
The Atlantic Meridional Overturning Circulation and its Variability in the Community Climate System Model version 4 (CCSM4)

Ming Feng
Decadal variability in the tropical Pacific and influences on the Leeuwin Current in the southeast Indian Ocean

Laura Feudale
African Monsoon variability at different time scales

Ed Hawkins
NCAS-Climate and University of Reading
Decadal predictability of the Atlantic: estimation of optimal perturbations

Ed Hawkins¹,² and Rowan Sutton²
¹NCAS-Climate, ²University of Reading
The potential to narrow uncertainty in regional climate predictions

Barry Heimlich
FAU Center for Environmental Studies
Effect of Sea Level Rise on Southeast Florida's Water Resources

Charles Jones
University of California
Decadal variations in the Madden-Julian Oscillation

Ambarish V. Karmalkar¹, Henry F. Diaz², and Raymond S. Bradley¹
¹Department of Geosciences, University of Massachusetts, ²NOAA/ESRL/CIRES, University of Colorado
Climate Change Projections for Central America

K. H. Kilbourne¹, T. M. Quinn², ³, R. S. Webb⁴, T. P. Guilderson⁵, ⁶, A. Winter⁷ and J. Nyberg⁸
¹University of Maryland Center for Environmental Science, ²Univ. of Texas, Austin, ³Univ. of Texas, ⁴Earth System Research Laboratory, NOAA, ⁵Lawrence Livermore National Laboratory, ⁶Univ. of California at Santa Cruz, ⁷Univ. of Puerto Rico, ⁸Geological Survey of Sweden
Caribean climate since 1469: a coral perspective on tropical Atlantic multidecadal variability

David H. Krantz
Center for Research on Environmental Decisions, Columbia University
Time horizons in decision making and decadal-scale climate variation

Shoji Kusunoki
Climate Research Department, Meteorological Research Institute
Future Change in Precipitation Intensity of East Asian Summer Monsoon

Sang-Ki Lee¹, David Enfield¹ and Chunzai Wang²
¹CIMAS, University of Miami, Miami, FL, ²AOML, NOAA, Miami, FL
Future Impact of Differential Inter-Basin Ocean Warming on Atlantic Hurricanes

Damianos Mantsis and Amy Clement
Rosenstiel School of Marine and Atmospheric Sciences, University of Miami
Simulated Multi-decadal Variability in the Mean Meridional Circulation Over the 20th Century

Christopher S. Moses¹, William Anderson¹, Colin Saunders², Carrie Rebenack¹
¹Department of Earth and Environment, Florida International University, ²South Florida Water Management District
Holocene Decadal to Multidecadal Hydrologic Variability in the Everglades: Climate and Implications for Ecosystem Management

Rym Msadek
GFDL/NOAA
Assessing the decadal predictability of climate in the GFDL coupled model

Ernesto Munoz
Decadal shift in Intra-Americas Sea climate and its relation with precipitation and tornadic activity in the eastern United States

Matt Newman
CIRES/CDC, University of Colorado, and NOAA/ESRL/PSD
Interannual to Decadal Predictability of Global SSTs

Intan Suci Nurhati¹, Kim M. Cobb¹ and Christopher D. Charles²
¹Georgia Institute of Technology, ²Scripps Institute of Oceanography
Multi-proxy coral records of decadal-scale tropical Pacific SST and hydrological variability during the 20th century

Joseph Park
SFWMD
Florida Extreme Tidal Statistics for Water Management Planning

Scott B. Power
Decadal climate variability in the Pacific - what causes it and can the variabiltiy be predicted?

Takashi T. Sakamoto¹, Masayoshi Ishii²,¹, Yoshiki Komuro¹, Hiroaki Tatebe¹, Akira Hasegawa³, Hideo Shiogama³, Takahiro Toyoda¹, Seita Emori³,¹,⁴, Hiroyasu Hasumi⁴ and Masahide Kimoto⁴
¹Japan Agency for Marine-Earth Science and Technology, ²Meteorological Research Institute, Japan Meteorological Agency, ³National Institute for Environmental Studies, ⁴Center for Climate System Research, University of Tokyo
MIROC4.0 - a high-resolution AOGCM for the near-term climate prediction

Scott Sellars and Tom Knutson
Geophysical Fluid Dynamics Laboratory
Southeast U.S. Multi-Decadal Surface Temperature Trends: Models vs Observations

Amanda J. Waite and Peter K. Swart
Stable Isotope Laboratory, Division of Marine Geology and Geophysics, Rosenstiel School of Marine and Atmospheric Science, University of Miami
Bahamian Sclerosponge Geochemistry Reveals Decadal to Multi-decadal Scale Climate Variability in the Atlantic over the Last 600 Yrs

Ian Watterson and P. H. Whetton
CSIRO
Simple predictions for the coming decade based on time series representing both forced change and variability and incorporating observed trends

Xiaosong Yang¹, Timothy DelSole¹, Ben Kirtman², and Dughong Min²
¹Center for Ocean-Land-Atmosphere Studies, ²Rosenstiel School for Marine and Atmospheric Science, University of Miami
Diagnosing the predictability in CCSM decadal simulations

Sayaka Yasunaka¹, Masayoshi Ishii²,⁴, Masahide Kimoto¹, Takashi Mochizuki², and Hideo Shiogama³
¹Center for Climate System Research, University of Tokyo, ²Japan Agency for Marine-Earth Science and Technology, ³National Institute for Environmental Studies, ⁴Meteorological Research Institute, Japan Meteorological Agency
Impact of new XBT depth correction on near-term climate prediction

Steve Yeager, G. Danabasoglu, J. Anderson, N. Collins, T. Hoar, J. Tribbia
National Center for Atmospheric Research
Estimating the Strength and Variability of the Atlantic Meridional Overturning Circulation in Recent Decades using Community Climate System Model (CCSM) Ocean Hindcast and Data Assimilation Simulations

Subrahmanyam Bulusu
Department of Earth and Ocean Sciences, University of South Carolina
Influence of Solar Radiation Absorbed by Phytoplankton on Ocean Circulation and Climate

This study objective is to quantify and understand how phytoplankton, by absorbing and reflecting solar radiation, may affect the thermal structure and circulation of the upper ocean, climate, and climate change. Results from our exploratory research to date suggest that the effects are significant and occur over large scales. They point to a potentially important climate biofeedback mechanism that is not included satisfactorily in coupled climate models at present.

Simulations from the North American Regional Climate Change Assessment Program (NARCAPP) Geophysical Fluid Dynamics Laboratory (GFDL) AM2.1 timeslice experiment, for current climate (1971-2000) and future climate (2041-2070), were compared and contrasted to assess how May through October accumulated rainfall is responding to climate change along the southern United States coast and in the Caribbean under the Intergovernmental Panel on Climate Change (IPCC) A2 emissions scenario (a scenario of relatively high emissions increase). The simulations were done on a global domain at a horizontal resolution of roughly 50 km.

There is an overall decrease of about 200 mm (30 percent) in the May through October rainfall in the region of the southern United States coast and the Caribbean. The absolute decrease is larger in the regions that receive the most rain. However, proportionally, the decrease is larger in the regions that receive the least rain. For the subregion of Florida, rainfall time series indicate a delay of the region's late wet period in the future climate. This shift needs to be further examined to determine its significance and underlying physical processes. Florida also received less rainfall in future climate, but the standard deviation of the early and late wet periods was found to be larger. This is in accord with the findings of the IPCC of an increase in global extremes as a result of climate change. In a future study, the time series for four other subregions will be analyzed.

The Atlantic Meridional Overturning Circulation (AMOC) and its variability are presented from the long, fully-coupled pre-industrial and present-day control integrations as well as from a five-member ensemble of the 20th Century simulations that use the Community Climate System Model version 4 (CCSM4). In CCSM4, the ocean model has a nominal 1 degree horizontal resolution with 60 vertical levels. The atmospheric model uses a finite volume dynamical core with a nominal 1 degree horizontal resolution and 26 vertical levels. We analyze the 1000-year pre-industrial control integration and the associated 20th Century simulations that use the standard CCSM4 configuration. In addition, we investigate the AMOC mean and variability in the 530-year pre-industrial and 600-year present-day control experiments in which the atmospheric model uses a nominal 2 degree horizontal resolution. An Empirical Orthogonal Function (EOF) analysis of the annual-mean AMOC time series reveals a dominant mode associated with a South Atlantic cell with a period of 4-5 years in all control experiments. This variability in AMOC appears to be highly correlated with the model's ENSO time series, also showing large variability at 4-5 year periods. After a five-year running-mean, the EOF analysis of the cases with the 2 degree atmospheric resolution produces a dominant mode associated with an overall strengthening and weakening of the North Atlantic Deep Water cell. This mode accounts for more than 60% of the total AMOC variance. The corresponding spectra of the principal component time series show statistically significant peaks at several decadal periods, including 25 years for the pre-industrial and 27 years for the present-day control simulations.

In both of cases, the AMOC variability has an amplitude of 1.2 Sv. Similar AMOC variability is seen in a preliminary analysis of the pre-industrial CCSM4 control simulation with the 1 degree atmospheric resolution. Model simulations also indicate sensitivity of the mean AMOC transport and its variability to a parameterization of the Nordic Sea overflows. Our analysis indicates that there are important differences in the AMOC behavior and variability between these CCSM4 simulations and those of the CCSM3.

Ming Feng
Decadal variability in the tropical Pacific and influences on the Leeuwin Current in the southeast Indian Ocean

Laura Feudale
African Monsoon variability at different time scales

Ed Hawkins
NCAS-Climate and University of Reading
Decadal predictability of the Atlantic: estimation of optimal perturbations

Future decadal climate forecasts will need to sample the uncertainty in initial conditions efficiently, and are likely to rely on ensembles initialised using small perturbations to oceanic and atmospheric conditions. In order to design efficient ensembles there is a need to identify those perturbations that grow most rapidly. Such perturbations may also be useful to identify where new ocean observations could improve forecast skill. We have employed two different methods to estimate such optimal perturbations for decadal forecasts of the Atlantic Ocean.

Firstly, we use linear inverse modelling (LIM) to find the initial condition anomalies which grow most rapidly for a range of GCMs. The regions identified as most sensitive to small perturbations tend to be located in the far North Atlantic. Significant non-normal amplification is found, and the mechanisms of amplification often involve a basin-wide overturning circulation response to the small perturbations.

Secondly, we are using an ensemble based technique which, unlike the LIM approach, enables optimal perturbations to be estimated for specific initial conditions, e.g. a high or low overturning strength. For HadCM3, the most sensitive regions are again identified as the far North Atlantic.

Ed Hawkins¹,² and Rowan Sutton²
¹NCAS-Climate, ²University of Reading
The potential to narrow uncertainty in regional climate predictions

Decision makers are increasingly seeking predictions of regional and local climate changes. An important issue for these decision makers, and for organisations that fund climate research, is what is the potential for climate science to deliver reductions in uncertainty in such predictions?

Uncertainty in climate predictions arises from three distinct sources: internal variability, model uncertainty and scenario uncertainty, which are quantified using the CMIP3 range of GCMs. For predictions of changes in temperature on decadal timescales and regional spatial scales, we show that uncertainty for the next few decades is dominated by sources that are potentially reducible through progress in climate science. Furthermore, we find that model uncertainty is of greater importance than internal variability for temperature predictions. However, for precipitation, the contribution from internal variability is larger, and adaptation decisions may need to be made in the context of high uncertainty concerning regional changes in precipitation.

These findings have implications for managing climate adaptation. Because adaptation costs are very large, and greater uncertainty about future climate likely requires more expensive adaptation, reducing uncertainty in climate predictions is potentially of enormous economic value. Climate science investments should therefore be targeted on the most promising opportunities to reduce prediction uncertainty.

Barry Heimlich
FAU Center for Environmental Studies
Effect of Sea Level Rise on Southeast Florida's Water Resources

The poster summarizes information from the section on sea level rise and its impacts on Southeast Florida’s water resources from a recent study conducted at Florida Atlantic University’s Center for Environmental Studies under the sponsorship of the National Commission on Energy Policy. Charts and graphs illustrate how sea level rise and other climate change impacts are likely to affect Southeast Florida’s ground waters, surface waters, and water supply in the next several decades. The findings show that sea level rise of as little at 3 to 9 inches could exacerbate saltwater intrusion, reduce groundwater flow, increase water tables, and significantly increase the risk of flooding during heavy rain events and storms.

Reference:
Heimlich, Bloetscher, Meeroff & Murley, 2009, Southeast Florida’s Resilient Water Resources: Adaptation to the Sea Level Rise and Other Impacts of Climate Change, http://www.ces.fau.edu/projects/climate_change

Charles Jones
University of California
Decadal variations in the Madden-Julian Oscillation

The Madden–Julian oscillation (MJO) is the most prominent form of tropical intraseasonal variability. This study investigates the variability of the MJO in the instrumental record as well as projections of MJO activity for the coming decades. The activity of the MJO is defined in terms of number of events occurring in a given interval of time and multiannual-to-decadal changes in the MJO are investigated in coupled global climate models and in a stochastic model of the MJO.

The stochastic model is driven by observed SST anomalies and a large ensemble of simulations was performed to infer the activity of the MJO in the instrumental period (1880–2008). The model is capable to reproduce the activity of the MJO during the reanalysis period. The simulations indicate that the MJO exhibited a regime of near normal activity in 1948–1972 (3.4 events year-1) and two regimes of high activity in 1973–1989 (3.9 events) and 1990–2008 (4.6 events). Stochastic simulations indicate decadal shifts with near normal levels in 1880–1895 (3.4 events), low activity in 1896-1917 (2.6 events) and a return to near normal levels during 1918–1947 (3.3 events). The results also point out to significant decadal changes in probabilities of very active years (5 or more MJO events): 0.214 (1880–1895), 0.076 (1896–1917), 0.197 (1918-1947) and 0.193 (1948–1972). After a change in behavior in the 1970s, this probability has increased to 0.329 (1973–1989) and 0.510 (1990-2008). The observational and stochastic simulations presented here call attention to the need to further understand the variability of the MJO on a wide range of time scales. The variability of the MJO during 1880-2008 is compared against simulation of the ECHAM model. Projections of MJO activity are performed using the stochastic model and simulations with the ECHAM model.

Future changes in precipitation amount and variability are among the most important and serious projected consequences of climate change. Central America shows most of its climate variability in precipitation. Thus, the large hydrological response to global warming can have negative consequences on agricultural activities and the ecosystem dynamics in the region. Indeed, Central America is considered to be a climate change hot-spot in the tropics mainly due to a future decrease in precipitation amount and an increase in precipitation variability as projected by the IPCC models. The coarse resolution general circulation models (GCMs) do not provide climate information at spatial scales appropriate for impacts assessment. Central America has high biodiversity, and it harbors high-value ecosystems that are susceptible to future changes in climate. In fact, conservation international has declared Central America as one of the largest biodiversity hotspots in the world. Therefore, it is important to provide realistic regional climate change scenarios to assist in adaptation and mitigation work in the region. A regional climate model PRECIS was used to carry out two experiments: (i) the baseline (present-day) run and (ii) the SRES A2 run, both performed at 25 km horizontal resolution.

The tropical North Atlantic Ocean, including the Caribbean Sea, is an integral part of the primary modes of Atlantic climate variability, including the Atlantic Meridional Overturning Circulation (MOC) and multidecadal-scale hemispheric surface temperature anomalies (Atlantic Multidecadal Variability, AMV). However because modern climate and ocean circulation records are relatively short, it is unclear if AMV noted in the 20th century is a persistent feature of the Atlantic climate system, if it is connected to MOC, and if the tropical Atlantic plays an active or passive role in the process.

Isotopic and elemental ratio data from two continuous coral cores that grew offshore from Puerto Rico are used to reconstruct climate variability in the Caribbean region since the beginning of the Little Ice Age. New data include annually-resolved coral Sr/Ca and _18O records spanning the years 1469 to1669. Spectral analysis of data from both cores indicate a persistent ~60 year period signal that appears to be primarily temperature-related in the earlier part of the record but is associated with a shift in the trade winds and the regional salinity gradient over the period of overlap with the instrumental record. We find that the northern Caribbean was relatively cool during most of the Little Ice Age, with a ~2° warming trend since about 1750, broadly consistent with other temperature reconstructions of this region. These new data demonstrate that AMV is a persistent feature of the modern tropical Atlantic Ocean, at least over the last 540 years, paving the way for more process-oriented studies on its relationship to MOC and global climate processes.

David H. Krantz
Center for Research on Environmental Decisions, Columbia University
Time horizons in decision making and decadal-scale climate variation

Should I purchase 100 excellent umbrellas (a 20-year supply, for me) at the very attractive price of $250? There is, of course, the question of where I would store so many. I also worry about extra moving costs if I enter an assisted-living facility in a few years. Also, perhaps precipitation will be much less and I won't really need that many (but I don't know whether I can trust long-term prediction of a drier climate). Furthermore, umbrella technology may improve over the next two decades. I also need to consider the opportunity cost of committing $250 now; compared with gradual spending across 20 years, while earning 5% interest, the true cost of buying the 100 umbrellas now comes to nearly $20 per year, rather than $12.50. In any case, should someone my age (71 years) even be thinking 20 years ahead?

With a one-day time horizon, a person might decide whether or not to carry an umbrella, on the basis of an uncertain weather forecast or perhaps just on the basis of climatology, the probability of rain at a specific location and season. The opportunity to purchase 100 excellent umbrellas might shift the time horizon for decision making from one day to 20 years. Likewise, the attempt to consider decade-scale shifts of climate regime might induce a long time horizon for decision making, and this in turn could lead to consideration of longer-term plans, whether or not there is actually any skill in a decadal-scale forecast.

The preceding incorporates a modern view of decision making. Choice is often highly dependent on context. The very plans considered, as well as their evaluation, depend on social context, which activates a variety of important social goals; plans and their evaluation also depend on situation-dependent time horizons, which strongly influence tradeoffs between jam today versus more jam tomorrow; and finally, plans and their evaluations depend on the way in which uncertainty about the future is incorporated into the decision process. Awareness of possible future plans, perhaps advertised commercially or by government or non-profit institutions, can also change what goals are activated and what time horizons are used in the decision process.

A change in time horizon of planning may affect important characteristics of human "irrationality", including myopia, hyperbolic discounting, loss aversion, probability neglect, and "not in my term of office" thinking. Both institutional changes and a focus on decadal-scale climate regimes may thus lead to better decision making in domains such as agriculture and protection against hazards.

Shoji Kusunoki
Climate Research Department, Meteorological Research Institute
Future Change in Precipitation Intensity of East Asian Summer Monsoon

The reproducibility of precipitation intensity by the Couple Model Intercomparison Project 3 (CMIP3) models was investigated for the East Asia summer monsoon rain band "the Baiu rain band". The Simple Daily precipitation Intensity Index (SDII) in June and July are calculated for the last ten years (1991-2000) of 20th Century Climate in Coupled Models (20C3M) simulations with respect to sixteen CMIP3 models. For the verification, we used the one-degree daily data of Global Precipitation Climatology Project (GPCP). Model's reproducibility was measured by the Taylor index (Taylor 2001). Models tend to underestimate precipitation intensity. One of the highest horizontal resolution models shows relatively higher reproducibility compared with lower horizontal resolution models, but the advantage of higher resolution models over lower resolution models was not evident. Models with higher reproducibility of precipitation climatology tend to show higher reproducibility of precipitation intensity.

Future change in precipitation intensity was investigated by the global warming projections for green house gas emission scenario A1B. Target periods are 10 years from 2051 to 2060 as well as 10 years from 2091 to 2100. After selecting five models with higher reproducibility of precipitation intensity for present-day climate, skill weighted ensemble average was calculated. Precipitation intensity increases almost all over the East Asia. Increase in precipitation intensity for the period 2091-2099 is larger than that for the period 2051-2060.

Sang-Ki Lee¹, David Enfield¹ and Chunzai Wang²
¹CIMAS, University of Miami, Miami, FL, ²AOML, NOAA, Miami, FL
Future Impact of Differential Inter-Basin Ocean Warming on Atlantic Hurricanes

Global climate model simulations forced by future greenhouse warming project that the tropical North Atlantic (TNA) warms at a slower rate than the tropical Pacific in the 21st century, consistent with their projections of the weakening Atlantic thermohaline circulation and Pacific Walker circulation. Here, we use an atmospheric general circulation model to advance a consistent physical rationale that the preferential warming of the tropical Pacific over the TNA increases the vertical wind shear and static stability over the main development region for hurricane, and thus deceases Atlantic hurricane activity in the 21st century. The preferential warming of the tropical Pacific accelerates the warming of the tropical troposphere over the TNA, via a tropical teleconnection mechanism, and thus increases the static stability and decreases the convection over the TNA. The anomalous diabatic-cooling, in turn, forces the formation of a stationary baroclinic Rossby wave northwest of the forcing region, consistent with the Gill’s simple model of tropical atmospheric circulations, and thus induces a secular increase of the vertical wind shear over the TNA.

Damianos Mantsis and Amy Clement
Rosenstiel School of Marine and Atmospheric Sciences, University of Miami
Simulated Multi-decadal Variability in the Mean Meridional Circulation Over the 20th Century

The variability in the zonally averaged atmospheric meridional circulation and its relation with the underlying SSTs during the last 150 years is investigated using two atmospheric GCMs. A multi-decadal signal, with an approximately 70 year timescale, is identified in the zonal mean cross-equatorial meridional wind. This signal is associated with a latitudinal shift in the ascending branch of the Hadley cell and precipitation in the tropics, as well as a change in the meridional heat transport of both atmosphere and ocean. These changes are well-correlated with the interhemispheric SST difference. When northern hemisphere SSTs are colder than in the southern hemisphere, the ITCZ and precipitation shift to the south in a zonal mean sense, and the northward atmospheric energy transport increases. Previous studies with idealized climate forcings have shown similar results, but the findings presented here highlight the potential relevance of the interhemispheric SST contrast for understanding 20th century climate changes.

The Florida Everglades are a complex, unique ecosystem. Adding to the complexity, a system of canals and gates control the flow of waters from central Florida southward into the Everglades, and ultimately Florida Bay and the Gulf of Mexico. With south Florida’s distinct wet and dry seasons, the hydrology has driven ecosystem evolution over the last 4-5 kya. However, since the 1920s the water content of the Everglades has largely been anthropogenically modulated, with the exception of the natural variability of evaporation and precipitation over the large area south of the Tamiami Trail. Because of the incredibly flat nature of the Everglades, small changes in the freshwater balance have substantial impacts on the diversity and distribution of organisms. Decadal and multidecadal variability in precipitation, hurricane incidence, and sea level rise all have important effects on the ecosystem. During the instrumental record, the natural precipitation across south Florida has been strongly influenced by combinations of the Atlantic Multidecadal Oscillation, Pacific Decadal Oscillation, and ENSO. Here we discuss evidence of natural climate variability impacts on the ecosystem beyond the anthropogenic hydrological controls. Proxy environmental data from seeds, charcoal, and trees, plus the sparse, but available, instrumental records provide evidence of changes in the ecosystem over the Holocene, and suggest considerations for future management

Rym Msadek, Tom Delworth, Keith Dixon and Tony Rosati
GFDL/NOAA
Assessing the decadal predictability of climate in the GFDL coupled model

Numerous modeling studies have shown evidence of decadal variability in the Atlantic Meridional Overturning Circulation (AMOC), suggesting a potential predictability of the associated climate impacts. The predictability attributed to this climate variability is investigated in this study using the GFDL coupled model.

First the multidecadal climate variability is estimated in a 700-yr long control integration of the CM2.1 global coupled model, in which the greenhouse gases concentration is held fixed to the value of 1860. Perfect model ensemble experiments are initialized from this control integration, every 100 years. Each ensemble consists of 10 experiments with slightly perturbed atmospheric but the same oceanic initial conditions. Our study focuses on the potential predictability of the AMOC and its surface and subsurface signatures. We show that not only the AMOC is predictable up to 20 years, but observable AMOC fingerprints are predictable too. Indeed previous studies showed that the leading mode of upper ocean heat content, subsurface temperature and sea surface height has the same dipolar structure in the North Atlantic, with opposite signs in the subpolar gyre and along the Gulf Stream path. This pattern was suggested to be a distinctive signature of AMOC fluctuations. Here we show that these fingerprints have the same decadal predictability skills than the AMOC. A weaker predictability is found for the basin-average North Atlantic sea surface temperature called the Atlantic Multidecadal Variability (AMV) index. Predicting decadal variations of AMOC fingerprints that are extensively observed is highly relevant for both monitoring and forecasting purposes. Moreover, subsurface and depth integrated variables are less affected by radiative forcing and may reflect AMOC variations to a larger extent than sea surface temperature. Although the largest predictive skills are found in the North Atlantic, predictable signals in precipitation and sea level pressure are also found over the western tropical Pacific, and might reflect low frequency changes associated with the model ENSO.

Assessing the potential predictability of climate relies on the understanding of the origin of the multidecadal variability. The mechanisms governing this variability are investigated in three versions of the GFDL global climate model, which differ mainly in the resolution of the atmospheric and oceanic components. Clear decadal to multidecadal fluctuations of the AMOC are found in CM2.1 and CM2.3, which have the same 1 degree ocean model but a different atmospheric component. In CM2.4 which is eddy-permitting in the ocean, the AMOC does not show much low-frequency variability. Because of the differences in the simulated North Atlantic variability, the regional climate impacts associated with the AMOC will likely be different in these three models. Our study suggests that decadal prediction skills critically depend on model formulation. It emphasizes that large uncertainties remain in understanding and modeling decadal natural variations in the ocean, which are critical for decadal predictability.

Ernesto Munoz
Decadal shift in Intra-Americas Sea climate and its relation with precipitation and tornadic activity in the eastern United States

Matt Newman
CIRES/CDC, University of Colorado, and NOAA/ESRL/PSD
Interannual to Decadal Predictability of Global SSTs

An empirical technique that fits and then tests a multivariate red noise model to the SST observations, called linear inverse modeling (LIM), provides an excellent approximation of the observed evolution of Pacific SST anomalies on time scales ranging from weeks to years. This technique is now applied to the global ocean, using observed simultaneous and one-year lag correlation statistics of July-June averaged SST from the Hadley Centre Global Sea Ice and Sea Surface Temperature (HadISST) dataset for the years 1900-2007. Multi-year forecast skill is assessed and shown to be comparable to and often better than that of coupled GCMs, suggesting that at a minimum the LIM is an important benchmark for decadal forecasts.

In the stable dynamical system determined by LIM, longer-term predictability and forecast skill result from the least damped eigenmodes of the dynamical operator. The propagating eigenmodes that strongly correspond to canonical ENSO and PDO patterns all have short decay times, so that they do not actually represent regular oscillations and correspond to little forecast skill beyond two years. Instead, long-term skill arises from a few stationary eigenmodes, one similar to the centennial trend and another that corresponds to multidecadal fluctuations of a pattern that potentially represents natural decadal variability. The observed LIM is also compared to LIMs constructed from the output of each 20th century (20c3m) ensemble member from the CMIP3 experiments used in AR4. It is found that in all the 20th century runs the second eigenmode is not only poorly captured but is also strikingly less persistent, suggesting that there remains predictable decadal variability yet to be captured by the models.

A comprehensive understanding of tropical Pacific decadal-scale climate variability, both natural and anthropogenic, would likely improve forecasts of future regional climate changes. Coral geochemical records from the tropical Pacific provide high-resolved, continuous climate archives that complement the instrumental climate datasets over the last century. We employ a multi-proxy approach to corals from three islands of the central tropical Pacific (Palmyra, Fanning and Christmas; 2ºN-6ºN, 157ºW-162ºW) to reconstruct SST (via Sr/Ca) and salinity (via oxygen isotopic composition of seawater or delta¹⁸O_sw) over the 20th century. On interannual timescales, the SST proxy records are highly correlated to the NIÑO3.4 index of El Niño-Southern Oscillation (ENSO) variability (R=0.7 to 0.9). The coral-based salinity proxy records exhibit correlations of roughly -0.5 with NIÑO3.4, consistent with mixed layer freshening expected in the central tropical Pacific during El Niño events. On decadal timescales, the century-long Palmyra SST proxy record shows a strong link with the North Pacific Gyre Oscillation (NPGO) Index (R=-0.46 for 1950-1998), which is the second principal component of North Pacific SST variability [Di Lorenzo et al., 2008], but is poorly correlated to the Pacific Decadal Oscillation (PDO), the first principal component of North Pacific SST variability. Surprisingly, our Palmyra salinity proxy record exhibits a strong connection (R=-0.57) to the PDO, indicating that central tropical Pacific hydrology is sensitive to PDO variability even though central tropical Pacific SST is linked to NPGO variability. The last forty years of the coral salinity proxy record are dominated by a freshening trend that is unprecedented in the last millennium [Cobb et al., 2003], consistent with an anthropogenic strengthening and/or an equatorward migration of the Intertropical Convergence Zone in the central tropical Pacific. The coral freshening trend is consistent with predictions of an enhanced hydrological cycle under continued anthropogenic forcing [Held and Soden, 2006]. Taken together, the new coral data imply strong linkages between tropical and extratropical Pacific decadal-scale variability over the last century, but suggest that hydrological shifts have dominated the response of the tropical Pacific climate system to anthropogenic greenhouse forcing.

References:
Cobb et al. (2003), Nature, 424, 271-276.
Di Lorenzo et al. (2008), Geophys. Res. Lett., 35, L08607
Held, I. M., and B. J. Soden (2006). J. Clim., 19, 5686-5699.

Joseph Park
SFWMD
Florida Extreme Tidal Statistics for Water Management Planning

Climate forced sea-level rise will impact coastal aquifer sustainability and management, and will also affect the performance and operation of coastal drainage water control structures used to mitigate floods. Long-term consequences of these effects can be anticipated to a certain degree based on model analysis. A less scrutinized process which may have more of an impact in the near-term than mean sea-level rise are extreme events in coastal sea-level. We analyze long-term tidal records at three Florida stations in order to extract statistics of extreme water levels. Correlations are assessed with respect to epoch and the AMO, and probability distributions of events are estimated. Such statistics may provide decision-support based on climate index forecasting applicable to water management planning and forecasting.

Scott B. Power
Decadal climate variability in the Pacific - what causes it and can the variabiltiy be predicted?

Preliminary results, especially mean climate and inter-annual to decadal variability, on a general circulation climate model, Model for Interdisciplinary Research on Climate (MIROC) version 4.0, were presented. The model was developed by the Center for Climate System Research (CCSR), the University of Tokyo; National Institute for Environmental Studies (NIES); and Japan Agency for Marine-Earth Science and Technology (JAMSTEC). MIROC4.0 is an updated model from the previous version MIROC3.2_hires, which was used to contribute to the IPCC AR4. Most of the model components are the same as MIROC3.2_hires, but the atmospheric component is changed to T213 spectrum model from T106 one. The ocean component is the same as that used in MIROC3.2_hires, whose horizontal resolution is 0.28125° zonally and 0.1875° meridionally, while the latitudinal range where the Gent-McWilliams (GM) parameterization were applied was changed in order to improve the climatological distribution of the sea surface temperature (SST). The other components, sea ice, land surface process, and river routing models, are also same as the previous model. To obtain the radiative balance, parameters associated with radiation, clouds, and aerosols are tuned. Using this model, spin-up and control experiments (100 years) under the condition of year 1950 without flux adjustment were conducted.

Scott Sellars and Tom Knutson
Geophysical Fluid Dynamics Laboratory
Southeast U.S. Multi-Decadal Surface Temperature Trends: Models vs Observations

In this analysis, a low frequency filter (10-yr running mean) was applied to IPCC AR4 historical climate forcing runs (20c3m). The purpose was to evaluate regional long-term surface temperature trends within 22 IPCC AR4 models for the Southeastern U.S. (SE U.S.). The SE U.S. is a very interesting region in terms of long-term surface temperature trends. In fact, the region has apparently cooled slightly since 1900 as estimated using simple linear trend analysis (e.g. IPCC AR4, Fig 3.9). This feature is quite unusual, as the global mean temperature, and surface temperatures in most regions, have increased over that time period (Exceptions are the SE U.S. and the high latitude North Atlantic).

Looking at the SE U.S. temperature time series in detail (HadCRUT3v), a warming occurred during the early 20th century, with relatively warm levels persisting from the 1930s through the 1950s, followed by a relatively cool period from the 1960s through the early 21st century. Adopting 1880-1920 as a base period for an analysis and comparison to models, the cool period from the 1960s on is quite unusual in comparison to the historical climate forcing runs from the CMIP3 archive. This is particularly so since about 1990, when the observed SE U.S. temperature has been cooler (relative to 1880-1920) than *any* of the 70+ runs examined in the archive.

Another unusually cool period in the SE U.S. record is centered in the 1860s. This earlier cool period is even cooler than the one beginning in the 1960s, and suggests that low-frequency internal climate variability may play a more dominant role than greenhouse warming in the region's past low-frequency climate variations. This early cool period plays a very important role in the regional long-term surface temperature trend analyses and further investigation of this early cool period was prompted.

We have begun a preliminary analysis of station data to try to assess the likely reliability and robustness of the 1860s cool period in the record using station data obtained from the Environmental Document Access and Display System (EDADS) - Forts Database, and results will be reported on at the meeting.

Recently, much attention has been focused on the use of high resolution paleoclimate proxies for the reconstruction of decade to century scale climate variability. In particular, this includes the use of tree rings and corals. Unfortunately, long-term (500 years or more) reconstructions from these archives are relatively sparse in the literature and those which do exist show marked disagreement prior to the instrumental period. The skeletons of long-lived sclerosponges offer another opportunity to investigate changes in climate over this time. In 1993, a ~600 year old sclerosponge specimen (LSI-4) of Ceratoporella nicholsoni was collected from 133 meters depth in Exuma Sound in the Bahamas. The sample was U/Th dated, and micromilled at a resolution of ~2 samples per year. The individual samples were then analyzed for stable carbon and oxygen isotopes, as well as minor and trace element compositions. Using published calibration equations, temperature and salinity have been reconstructed from the geochemical measurements of this specimen.

The temperature reconstruction of LSI-4 indicates a warming of approximately 1°C over the last 150 years, which is consistent in amplitude and small-scale variability with other published sclerosponge records from the Caribbean. The reconstructions from this sclerosponge also reveal pronounced multi-decadal variability at a number of periodicities. The temperature reconstruction from Sr/Ca of the Exuma Sound sclerosponge shows periodicities of 15 and 30 years. The Sr/Ca derived temperature and _18O of the skeleton are then used to calculate paleo-salinity. The reconstructed salinity from LSI-4 depicts cyclicity with a period of ~20 years between 1400 and 1790. From 1790 to 2000, however, the dominant mode appears to switch to a roughly 60 year periodicity. This 60 year periodicity in the salinity is consistent with that of the Atlantic Multi-decadal Oscillation (AMO).

The sensitivity of the LSI-4 salinity record to the AMO is likely controlled by the presence of the North Atlantic Salinity Maximum Waters (SMW) in the region. Changes in the evaporation and precipitation budget over the source region for the SMW result from phase changes in the AMO, which is ultimately communicated to the Bahamas via the salinity, strength, and extent of the SMW. In keeping with this, the ~.5 psu increase in salinity seen in the reconstruction from LSI-4 over the last 50 years is consistent with observed changes in the SMW. Assuming that the geochemistry of LSI-4 is responding to AMO related variability over its lifespan, then this record is the longest reconstruction to date. However, like some of the recently generated coral reconstructions, it disagrees with tree ring indices prior to 1850, and changes periodicity around 1790. This raises many important questions regarding the origins and longevity of the AMO and further highlights the need for additional long-term paleo proxy reconstructions.

Ian Watterson and P. H. Whetton
CSIRO
Simple predictions for the coming decade based on time series representing both forced change and variability and incorporating observed trends

Probabilistic projections of change in temperature and precipitation over the globe for the decade 2090-2099 relative to a base climate from 1980-1999 have been derived from the CMIP3 simulations for the A1B scenario. By combining these with the best estimate for global warming in intervening years, time series representing the range of forced change for a location can be constructed. This range can be augmented to allow for natural variability of decadal means. The approach is applied to the cases of locations in both the US and southeast Australia, where temperatures have risen and rainfall declined sharply in the past two decades. It is then extended to depict the full uncertainty in change relative to an unforced base climate assumed to apply up to 1900. By putting these series in the context of the full observational record, a further representation of the absolute values of the quantities is made. This provides a determination of the likely range of the unforced climate itself. Based on the time series, one can predict, for example, that the southeast Australia temperature for the coming decade has a probability of 0.5 of being even higher than the record value of the past decade. Rainfall is unlikely to be as low as it was then, but has a probability of 0.7 of being lower than the average for the rather moist 1961-1990 period.

We propose a new histogram method to diagnose decadal predictability and apply it to hindcasts from CCSM. The histogram of the signal variance to total variance can be generated as a function of forecast lead time, and the significance test of the histogram can be done by a Monte Carlo method with reshuffling all ensembles. We also have developed an index, based on the month-to-month lag-correlation, to detect the regions with higher predictability. The index can indicate the persistence of a signal, as well as the reemergence mechanisms of SST. The histogram method and the month-to-month lag-correlation index both demonstrate that the CCSM3 has the predictability ranging from interannual to decadal timescales.

The impact of expendable bathythermograph (XBT) depth bias correction introduced by Ishii and Kimoto (2009) on near-term climate prediction is presented. Using a coupled atmosphere-ocean climate model, we perform two sets of 10-member data assimilation runs and ensemble hindcast experiments with two different observational data sets of ocean temperature; one is the data set with the bias correction and the other is that without it. Differences between the two sets of assimilated ocean temperatures are significant along the thermocline in the tropics and the subtropics, and differences of two observational data sets substantially remain in the assimilations. Root-mean-square errors of ocean temperature in the hindcasts are reduced thanks to the correction especially in the central-to-eastern tropical Pacific and the central North Pacific, which might be caused by an improvement of the initialization at the equator. The hindcasts of the Pacific Decadal Oscillation are also better agreement with the observations by the correction.

We demonstrate the sensitivity to surface boundary conditions of model simulations of the Atlantic Meridional Overturning Circulation (AMOC) by comparing a suite of historically-forced (1948-2007) ocean only and coupled ocean-ice experiments which differ only in their surface freshwater forcing and sea ice treatments. In the ocean-only configuration, daily sea ice extent is prescribed from satellite observations, and in the coupled ocean-ice configuration, a freely-evolving dynamic and thermodynamic sea ice model is used. The latter configuration represents the CCSM contribution to the Co-ordinated Ocean-Ice Reference Experiments, number 2 (CORE-II) project of the CLIVAR Working Group on Ocean Model Development. For each model configuration, four realizations of the hindcast are compared corresponding to a restoring timescale of surface salinity to observed climatology of 30 days, 1 year, 4 years, and infinity. Common AMOC variations are ascribed to specific forcing terms and high latitude transformations, while the differences in circulation associated with changes in ad hoc and/or poorly-known boundary conditions are used to discriminate between robust and spurious AMOC variability. We quantify the effects of ocean-cryosphere interactions on AMOC variability for both prescribed and freely-evolving sea ice boundary conditions.

The systematic analysis of AMOC sensitivity outlined above is intended to help inform the choice of initial conditions for future projection experiments using CCSM4. An accurate representation of the present state of the AMOC is thought to be essential for the initialization of fully coupled decadal projection simulations to be run for the Intergovernmental Panel on Climate Change (IPCC) Fifth Assessment Report (AR5). The use of ocean data assimilation techniques offers another possibility for increasing the fidelity of historical AMOC simulations. We present results from 21st century CCSM ocean hindcast ensembles which are strongly constrained by ocean observations using Data Assimilation Research Testbed (DART) algorithms. Both types of ocean hindcast (DART and CORE-II) are being considered for initializing future projection experiments. We compare the AMOC strength and variability obtained in each and show preliminary results from fully coupled projection experiments which use the different initializations.

last updated Fri, Jan 8, 2010 by Anna Pirani

WCRP Decadal Prediction

Predicting The Climate Of The Coming Decades

RSMAS, Miami, USA January 11-14, 2010

Poster Abstracts

RSMAS, Miami, USA
January 11-14, 2010