Climate Analysis Section Narrative

Climate Analysis Section (CAS) research has the goal of increasing our understanding of atmospheric and climate variability and climate change through parallel development and analysis of observational, assimilated, model-generated, and model-forcing datasets; and, by using these datasets for empirical studies, diagnostic analyses, model experimentation, and model evaluation, to document comprehensively the variability, its causes, and the processes involved. CAS staff are also heavily involved in designing and advocating the global climate observing system and national and international programs.

In CAS a central ongoing thrust of considerable importance to the research and university communities is the acquisition, evaluation, improvement, and restructuring of datasets, development of climatologies and high-level derived products, while facilitating access to data and documentation in catalogs available through the Internet. The datasets are used extensively in diagnostic, theoretical, and modeling studies, including validation of the Community Climate System Model (CCSM). CAS collaborates closely with the Data Support Section (DSS) of NCAR’s Scientific Computing Division (SCD) in all observational data-related activities. Tools to increase access to and display of data are being developed in conjunction with SCD’s Data Portal activity https://dataportal.ucar.edu:8443/cdp/index.jsp.

CAS research is focused on the atmosphere and its interactions with the surface of the Earth and oceans on a wide range of temporal and spatial scales. Included are studies of the diurnal cycle, intraseasonal variability, interannual variations such as the El Niño-Southern Oscillation (ENSO), the North Atlantic Oscillation (NAO), annular modes, the Tropospheric Biennial Oscillation (TBO), extreme events, and interdecadal variations and longer period trends including paleoclimate. Also included are comprehensive diagnostic studies of the global and regional atmospheric moisture, heat, energy, and momentum budgets; observational analysis of upper ocean and sea ice variations; heat and fresh water budgets in the ocean, and the water cycle on land. CAS scientists interact with other CGD sections especially through the analysis, validation, and diagnosis of divisional climate models.

CAS research spans many topics, but includes several studies that are considered to be part of national and international programs including the Climate Change Science Program, International Satellite Cloud Climatology Project (ISCCP), Global Energy and Water Cycle Experiment, Climate variability and predictability (CLIVAR), Past Global Changes (PAGES), Intergovernmental Panel on Climate Change (IPCC), International Geosphere-Biosphere Programme, and the World Climate Research Programme as a whole. CAS is also involved in developing the Global Climate Observing System, the Global Earth Observing System of Systems, and the United States component through the Interagency Working Group on GEO (IWGEO). Of special note is that CAS is contributing results of data analyses to the Fourth Scientific Assessment of IPCC (AR4); all CAS scientists are contributing authors, while Kevin Trenberth (CAS) and Gerald Meehl (joint with CAS and Climate Change Research Section, CCR) are convening lead authors.

Dataset Development

CAS staff members, and Dennis Shea, Lesley Smith, and Adam Phillips, in particular, continue to collaborate with SCD’s DSS to acquire, evaluate, and reformat data. David Stepaniak has moved from CAS to DSS, which strengthens these collaborations. Ongoing efforts in these areas include analyses of conventional, satellite, and model data. In addition to the normal updates to existing datasets, some recent additions to the CAS data catalogs (http://www.cgd.ucar.edu/cas/dcats.html) include 1) Global Precipitation Climatology Project version 2 combined precipitation dataset updated through the end of 2003; 2) Palmer Drought Severity Index (PDSI) updated through 2003; 3) ERA-40 Adjusted Surface Pressure Anomalies; and 4) SSM/I Monthly Mean Water Vapor version 5 (Remote Sensing Systems) for July 1987–January 2004. A top-level catalog was created by SCD and CAS for the NCAR Community Data Portal outlining all CAS catalog datasets. See: https://dataportal.ucar.edu:8443/cdp/index.jsp. Stepaniak developed a new Web page devoted to "Vertically Integrated Mass, Moisture, Heat, and Energy Budget Products Derived from the NCEP (National Centers for Environmental Prediction)/NCAR Reanalysis," providing the community with complete background and access to a total of 36 monthly mean fields spanning 1979 to 2001 – see http://www.cgd.ucar.edu/cas/catalog/newbudgets/. Shea continues to update the NCAR Instructional Aid entitled "An Introduction to Atmospheric and Oceanographic Datasets," which is available at http://www.cgd.ucar.edu/cas/tn404/. It serves as a ‘data-primer’ for students and those in other fields of research. It describes the general characteristics of atmospheric and oceanographic datasets and how to determine what data are available and where they are located.

Smith has responsibility for the online dataset catalog on the Web, which has been upgraded and added to in the past year. It documents all the observationally-based global datasets developed in CAS in the network Common Data Format (netCDF). Many datasets have been revamped and reconstructed into new formats. Continuing development of expert user guidance and commentary on quality and utility of datasets related to climate variability is led by Phillips, Clara Deser, and James Hurrell (all CAS). The primary purpose of this activity is to catalog and provide easy access to a basic set of historical datasets for interdisciplinary studies of climate variability. The need for this activity has been articulated at several U.S. CLIVAR planning workshops.

Sylvia Murphy (CCSM) and Shea continue to participate in development of a general purpose data processing tool based on SCD’s NCAR Command Language. Processing and graphical capabilities continue to be added to facilitate climate research, in particular, model-to-model and model-to-observed comparisons. A Web site (http://www.cgd.ucar.edu/csm/support/) contains many up-to-date examples. Several activities have been pursued including coordination of a diagnostics working group to facilitate the creation and dissemination of model component (atmosphere, ocean, ice, and land) diagnostics packages; the ability to access and process data in the Hierarchical Data Format-Earth Observing System format (HDF-5); and tutorial workshops on data processing and visualization. Five three-day workshops were held during 2004, four at NCAR and one at the National Ocean Service in Silver Spring, Maryland.

Trenberth, Smith, and Stepaniak continue to evaluate global reanalyses from NCEP/NCAR and from the European Centre for Medium-range Weather Forecasts (ECMWF). The focus has shifted to the new ERA-40 re-analyses from ECMWF. The vertically-integrated mass, moisture, heat, and energy budgets for the atmosphere documentation of products is at http://www.cgd.ucar.edu/cas/catalog/tn430/. The data can be downloaded from the Web browser, and all of the fields are available through anonymous ftp, via ftp://ftp.ucar.edu/cgd/cas/tn430/. The products are all global vertically-integrated grids as individual monthly mean time series from 1979 through 1993 (ERA-15), or 2001 for NCEP and monthly, seasonal, and yearly averages (climatologies). Fields from ERA-40 will be developed in the near future. Trenberth, Smith, and John Fasullo (CAS, who replaced Stepaniak) have begun preliminary evaluations of ERA-40 fields. Initial work is on vertically-integrated mass fields, including water vapor.

Hurrell, with James Hack (Climate Modeling Section, CMS), James Rosinski (CMS), Julie Caron (CMS), and Shea, finished the construction of a new lower boundary forcing dataset for studies with the Community Atmosphere Model (CAM). A method was developed for “blending” sea surface temperature (SST) products from NCEP and the Hadley Centre, and for reducing the number of spurious sea ice concentrations associated with the original SST products. In addition, CAM simulations to assess the atmospheric impact of the new relative to old SST datasets were completed. The need for this documentation lies not only in the fact that CAM is a widely-used community tool, but that other major modeling centers (e.g., the Geophysical Fluid Dynamics Laboratory (GFDL) and NASA) are also employing the new SST product in their modeling activities.

Climate Observations and Analysis

Trenberth was a lead author in the Second Report on the Adequacy of the Global Observing Systems for Climate in support of the United Nations Framework Convention on Climate Change, and is a lead author also on the follow-on Implementation Plan that was made available for review and is in final stages of revision. He has also been extensively involved with the IWGEO, as well as service on several relevant NOAA committees.

Climate Diagnostics

Hurrell, Trenberth, and Shea continue work to carefully evaluate variations in storm tracks associated with global patterns of circulation variability. They are making use of nearly 50 years of NCEP/NCAR reanalysis data band-passed filtered to retain synoptic time scales. The quality of the reanalyses for such a task is also being assessed.

Postdoctoral fellow Jeffrey Yin (CAS) published work from his Ph.D. thesis with David Battisti (University of Washington) on the structure of baroclinic waves in the NCEP/NCAR reanalysis. Using regression analysis of 6-hourly data, they demonstrated that baroclinic waves tilt poleward with height, although the observed tilt is smaller than that predicted by previous theoretical and modeling studies. They also produced an improved estimate of the meridional ageostrophic wind, which enhances the poleward tilt of baroclinic waves. Yin is currently using the lag regression and eddy energy budget analysis techniques developed in his thesis to evaluate the structure of baroclinic waves and storm tracks in the NCAR CAM version 3 (CAM3) and the GFDL Atmosphere Model version 2 (AM2) compared to the ERA-40 and NCEP/NCAR reanalyses.

Trenberth and Smith analyzed the total mass of the atmosphere, which varies mainly from changes in water vapor loading, using global mean surface pressure, specific humidity, and precipitable water from the ERA-40 reanalyses from ECMWF. The mass of the dry atmosphere is estimated to be constant for the equivalent surface pressure to within 0.01 hPa based on changes in atmospheric composition. Global reanalyses satisfy this constraint for monthly means for 1979–2001 with a standard deviation of 0.065 hPa. New estimates of the total mass of the atmosphere and its dry component, and their corresponding surface pressures are larger than previous estimates owing to new topography of the Earth’s surface that is 5.5 m lower for the global mean. Global mean total surface pressure is 985.50 hPa, and the total mean mass of the atmosphere is 5.1480 x 10¹⁸ kg, with an annual range due to water vapor of 1.5 x 10¹⁵ kg. The mean mass of water vapor is estimated as 1.27 x 10¹⁶ kg, and the dry air mass as 5.1352 ± 0.0003 x 10¹⁸ kg. The water vapor contribution varies with an annual cycle of 0.29 hPa range, a maximum in July of 2.62 hPa, and a minimum in December of 2.33 hPa. During the 1982–1983 and 1997–1998 El Niño events, water vapor amounts, and thus total mass, increased by about 0.1 hPa in surface pressure or 0.5 x 10¹⁵ kg for several months. Some evidence exists for slight decreases following the Mount Pinatubo eruption in 1991, and also for upward trends associated with increasing global mean temperatures, but uncertainties due to the changing observing system compromise the evidence.

Using the ERA-40 reanalyses for 1958 to 2001, an analysis by Trenberth, Smith, and Stepaniak was made of global patterns of monthly mean anomalies of atmospheric mass to examine whether global patterns of behavior exist requires analysis of all seasons together. Empirical orthogonal function (EOF) analysis, R-mode varimax rotated EOF analysis, and cyclostationary EOF (CSEOF) analysis tools were used to explore patterns and variability on interannual and longer time scales. Clarification is given of varimax terminology, and procedures that have been previously misinterpreted. The dominant global monthly variability overall is associated with the Southern Annular Mode (SAM), which is active in all months of the year and exhibits a great deal of natural unforced variability. The third most important pattern is the Northern Annular Mode (NAM) and associated NAO. Neither of these are really global modes, although they co-vary on long timescales in association with tropical or external forcing. For monthly data, the second mode corresponds to ENSO, which is truly global in extent. It exhibits more coherent evolution with time and projects strongest onto the interannual variability, where it stands out by far as the dominant mode in the CSEOF analysis. Unlike CSEOF analysis, standard EOF and varimax analyses are not able to evolve with time of year unless the analysis is stratified by season. Varimax analysis is able to extract the SAM, NAM, and ENSO modes very well, however.

Further analysis of the mean annual cycle and variability of vertically-integrated atmospheric energy and heat budgets by Trenberth and Stepaniak has been featured in the Symons lecture of the Royal Society. The focus is the flows of energy through the climate system, whereby the incoming radiant energy is transformed into various forms (internal heat, potential energy, latent energy, and kinetic energy) moved around in various ways primarily by the atmosphere and oceans, stored and sequestered in the ocean, land, and ice components of the climate system, and ultimately radiated back to space as infrared radiation. Results are presented for the energy transformations, transports, uptake, storage and release, and the processes involved for the region 60°N to 60°S for the solstitial seasons and their differences to highlight the annual cycle. Values are provided for the seasonal uptake and release of heat by the ocean that substantially moderate the climate in maritime regions. Challenges in better determining the surface heat balance and its changes with time are discussed.

This figure shows vertically integrated atmospheric diabatic heating for 1979–2001 averages computed from NCEP/NCAR reanalyses as a residual. The bottom panel highlights the drivers of the monsoon systems and its seasonal reversal, while the top panels show the main heating for the mean climate in each season. The latent heating dominates the diabatic heating patterns. Trenberth et al. (2004)

Advanced Study Program (ASP) postdoctoral fellow Sungsu Park, together with Deser and Michael Alexander (Climate Diagnostics Center, CDC/NOAA), have completed a comprehensive study on estimation of the surface heat flux response to SST anomalies over the global oceans. The surface heat flux feedback is a key quantity needed for understanding and modeling the nature of large-scale air-sea interaction in climate variability studies. This work uses ship-derived estimates of turbulent energy fluxes and satellite-derived estimates of radiative fluxes to derive the surface heat flux response to local SST anomalies. The remote influence of ENSO is accounted for through simple linear regression analysis. Over most of the global oceans, the feedback is negative, implying a damping of SST anomalies via surface heat fluxes. However, in particular regions and seasons the feedback may be positive, implying an amplification of SST anomalies. Positive feedbacks are mainly associated with the following atmospheric responses to positive SST anomalies: (1) reduced surface wind speed (positive turbulent heat flux feedback) over the tropical western North Atlantic and Indian Oceans, (2) reduced marine boundary layer stratocumulus cloud fraction (positive shortwave radiative flux feedback) over the Namibian stratocumulus deck, and (3) enhanced atmospheric water vapor (positive longwave radiative flux feedback) in the vicinity of tropical deep convection region over the Indian Ocean that exceeds the negative shortwave radiative flux feedback associated with enhanced cloudiness.

By combining a stochastically-forced entraining ocean mixed layer model with the ISCCP satellite-derived radiation data, Park is trying to understand how marine stratocumulus (MSC) feed back on the climate system through air-sea interaction over the midlatitude North Pacific Ocean in summer. A preliminary analysis indicates that MSC significantly increase year-to-year persistence of the SST anomalies during summer in the midlatitude North Pacific. The relative contributions of the other components, including turbulent heat flux feedback, remote ENSO forcing, and oceanic entrainment (or reemergence) on the SST persistence are also being examined to assess the importance of the MSC-SST feedback.

Meehl, in collaboration with Harry van Loon (Colorado Research Associates), and Julie Arblaster, diagnosed decadal timescale variability in the latter part of the 20th century from NCEP/NCAR reanalyses and gridded precipitation from Climate Modeling and Analysis Program, and found consistency with an earlier modeling study in that the signal of increased solar forcing on the timescale of the 11 year solar cycle involves an intensification of the climatological precipitation maxima in the tropics, as well as an alteration of the vertical temperature structure also consistent with the earlier modeling study.

The Water Cycle

Aiguo Dai (CAS), along with Trenberth and Taotao Qian (CAS), have derived a monthly dataset of PDSI from 1870 to 2003 using historical precipitation and temperature data for global land areas on a 2.5° grid. Over Illinois, Mongolia, parts of China, and the former Soviet Union where soil moisture data are available, the PDSI is significantly correlated (r = 0.5 to 0.7) with observed soil moisture content within the top 1 m depth during warm season months. The strongest correlation is in late summer and autumn, and the weakest correlation is in spring when snowmelt plays an important role. Basin-averaged annual PDSI co-vary closely with streamflow. The results suggest that the PDSI is a good proxy of both surface moisture conditions and streamflow. An EOF analysis of the PDSI reveals a fairly linear trend resulting from trends in precipitation and surface temperature and an ENSO-induced mode of mostly interannual variations as the two leading patterns. The global very dry areas, defined as PDSI < –3.0, have more than doubled since the 1970s, with a large jump in the early 1980s due to an ENSO-induced precipitation decrease and subsequent expansion primarily due to surface warming while global very wet areas (PDSI > +3.0) declined slightly during the 1980s. Together, the global land areas in either very dry or very wet conditions have increased from 20% to 38% since 1972, with surface warming as the primary cause after the mid-1980s. These results provide observational evidence for increasing risk of droughts as anthropogenic global warming progresses and produces both increased temperatures and increased drying.

This figure shows leading modes of PDSI variations correspond to a trend and ENSO. Dai et al (2004).

Dai led a team of prominent researchers, including Peter Lamb (University of Oklahoma), Trenberth, Mike Hulme (University of East Anglia, United Kingdom), Phil Jones (University of East Anglia, England), and Pingping Xie (Climate Prediction Center, NCEP/NOAA), to rebut erroneous published conclusions concerning the reality of the drought in the Sahel. Using rainfall data from a fixed station network, they show that decreasing rainfall trends are widespread in the Sahel during the later part of the 20th century, and that the decreasing trend in Sahel regional rainfall is not an artifact of changing station networks. They show that the rainfall model used in Chappell and Agnew is incorrect, their modeled rainfall time series is totally unrepresentative of a Sahel rainfall average, their conclusion about the Sahel rainfall trends being an artifact of changing station locations is emphatically wrong, and their speculative statements about the implications of their results for other studies and other regions of the world are completely unfounded.

Trenberth continues work on manifestations of global climate change on accelerating the hydrological cycle and the prospects for increasing extremes. The issue of how the hydrological cycle will change as the climate changes is complicated considerably by somewhat independent changes in pollution (aerosol), whereby increases typically act to short-circuit the hydrological cycle. However, with warming, the main prospect is for increased water holding capacity and associated increased water vapor in the atmosphere. The rate of increase of 7% K-1 is much greater than projected rates of increase in overall evaporation and precipitation, which are more like 1–2% K-1 and are governed by the surface heat budget. Hence, the prospects are for increases in precipitation intensity but decreases in duration or frequency of precipitation. In addition, increased rain at expense of snowfall is likely, with major impacts on snow pack (storage) and soil moisture. The changes directly impact partitioning into soil moisture and runoff, and make prospects of both droughts and floods more likely.

Dai and Trenberth have used the estimated river discharge data and the E-P fields derived from the ECMWF and NCEP/NCAR reanalyses to derive new estimates of meridional transport of fresh water by the oceans. Preliminary results show improved agreement with in situ data based on direct estimates. Compared with earlier indirect estimates, the new estimates show increased southward transport in the Atlantic Ocean and increased northward transport in the South Pacific. However, the E-P fields are being evaluated.

Dai, together with Qian and Trenberth, has started to estimate the global continental fresh water discharge since 1950 by combining historical records of streamflow and model-simulated surface runoff (forced by reanalysis and station data). The Community Land Model (CLM) has been run offline forced by observation-based 3-hourly surface data (precipitation, surface solar radiation, surface air temperature, wind speed, and specific humidity). Surface observations of cloud cover were used to correct the erroneous trends in the reanalysis surface solar radiation. The satellite-based 3-hourly estimates of surface solar radiation for 1983–2001 from ISCCP were used to adjust the biases in the reanalysis solar data. The 6-hourly mean precipitation rates from the reanalysis were adjusted to represent realistic combination of frequency and intensity using observed precipitation frequency maps. The CLM was first run hundreds of years using recycled forced data to get a stable state for deep soil water. The historical CLM simulations (from 1948–2002), together with available observational datasets, will be used to study the changes and variations in terrestrial water fluxes (evaporation, precipitation, runoff, streamflow, and continental discharge) and surface moisture conditions (soil moisture contents, surface humidity, drought indices) during the last 50 years, especially since the mid-1970s when rapid global warming has occurred.

Dai collaborated with X.-Z. Liang and others at the Illinois State Water Survey and University of Illinois at Urbana-Champaign in using the MM5-based regional climate model (CMM5) to study the diurnal cycle of United States summer precipitation. They found that the model performance is very sensitive to the choice of cumulus parameterization schemes whose skills are highly regime selective. The Grell scheme realistically simulates the nocturnal precipitation maxima and the associated eastward propagation of convective systems over the Great Plains, where the diurnal timing of convection is controlled by the large-scale tropospheric forcing; whereas the Kain-Fritsch scheme is more accurate for the late afternoon peaks in the southeast U.S. where moist convection is governed by the near-surface forcing. In radar rainfall data and the simulation with the Grell scheme, another weaker eastward-propagating diurnal signal is evident from the Appalachians to the east coast. The result demonstrated the importance of cumulus schemes and provided a realistic simulation of the central United States nocturnal precipitation maxima.

El Niño-Southern Oscillation and Other Tropical Studies

Fasullo examined biennial characteristics of Indian monsoon rainfall to explore the coincidence of ENSO with the near 3-year peak the all-India rainfall spectrum. He found the oft-cited, weak-to-strong transitions in monsoon rainfall to be uniquely associated with El Niño-La Niña transitions, and showed the 3-year peak in rainfall’s power to coincide with the ENSO spectrum. The temporal and spatial structure of biennial power within India was shown to mirror the strength of the ENSO-monsoon teleconnection. Statistically robust and previously undocumented occurrences of biennial persistence were also identified. The role of an independent Indian Ocean mode in instigating biennial fluctuations in the monsoon is therefore called into question.

Dai’s work in collaboration with Adam Monahan (University of Victoria, Canada), has been completed and published. They examined the spatial and temporal structures of ENSO nonlinearity in several observed and model datasets, and found that the anomaly spatial pattern that changes sign between El Niño and La Niña events (the "linear" signal) strongly resembled that of the first EOF, while that which does not change sign (the “nonlinear” signal) resembled the pattern of the second EOF. Of several coupled models considered, the spatial structure of the El Niño/La Niña asymmetry was partly captured only by the GFDL R30 model.

Deser, together with Phillips and Hurrell, published a comprehensive study on the observed linkages between multi-decadal variability of the Aleutian Low in wintertime and conditions in the tropical Indo-Pacific during the last ~100 years. Their results provide strong evidence that the tropical Indo-Pacific Ocean is a key driver of North Pacific climate change on multi-decadal time scales. Modeling experiments are underway to address the mechanisms for this linkage.

This figure shows (left) epoch difference maps of winter SLP (contours) and land precipitation (color shading) for high-low North Pacific Index (NPI) regimes. The right shows the North Pacific 1899–2002: -NPI; winter precipitation difference coastal Alaska-Japan; winter surface air temperature NW Canada and Alaska; and EOF1 of winter-spring monthly SST anomalies. Deser et al. (2004).

In related work, Deser, together with Rosanne D’Arrigo (Lamont-Doherty Earth Observatory, Tree-Ring Laboratory) and other co-authors, completed an analysis that extends the observational work to paleoclimate records covering nearly four centuries. Specifically, they analyzed a suite of tree-ring chronologies from land areas surrounding the North Pacific Rim and a suite of coral records from the tropical Indo-Pacific Ocean to assess their degree of association on decadal time scales. The results show that the strength of association between the low and high latitude proxy records varies over time, with the 20th century showing the most coherent linkages.

Park, Deser, and Alexander, are completing a study of the influence of ENSO and local cloud-SST feedbacks upon SST anomaly persistence in the North Pacific. This work uses new estimates of surface heat flux feedbacks (see Climate Diagnostics), and a simple entraining stochastic climate model to investigate why summer SST anomalies in the North Pacific are so persistent despite their weak thermal inertia due to a shallow mixed layer.

The North Atlantic

Deser, with Ramalingam Saravanan (CDP) and Gudrun Magnusdottir (University of California at Irvine), published a study based on a suite of experiments using the CCM3 to investigate the relative sensitivity of the model's atmospheric circulation to observed trends over the past 50 years in winter sea ice extent and SST over the North Atlantic/arctic sector. They found a stronger sensitivity to the sea ice anomalies than SST anomalies. The response is a weak negative feedback on the NAO. The total response resembled closely the leading pattern of internal circulation variability in the model, although an additional “direct” near-local response was also apparent.

In a follow-up study Deser, together with Shiling Peng (CDC/NOAA), is using a simple modeling framework to gain insight into the dynamical processes responsible for the hemispheric atmospheric circulation response to sea ice and SST anomalies seen in the CCM3 experiments described above.

In related work Deser, together with Robert Tomas (CCR and CAS), is conducting a 100-member ensemble of short (4-month) integrations with the same SST/sea ice forcing and saving the output daily. Each ensemble member begins from a different initial condition on December 1, selected from a long control run, and is paired with a control integration that begins from the same initial condition but has no sea ice/SST forcing. These integrations should shed light on the evolution of the response to its equilibrium structure and on the dynamical processes at work.

Christophe Cassou (CAS), together with Deser and Hurrell, published two studies on the NAO. The first documented the spatial asymmetry of the two phases of the NAO and their associations with tropical and North Atlantic SSTs. The second examined the role of summer SST anomalies in the North Atlantic and their impact upon the following early-winter NAO. They found a statistical association in the observational data for an impact via a tropical atmospheric bridge. The physical basis for this empirical association was examined by means of modeling experiments using CCM3.

Hurrell, along with Martin Hoerling (CDC/NOAA), Phillips, and Taiyi Xu (CDC/NOAA), completed two papers examining the role of tropical forcing in boreal winter North Atlantic climate change over the last half of the 20th century. In the first paper diagnoses of ensembles of atmospheric general circulation model (AGCM) experiments demonstrate that the observed upward trend in the winter NAO index since 1950 is a virtually deterministic response to the temporal history of SSTs and that tropical SST forcing is of primary importance. In the second paper it is further established through idealized SST experiments that progressive warming of the Indian Ocean since 1950 is the principal contributor to the NAO changes. Moreover, it is demonstrated from coupled model experiments that the Indian Ocean warming likely has a strong anthropogenic component.

This figure shows trends in the NAO stem from trends in tropical sea surface temperatures and their effects on the atmosphere via precipitation and latent heat release. Hurrell et al. (2004)

Hurrell, Hoerling, and Phillips are nearing completion of a new study that examines the nature and causes for the regional downward trajectories of both northern and southern African monsoon rainfall over the latter half of the 20th century. Used are ensembles of five different AGCMs forced by the observed evolution of the global oceans since 1950, as well as control simulations with fully coupled ocean-atmosphere CCSM. One main finding is that the observed drying trends during both monsoon seasons fall within the distribution function of the simulated trends drawn from the AGCM experiments. They are not, however, consistent with the probability distribution function (PDF) of 50-year trends occurring from the unforced CCSM. In fact the PDFs of 50-year trends from these two different simulation datasets are almost mutually exclusive. This suggests that the observed downward trajectories of African monsoon rainfall since 1950 have been intimately controlled by the observed trajectory of global SSTs, though the responsible air-sea interactions may not be those occurring from natural variations alone. The possibility cannot be dismissed, however, that the simulated natural variability of the CCSM is too weak. Analysis of long integrations using other coupled models is in progress to assess robustness.

Hurrell, Phillips, Cassou, Chris Folland (Hadley Centre, United Kindom Meteorological Office), and Simon Brown (Hadley Centre, United Kingdom Meteorological Office) have completed their analysis of an ensemble of numerical experiments designed to assess the limitations of forcing AGCM with observed SSTs. The work primarily examines how coupling affects the variance and persistence of extratropical atmospheric properties, but it was extended to examine in more detail the effects of interactive SSTs on the tropical climate as well.

Hurrell and Folland continue their examination of the annual cycle of climate and climate change over the Atlantic investigating the mechanisms responsible for the variability through analyses of both observed and climate model data. They find a significant change in the summer European climate, whereby a summer season of higher-than-average surface pressure over northern Europe is accompanied by reduced rainfall over the tropical North Atlantic and North Africa. Based on analysis of variance techniques that separate climate variability into forced (i.e., due to SST variations) and unforced (i.e., due to internal atmospheric dynamics) components, their results suggest the observed low-frequency extratropical changes in summer climate arise indirectly from processes that affect tropical Atlantic precipitation on long time scales, such as the inter-hemispheric gradient in tropical Atlantic SST.

Hurrell and Mark Rodwell (ECMWF), a scientific visitor to CAS over the summer of 2003, published their work examining the response in five different atmospheric general circulation models (GCM) (including CAM) to realistic, optimally chosen North Atlantic SST anomalies. Together with Marie Drevillon (CERFACS), Claude Frankignoul (Universite Pierre et Marie Curie), Holger Pohlmann (Max Planck Institute), Martin Stendel (Danish Meteorological Institute), and Rowan Sutton (Reading University), they illustrate a consistent response among the models that is similar to observational estimates.

Hurrell has published several articles dealing with climate variability and ecology. He also served as a co-editor of a 2004 Oxford University Press book entitled Marine Ecosystems and Climate Variation with Nils Chr. Stenseth (University of Oslo), Geir Ottersen (Institute for Marine Research, Bergen Norway), and Andrea Belgrano (University of New Mexico). With Ottersen and Stenseth, Hurrell wrote the book’s introductory chapter that overviews the main currents and hydrography of the Atlantic Ocean and how the Atlantic is affected by large-scale variations in leading patterns of atmospheric variability. The chapter also provides an overview of oceanographic processes believed to be of particular importance to marine ecology, and it concludes with an account of how diverse the responses of ecology are to coupled atmosphere-ocean climate variability. In the same book Hurrell and Robert Dickson (Centre for Environment, Fisheries and Aquaculture Science) published a chapter that focused on the NAO and its forcing of the North Atlantic Ocean.

With Stenseth, Ottersen, Atle Mysterud (University of Olso), Mauricio Lima (Center for Advanced Studies in Ecology and Biodiversity, Chile), Kung-Sik Chan (University of Iowa), Nigel Yoccoz (Division of Arctic Ecology, Norwegian Institude for Nature Research), and Bjørn A°dlandsvik (Institute of Marine Research, Bergen Norway), Hurrell published an overview of climate patterns within the context of the ecological effects of climate variability by making use of climate indices. Whereas the considerable influence of both the ENSO and the NAO on ecological processes has been demonstrated, several other large-scale climate patterns are also of ecological interest. Both the advantages and disadvantages of using climate indices, which by definition reduce complex space and time variability into simple measures, in ecological studies are discussed.

Hurrell, in collaboration with Stenseth, Chan, Amir Shabbar (Meteorological Service of Canada), Stan Boutin (University of Alberta), Eli Rueness, Dorothee Ehrich, Kjetill Jakobsen, and Ole Chr. Lingjærde (all with the University of Oslo), studied how the dynamics of Canadian lynx (Lynx canadensis) abundance are geographically structured according to the influence of large-scale climatic regimes. This structuring matches zones of differential snow conditions, in particular surface hardness, as determined by the frequency of winter warm spells. Through a modified functional response curve it is shown that various features of the snow may influence lynx interaction with its main prey species, the snowshoe hare (Lepus americanus): softer snow reduces the killing capacity of lynx. This study highlights the importance of snow, and exemplifies how large-scale climatic fluctuations can – mechanistically – influence population biological patterns.

Decadal Variability

Deser, together with Shoshiro Minobe (Hokkaido University, Japan), Niklas Schneider (University of Hawaii), and four others, completed a review paper on Pacific decadal variability. This study, solicited by the International CLIVAR Program, summarizes the observed features, proposed mechanisms and current state of predictability of Pacific climate variability on decadal time scales.

Deser, together with Antonietta Capotondi (CDC/NOAA), Alexander, and Arthur Miller (Scripps Institution of Oceanography), completed a study on low-frequency pycnocline variability in the Gulf of Alaska using output from the NCAR CSM Ocean Model (NCOM) forced with time-varying atmospheric conditions during 1958–1997. They found that a large fraction of the pycnocline variability can be explained by local Ekman pumping due to local wind stress curl anomalies, while wave processes also contribute in coastal regions.

Deser, along with Capotondi, Alexander, and Michael McPhaden (Pacific Marine Environmental Laboratory/NOAA), completed a study on the anatomy and decadal evolution of the Pacific subtropical ocean circulation cells using output from the NCOM forced with time-varying atmospheric conditions during 1958–1997. They found that the interior equatorward mass transports of these circulation cells decreases after the mid-1970s, consistent with observations, and that the poleward western boundary current transports increase. The meridional mass transport convergence into the equatorial zone is strongly correlated with SSTs in the central and eastern Pacific.

Dai, in collaboration with Aixue Hu (CCR), Meehl, and Warren Washington (CCR), analyzed the variations of the Atlantic thermohaline circulation (THC) in a 1200-year control run and future climate change simulations using the Parallel Climate Model (PCM). The Atlantic THC intensified during the first century of the control run due to rapid cooling in the upper North Atlantic Ocean, and then became increasingly shallower and weaker due to freshening in surface oceans. Through potential vorticity conservation, the subpolar ocean gyre centered at the Labrador Sea contracted from (expanded to) the east, and the North Atlantic Current shifted southward (northward) as the THC gained (lost) strength and depth. A strong and deep THC leads to increased oceanic heat convergence and thus warming in the northern North Atlantic Ocean, thereby reducing the North Atlantic Deep Water production and eventually weakening the THC, a mechanism that might have contributed to large multi-decadal (~24 year) oscillations in the THC’s strength. In the greenhouse gas forced simulations, the THC weakens during the 21st century. In the 22nd century the THC continues to weaken if CO₂ keeps rising, but stabilizes if the greenhouse gases level off. This THC weakening results from larger warming in the upper and North Atlantic Ocean than the deeper and southern Atlantic basin, while salinity changes are small.

Meehl collaborated with Hu to examine aspects of multi-decadal variability in the PCM as seen in observations. In particular, sustained droughts and “megadroughts" in the Indian monsoon and southwestern United States regions were connected to the dominant pattern of multi-decadal SST variability in the Pacific in model and observations. The mechanism producing such multi-decadal climate variations in Pacific SSTs in the model involves coupled air-sea and tropical-midlatitude processes involving slowly propagating wind-forced ocean Rossby waves near 20°N and 20°S.

Model Evaluations

The study by Dai and Trenberth on the diurnal cycle in the CCSM version 2 (CCSM2), with a focus on precipitation and moist convection, was completed and published. The CCSM2 captures the diurnal amplitude and phase of surface air temperature over land, but over the ocean the amplitude is too small. The CCSM2 overestimates the mean total cloud amount from ~15°S–15°N and over northern mid- and high-latitude land areas in winter, whereas it underestimates the cloud amount by 10–30% in the subtropics and parts of the mid-latitudes. Problems also exist in the marine stratocumulus regions west of the continents. In the CCSM2, warm-season daytime moist convection over land starts about four hours too early compared with observations, and plateaus from 1100 to 1800 LST, in contrast to a sharp peak around 1600–1700 LST in observations.

Dai collaborated with Ying Sun, Susan Solomon, and Robert Portmann of NOAA Aeronomy Laboratory in a study to evaluate current climate models’ ability to simulate precipitation frequency, intensity, and the number of heavy-rain days. They analyzed daily precipitation data from worldwide rain gauges and seven fully coupled climate models. Although the current models are able to simulate the total precipitation amount well, most of them are unable to fully capture the spatial patterns of precipitation frequency and intensity. For light precipitation (1–10 mm day-1), most models overestimate the frequency, but produce patterns of the intensity that are in broad agreement with observations. For heavy precipitation (> 10 mm day-1), most models significantly underestimate the intensity but simulate the frequency well. The average number of the heavy rain days, which accumulate about two thirds of annual rainfall on average, is a simple index capturing the combined effects of frequency and intensity on water supply. Most of the models show large deficiencies in simulating the spatial distribution of the heavy rain days. The results indicate that precipitation variability remains a substantial challenge for most current climate models.

Deser, together with Capotondi, Saravanan, and Phillips, are contributing a paper to the Journal of Climate special issue devoted to the documentation of the CCSM3 and its component models. This study will document aspects of the ENSO phenomenon in CCSM3, including the spatial and temporal patterns of SST, precipitation, sea level pressure, surface wind, and ocean thermocline depth anomalies associated with ENSO over the global tropics. Additionally, the atmospheric teleconnections to the extra-tropics will be documented. The relationship between the temporal scale (e.g., frequency) and meridional scale of ENSO in the coupled model will also be explored within the dynamical framework of the delayed oscillator paradigm. Tropical Atlantic variability independent of ENSO will also be documented.

Young-Oh Kwon (ASP postdoctoral fellow) and Deser are analyzing Pacific decadal variability in the atmosphere and ocean from an 800-year segment of the CCSM2 control integration. Prominent spectral peaks at ~16 years and ~40 years are evident in oceanic variables such as SST and subsurface temperature throughout the North Pacific Ocean. The temporal and spatial structure of these decadal oscillations and the nature of the upper ocean heat budget during a composite decadal cycle are being evaluated to gain insight into the mechanisms responsible for the variability.

Meehl collaborated with Claudia Tebaldi (Research Applications Program) and Doug Nychka (GSP) to diagnose patterns of frost days and heat waves in the 20th century related to processes that produce these phenomena. They then related these processes to possible future changes. They found that the observed trend in frost days in the latter part of the 20th century (greater decreases of frost days in the western United States compared to the eastern United States) was most directly associated with changes in atmospheric circulation over that time period. Similarly, the observed pattern of heat waves over the United States and Europe in the latter part of the 20th century, derived from NCEP/NCAR reanalysis data, was related to the climatological atmospheric circulation, with greater heat wave occurrence over the southwestern United States, southern and upper midwest, and eastern seaboard of the United States.

Meehl also collaborated with Ping Liu (University of Hawaii), Bin Wang (University of Hawaii) and Ken Sperber (Lawrence Livermore National Laboratory, LLNL) in a study of the Madden-Julian Oscillation (MJO) in the CAM. In the standard CAM version, MJO variability is very weak. By substituting another convection scheme in the model that is known to better simulate MJO variability, the MJO simulation in this modified version of CAM becomes much more realistic. This shows that the convection scheme is most directly responsible for simulating the MJO in an atmospheric model and suggests future improvements in the CAM.

Hurrell, Phillips, Hack, Caron, and Yin are contributing a new paper to the Journal of Climate special issue devoted to documentation of the CCSM3 and its component models. In particular, these authors are examining the dynamical aspects of CAM in uncoupled mode, including the seasonal variation of its mean state and its intraseasonal and interannual variability. Comparisons are made to not only observations, but also to an earlier version of the CCM3.

Tom Wigley (CAS), in collaboration with Caspar Ammann (CCR), Ben Santer (LLNL), and Sarah Raper (Alfred Wegener Institute for Polar and Marine Research), has considered the effects of volcanic eruptions on climate using both PCM and Model for the Assessment of Greenhouse-gas Induced Climate Change (MAGICC). The results from 16 PCM simulations are used to reduce internally-generated noise and obtain an improved estimate of the underlying response of 20th century global-mean temperature to volcanic forcing. MAGICC is then used with the same forcing and the same climate sensitivity as the AOGCM to emulate the AOGCM results. The upwelling-diffusion energy balance model (UD EBM) and AOGCM results are in excellent agreement, justifying the use of the UD EBM to determine the volcanic response for different climate sensitivities. The maximum cooling for any given eruption is shown to depend approximately on the climate sensitivity raised to power 0.37. After the maximum cooling for low-latitude eruptions, the temperature relaxes back towards the initial state with an e-folding time of 29–43 months for sensitivities of 1–4°C. Comparisons of observed and modeled coolings after the eruptions of Agung, El Chichon, and Pinatubo give implied climate sensitivities that are consistent with the IPCC range of 1.5–4.5°C equilibrium warming for 2 x CO₂. The cooling associated with Pinatubo appears to require a sensitivity above the IPCC lower bound of 1.5°C, and none of the observed eruption responses rules out a sensitivity above 4.5°C.

Anthropogenic Climate Change and Detection

Dai, Meehl, Washington, and others analyzed the climate change simulations under the Department of Energy-sponsored Accelerated Climate Prediction Initiative Program using the PCM. An initialization to 1995 ocean conditions removes a large part of the unforced oceanic temperature and salinity drifts that occurred in the standard 20th century integration. The results suggest that the affect of small errors in the oceans (such as those associated with climate drifts) on coupled GCM-simulated climate changes may be negligible.

Article 2 of the United Nations Framework Convention on Climate Change has as its objective the stabilization of greenhouse gas concentrations in the atmosphere. The choice of a CO₂ concentration stabilization target, considered by Wigley, depends on the following: what is considered to be ‘dangerous anthropogenic interference with the climate system’; the forcings that might arise from non-CO₂ gases; and the climate sensitivity. These three factors are specified probabilistically, as probability density functions, and combined to produce a PDF for the CO₂ concentration target. There is a probability of 17% that the stabilization target should be less than the present level, and the median target is 536 ppm. The effects of reducing the emissions of non-CO₂ gases and/or implementing adaptation strategies are considered probabilistically and shown to alter these figures significantly.

This work has been extended by Wigley in collaboration with Richard Richels (Electric Power Research Institute) and Alan Manne (Stanford University) to consider the issue of climate stabilization rather than concentration stabilization. Thermal inertia of the oceans requires concentrations to decrease after a global-mean temperature stabilization point has been reached in order to maintain a stable climate. Focusing on temperature change is judged to be more meaningful than focusing on changes in atmospheric composition. The analysis explicitly incorporates several uncertainties critical to future temperature change: economic growth, climate sensitivity, and the rate of ocean heat uptake. Concentration ceilings necessary to limit temperature change to 2°C and 3°C are considered. The influence on investments in technology is addressed in the analysis. Differences in the assumed technological future that correspond to billions of dollars in investment costs can translate into trillions of dollars worldwide in terms of mitigation costs over the 21st century.

As an application of MAGICC, Wigley has considered the issue of the climate change commitment. Due to oceanic thermal inertia, even if the composition of the atmosphere were fixed at present levels, global-mean warming and sea level rise would continue. These constant-composition (CC) commitments and their uncertainties are quantified. The constant emissions (CE) commitment is also considered. The eventual CC warming commitment lies between 0.2°C and more than 1°C. The CE warming commitment is much larger (2–6°C by 2400). The CC sea level rise commitment ranges from negligible to more than 30 cm/century (central estimate, 10 cm/century). The corresponding CE commitment ranges from 6 cm/century to more than 50 cm/century (central estimate, 25 cm/century). Avoiding these changes requires, eventually, a reduction in emissions to substantially below present levels. For sea level rise a substantial long-term commitment may be impossible to avoid.

To investigate sea level rise over many centuries under CO₂ concentration stabilization scenarios, Wigley and Raper have developed a new Glacier and Small Ice Cap (GSIC) model. To justify the need for such a model, they show that the corresponding model used in the IPCC Third Assessment Report (TAR), if applied to times beyond 2100, imposes an unrealistic upper bound on GSIC melt. A modification to the TAR model is introduced that allows it to be extended beyond 2100 with asymptotic melt equal to the initially available ice volume (V0). The modification has a negligible effect on the original TAR formulation out to 2100 and provides support for the IPCC method over this time period. The sensitivity of GSIC melt to uncertainties in V0 and mass balance sensitivity is examined, and melt results are given for a range of CO₂ concentration stabilization cases. Approximately 73–94% of GSIC ice is lost by 2400.

The issue of natural climate variability is addressed by Peter Foukal (Heliophysics, Inc.), Gerald North (Texas A&M University), and Wigley. They assess the likelihood of significant low-frequency solar irradiance changes using recent astronomical evidence. Recent reconstructions of solar irradiance extending back to the 17th century have relied mainly on reconstructions of relatively large, multi-decadal irradiance variations based on photometry of other Sun-like stars. In fact, very few true Sun-like stars have been identified, so the basis for these reconstructions is suspect. If these speculative low-frequency irradiance variations are ignored, the fraction of past climate changes attributable to solar variations is much reduced, with important implications for our understanding of features like the Little Ice Age and the early 20th century warming.

In collaboration with Santer, Wigley has produced a review of ‘fingerprint’ (i.e., pattern correlation) method for detection and attribution. The work uses tropopause height as a detection variable to illustrate the ‘constant pattern’ approach pioneered by Klaus Hasselmann (Max-Planck-Institut fur Meteorologie). Observed changes from the NCEP/NCAR and ERA-15 reanalyses are compared with fingerprints for the sum of anthropogenic and natural (solar plus volcanic) forcing derived from the NCAR/DOE PCM. Both ‘raw’ and ‘optimized’ fingerprints are considered, and full spatial patterns, and patterns with global-mean changes removed are considered. Observed-model agreements are always strong and highly statistically significant. The primary causes for decadal time scale tropopause height changes are increases in well-mixed greenhouse gases (that warm the troposphere) and stratospheric ozone depletion (that cools the stratosphere). Both factors are required to explain the observed height changes.

Unravelling the causes of recent climate change is also the focus of a paper by Santer, Wigley and 16 other authors from a number of institutions (including Ammann, Meehl and Washington). This work examines changes in the height of the tropopause using the ERA-40 reanalysis, and uses optimized pattern correlation methods to identify the fingerprint of human-induced changes in the observational record. Over 1979–2001, the height of the tropopause increased in the ERA-40 data by nearly 200 m. The spatial pattern of height increases is consistent with predictions from PCM for the response to anthropogenic forcing factors and cannot be explained by natural variability alone. ERA-40 temperature data for the troposphere and the stratosphere are shown to agree well with satellite-based microwave sounding unit data from two independent sources. The results support the reality of significant human-induced temperature changes in both the troposphere and the stratosphere, and confirm earlier results that changes in regions have contributed to the observed tropopause height increase.

Wigley, in collaboration with Steve Smith (Joint Global Change Research Institute, JGCRI) and Hugh Pitcher (JGCRI), has developed new projections for future SO₂ emissions using the MiniCAM integrated assessment model. An income-based parameterization for future SO₂ emissions controls is developed using purchasing power parity (PPP) income estimates and their relation to historical controls. As a result of increases in both concerns about pollution and the economic ability to institute emissions controls for SO₂ that accompany increasing per capita wealth, SO₂ emissions for almost all of the IPCC Special Report on Emissions Scenarios (SRES) scenarios eventually decline over the next century to levels substantially below those of today – except in scenarios where future incomes in developing nations increase slowly. Under a climate policy that limits CO2 emissions, SO₂ emissions lie in a relatively narrow range, in part because of their common fossil-fuel sources.

Also in collaboration with Smith, Wigley has examined the role of climate forcing agents other than carbon dioxide using the MiniCAM integrated assessment model for no-climate-policy and policy emissions scenarios. Non-CO₂greenhouse-gas forcing is dominated by methane and tropospheric ozone. Assumptions about the prevalence of methane recovery and local air pollution controls in the no-policy cases are a critical determinant of methane and ozone-precursor emissions. When these factors are considered, emissions of these gases are reduced substantially relative to earlier estimates, such as those in the SRES. Climate change predictions based on at least some of the SRES scenarios are therefore likely to be overestimates. The indirect influence of mitigation policies on SO₂ emissions is also considered. While aerosols have only relatively small effects on climate by the end of the century, there is a significant interaction in the early 21st century between policies to reduce CO₂ emissions and SO₂ emissions, even in the presence of SO₂-related pollution control policies. The attendant reduced aerosol cooling more than offsets the reduction in warming that accrues from reduced CO₂.

This figure shows tropopause height changes: Model (PCM) vs observed from ERA-40 after 1958. Means are forced to be the same 1958–1999. Bold lines are low pass filtered. The orange and blue envelopes are measures of the variability. The model, with all forcings, follows the analyses well, although volcanic events are overdone. (Santer, Wigley et al. 2004)

General Overviews

Trenberth, along with Berrien Moore (University of New Hampshire), Tom Karl (National Climatic Data Center/NOAA), and Carlos Nobre (Centro de Pronostico del Teimpo y Estudios Climatico, del Instituto Nacional de Investigaciones Espaciales), has put forward a future perspective on monitoring and prediction of the Earth’s climate as a keynote talk at the International CLIVAR conference. They point out that, not only is the climate changing and will continue to do so regardless of any mitigation actions, but observations and information available are also changing as technological advances take place. Accordingly, they highlight the need for a climate information system that embraces a comprehensive observing system to observe and track changes and the forcings of the system as they occur, and which develops the ability to relate one to the other and understand changes and their origins. Observations need to be taken in ways that satisfy the climate monitoring principles and ensure long-term continuity, and which have the ability to discern small but persistent signals. They propose some benchmark observations to anchor space-based observations and trends, including a much-needed step forward in the quality of water vapor observations. The health of the monitoring system must be tracked and resources identified to fix problems. Fields must be analyzed into global products and delivered to users while stakeholder needs are fully considered. Data should be appropriately archived with full and open access, along with metadata that fully describe the observing system status and environment in which it operates. Reanalysis of the records must be institutionalized along with continual assessment of impacts of new observing and analysis systems. Some products will be used to validate and improve models, as well as initialize models and predict future evolution on multiple time scales using ensembles. Attribution of changes to causes is essential, and it is vital to fully assess past changes and model performance and results in making predictions to help appraise reliability and assess impacts regionally on the environment, human activities, and sectors of the economy. They claim that such a system will be invaluable and further provides a framework for priority setting new observations and related activities.

Trenberth, with Phil Arkin and Eugenia Kalnay (University of Maryland), James Laver (Climate Prediction Center/NOAA), and Siegfried Schubert (Goddard Space Flight Center/NASA), wrote up results of a workshop on “Ongoing analysis of the climate system.” They point out that one essential aspect of a comprehensive climate observing system is the capability to synthesize observations into a coherent, internally consistent depiction or analysis of the evolution and present state of global climate. To further the planning that will ensure such a national capacity, the workshop was held involving about 65 scientists and agency managers. The workshop provided guidance on the steps needed to ensure that ongoing developmental ocean, atmosphere, and land surface data assimilation/reanalysis efforts remain complementary, and identified near-term, high priority actions required for future atmospheric reanalyses that will: 1) deal more effectively with the changing climate observing system and with uncertainties in analyzed fields; 2) improve the description of atmospheric interactions with the land, ocean and cryosphere; and 3) improve the description of the hydrological cycle.

Trenberth reviewed the role of the ocean in climate as the opening keynote introduction to NOAA’s new Annual Report of the State of the Ocean and Ocean Observing System for Climate. He discussed the importance of the ocean being wet, having large heat capacity, its geographic distribution, its role in the hydrological cycle, the role of currents in transporting heat, fresh water, and chemicals, the thermohaline circulation, the annual cycle, sea ice, coupled ocean-atmosphere interactions, and sea level changes. Ultimately, he provided a comprehensive justification for better observations of the ocean.

Trenberth, John Overpeck (University of Arizona) and Susan Solomon (Aeronomy Lab, NOAA) wrote a full report and a summary for Eos on a workshop on “Exploring drought and its implications for the future.” The meeting was titled more fully “A Multi-millennia Perspective on Drought and Implications for the Future” and was organized on behalf of the international CLIVAR-PAGES joint working group. The meeting was attended by approximately 70 people and brought a focus of new ideas, observations, analyses, and theories about drought to improve understanding, analysis approaches, and predictive capabilities. The main focus regionally was on North America and Northern Africa, the two regions with the largest amount of available, drought-related paleo-data and research, as well as serious ongoing droughts. For a full report see http://ipcc-wg1.ucar.edu/meeting/wg1/Drght

Trenberth points out that, while urbanization and other changes in land use have an impact on surface air temperatures, the Kalnay and Cai Nature report claim that urbanization and land-use change have a major effect on the climate in the United States is flawed. They used surface temperatures obtained from NCEP/NCAR 50-year reanalyses (NNR), and their difference compared with observed station surface temperatures as the basis for their conclusions on the grounds that the NNR did not include these anthropogenic effects. However, the NNR also overlooked other factors, such as known changes in clouds and in surface moisture, which are more likely to explain Kalnay and Cai’s findings. Although urban heat-island effects are real in cities, direct estimates of the effects of rural land use change indicate a cooling rather than a warming influence that is due to a greater reflection of sunlight.

Karl and Trenberth reviewed “Modern global climate change” as part of Science’ State of the Planet series in December 2003. They point out that modern climate change is dominated by human influences, which are now large enough to exceed the bounds of natural variability. The main source of global climate change is from human-induced perturbations of the composition of the atmosphere primarily resulting from emissions associated with energy use, but on local and regional scales urbanization and land use changes are also important. Although there has been recent progress in monitoring and understanding climate change, there remain many scientific, technical, and institutional impediments for precise planning for, adapting to, and mitigating the effects of climate change. There still exists considerable uncertainty about the rates of change that can be expected, but it is clear that these changes will be increasingly manifested in important and tangible ways, such as changes in extremes of temperature and precipitation, decreases in seasonal and perennial snow and ice extent, and sea level rise. Anthropogenic climate change is now likely to continue for many centuries into the future. They suggest that we are venturing into the unknown with climate and its associated impacts could be quite disruptive.

Wigley has produced a review of simple climate models, from one-box energy balance models (EBMs) to multi-box upwelling-diffusion (UD) EBMs. The latter is illustrated using the MAGICC mode (downloadable from www.cgd.ucar.edu), which couples a UD EBM to a range of gas-cycle models to investigate future climate change due to anthropogenic emissions of greenhouse gases and aerosol precursors. Because of their computational efficiency, simple climate models are valuable when a large number of climate simulations are required, as is the case for probabilistic projections of future warming. Some examples of probabilistic projections are given in this review.

Wigley has compared future climate change under no-policy and policy emissions pathways. Future emissions under the full range of SRES scenarios are used as examples of no-climate-policy scenarios, while the WRE CO₂ concentration profiles are given as examples of stabilization pathways. An alternative ‘overshoot’ pathway is introduced where concentrations initially rise above the eventual stabilization level. Probabilistic projections (as PDFs) for global-mean temperature under the SRES scenarios are given. The relative importance of different sources of uncertainty is determined by removing individual sources of uncertainty and examining the change in the output temperature PDF. Emissions and climate sensitivity uncertainties dominate, while carbon cycle, aerosol forcing, and ocean mixing uncertainties are shown to have only small effects. Uncertainties in regional climate change are also considered by comparing normalized changes (i.e., changes per 1°C global-mean warming) across multiple models and using the inter-model standard deviation as an uncertainty metric.

Wigley has also produced an extensive review of input needs for downscaling of climate data for the California Energy Commission. Downscaling refers to the production of high spatial resolution information from coarser-resolution AOGCM output. Statistical and dynamical (i.e., the use of regional climate models) methods, and their relative strengths and weaknesses are reviewed. The analysis concentrates on the western USA.