Restructuring Federal Climate Research to Meet the Challenges of Climate Change (2009)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Appendix E Research Priorities for Improving Our Understanding of the Natural Climate System and Climate Change Antonio J. Busalacchi, University of Maryland Ian Kraucunas, National Research Council Note: The committee commissioned the following discus- sion paper from the staff and chair of the National Research Council Climate Research Committee. Their views, as expressed below, may not always reflect the views of their committee, the Committee on Strategic Ad- vice on the U.S. Climate Change Science Program, or vice versa. BACKGROUND The National Academies’ Committee on Strategic Advice on the U.S. Climate Change Science Program (CCSP Advisory Committee) is charged to “examine the program elements de- scribed in the Climate Change Science Program strategic plan and identify priorities to guide the future evolution of the program in the context of established scientific and societal objectives.” These priorities may include “adjustments to the balance of science and applications, shifts in emphasis given to the various scientific themes, and identification of program elements not supported in 203 

204 APPENDIX E the past.” To help develop its response to this charge, the CCSP Advisory Committee has requested input on: 1. Top priorities that have been identified by the Climate Re- search Committee (CRC) and elsewhere that focus on understanding the climate system and that take into consideration science require- ments from stakeholders 2. Sources/references for these priorities and criteria for se- lecting them 3. Items in the existing CCSP strategic plan that could be deemphasized This discussion paper identifies 15 priorities, with an emphasis on improving our scientific understanding of the natural climate system. To develop this list of priorities, we reviewed the docu- ments listed in the “Context” section below, solicited input from members of the CRC and its parent Board on Atmospheric Sci- ences and Climate (BASC), and used an informal set of criteria, listed in the “Selection Criteria” section, to select 15 priorities from among the many ideas collected. We identified two overarch- ing priorities, which emerged as the most critical issues facing the CCSP from the perspective of natural sciences research, then cate- gorized the remaining 13 as either existing priorities (i.e., those already reflected in the 2003 CCSP strategic plan), emerging pri- orities (those that have surfaced or increased in importance during the past 5 years), or crosscutting (infrastructural and organiza- tional) priorities. Although an attempt was made to cover the full range of ac- tivities needed to facilitate progress in understanding the physical basis of climate change and to support climate-related decision making, this appendix does not attempt to provide a comprehen- sive review of the priorities listed in other documents (e.g., the recently proposed draft revisions to the CCSP strategic plan), and it is likely that some important priority areas have been over- looked. Also, this appendix focuses on priorities related to the natural (physical-biogeochemical) climate system; a companion paper (Appendix D) discusses priorities for the human dimensions of global change research. The CCSP Advisory Committee will also be considering priorities for climate science applications, 

APPENDIX E 205 which were the focus of a workshop held in October 2007. These collections of ideas are all intended to serve as a starting point for discussions at the CCSP Advisory Committee’s March 2008 work- shop, where that committee will begin developing its final report on future priorities for the CCSP. CONTEXT Our understanding of the climate system and climate change has evolved rapidly over the past several decades. Significant pro- gress has been made in many areas, such as measuring the precise concentrations of different greenhouse gases and determining their impact on Earth’s radiative balance, while other questions have proven more challenging to answer. A number of documents pro- duced by the National Research Council (NRC) and other groups have attempted to assess progress in different areas of climate change science and to identify the critical research advances needed to further improve our understanding of past, current, and projected future climate changes; the impacts of these changes on both human and natural systems; and the infrastructure, organiza- tional structures, and strategic frameworks needed to promote progress. Some of the most important sources consulted during the development of this paper are listed at the end. This document was also informed by discussions held at CRC meetings during the past several years, including • CRC Forum on Seamless Prediction and Year of Tropical Convection, May 17, 2007 • BASC/CRC Forum on IPCC AR4: Key Research Ques- tions and High-Priority Research Needs, May 17, 2007 • CRC Forum on Integrated Earth System Analysis, March 22, 2006 • CRC Forum on Development of an Abrupt Climate Change Early Warning System, December 1, 2006 

206 APPENDIX E SELECTION CRITERIA This paper attempts to distill the findings and recommenda- tions from the reports and other sources listed in the preceding section down to a short list of the most important priorities for en- suring continued progress in understanding the natural climate system and climate change. We selected 15 priorities based on an informal, subjective consideration of the following criteria: • How important is the priority for documenting and under- standing current climate change and its impacts? • How important is the priority for improving predictions and projections of climate variability and future changes in the climate system? • How relevant is the priority to decision support activities, including efforts to develop or evaluate strategies for responding to climate change? • What is the current level of support (within the CCSP, in the United States, and internationally) for progress on the priority, relative to its perceived scientific importance? • Can progress be made in the next 5 years given our current basis of understanding and available or potentially available tech- nology and human resources? We divided our 15 priorities into four categories: overarching priorities, which emerged as the most critically important for en- suring continued progress in climate change research; existing priorities from the CCSP’s 2003 strategic plan, some of which have resulted in significant progress during the past 5 years while others have seen less; emerging priorities, or areas that have sur- faced or increased in importance during the past 5 years; and crosscutting priorities, which involve the infrastructural and or- ganizational frameworks needed to conduct climate research and connect the results to stakeholders. The following section provides a brief description of each priority. Our two overarching priorities are listed first and are viewed as top overall priorities for the CCSP. The remaining priorities are numbered for convenience but are not ranked, either with respect to one another or to priorities related to other aspects of the program. It should also be reiterated 

APPENDIX E 207 that these 15 priorities represent only a partial list of the full spec- trum of valuable and worthwhile activities that could be undertaken to improve scientific progress on climate change, and are intended to serve as a starting point for continued discussion. OVERARCHING PRIORITIES 1. Observations. Long-term, stable, and well-calibrated ob- servations are fundamental to all climate research, prediction, and applications—as noted in the 2003 CCSP strategic plan (CCSP, 2003), “observations of the underlying physical state of the Earth system … are required before questions about climate or global change can be addressed.” Observations are likewise essential to improving our understanding and ability to model the individual components of the climate system, as well as how these elements interact. Hence, in our view, the single most important overall pri- ority for the CCSP is the development of a sustained, integrated, and well-calibrated climate observing system that includes a broad spectrum of in situ and remotely sensed measurements from plat- forms in space, on land, at sea, and in the air. This includes maintaining the continuity of existing observations in order to de- termine anomalies and detect long-term changes, developing new observational capabilities targeting critical gaps, and integrating these elements to ensure and enhance the accuracy and compre- hensiveness of the observing system. The value of stable, homogeneous, long-term climate datasets has been recognized for many years (e.g., NRC, 1995, 1999, 2004a). Some examples of the research advances made possible by the current observing system can be found in the annual updates to Our Changing Planet (e.g., CCSP, 2007a), by comparing the most recent reports by the Intergovernmental Panel on Climate Change (IPCC, 2007) with previous assessments, and also in retrospective reports such as the recently released Scientific Accomplishments of Earth Observations from Space (NRC, 2008). Despite these impor- tant accomplishments, the use of existing observations for climate applications remains hampered by problems with precision, conti- nuity, and spatial and temporal coverage, as well as a dearth of observations for key systems and an overall lack of coordination 

208 APPENDIX E and integration. For example, despite repeated calls for invest- ments in upgrading the surface observations network (e.g., NRC, 1999), climate trends from surface-based climate observations still contain significant errors (IPCC, 2007). More alarming, when the National Polar Orbiting Environmental Satellite System (NPOESS) experienced severe development problems and cost overruns, the climate research and monitoring instruments were demanifested; this and other concerns led the NRC Committee on Earth Sciences and Applications from Space to conclude that the current U.S. envi- ronmental satellite system is “at risk of collapse” (NRC, 2005a). Moreover, these ongoing and in some cases worsening problems have all occurred even though the 2003 CCSP strategic plan ac- knowledged the limitations of observing systems, dedicated an entire chapter to “Observing and Monitoring the Climate System,” and called for the program to “expand observations, monitoring, and data/information system capabilities” (CCSP, 2003). A number of factors have hampered the development of an in- tegrated climate observing system. Climate observations demand dedicated long-term observational campaigns to evaluate climate variability of different timescales and estimate long-term trends. However, many “climate” observations are (or were) originally collected to support operational forecasting or other short-term applications that have less stringent accuracy and calibration re- quirements than climate monitoring and prediction. Maintaining long-term measurements also requires long-term agency attention and funding commitments, yet continuous climate observations unrelated to weather prediction do not appear to be the lead re- sponsibility of any single agency. For example, for space-based platforms, the National Oceanic and Atmospheric Administration (NOAA) has traditionally performed operational data collection while the National Aeronautics and Space Administration (NASA) has conducted research data collection, but the relative role of each agency in maintaining records of suitable quality for climate re- search is less clear; this lack of clear responsibility may have led, in part, to the demanifestation of climate sensors from NPOESS and the Geostationary Operational Environmental Satellite-Series R (NRC, 2007f). Climate-relevant in situ data are managed by an even larger range of federal agencies, state agencies, international organizations, and other groups, creating additional obstacles to 

APPENDIX E 209 the development of an integrated climate observing system (e.g., NRC, 2007c). The Group on Earth Observations framework, a 10- year initiative launched in 2003, is intended to support and encour- age international coordination in the development of an integrated observing system, but its progress to date and prospects for success given current funding levels remain unclear. Without a coordinated, well-funded effort to design and build an integrated climate observing system, the climate record will remain fragmentary and inaccurate, compromising progress in all five CCSP goal areas. As the Climate Research Committee wrote in a letter to Senator Tim Wirth more than 10 years ago, “Without this record, we cannot credibly assess natural climate variations, estimate anthropogenic effects on climate, judge the efficacy of negotiated mitigation efforts, or consider appropriate mid-course policy options” (NRC, 1997). Observations are also critical to every single one of the suggested future priorities that follow; for instance, regional and mesoscale observing networks will be needed to support the development of regional climate models. Thus, in our view, observations should be made the top priority in the next CCSP strategic plan. 2. Regional climate modeling. Given a sound operational foundation in climate observations, a second overarching priority for U.S. climate science that we identified in many of the input documents listed in the “Context” section is improved regional- scale climate change predictions and projections. The regional level is where climate change, land use, and human pollutants in- tersect, and where climate change science connects most directly with decision makers and other stakeholders (e.g., NRC, 2003b). Our ability to model climate change at these scales has simply not kept pace with the increasing demand for this information. In our view, the lack of a national strategy to develop regional climate modeling capabilities is a fundamental shortcoming of the CCSP, and it means that the program has, to date, missed a major oppor- tunity to connect climate change science to the citizenry. Climate change projections from the current generation of global climate models are only reliable on continental-to-global scales, mainly because these models do a poor job of resolving smaller-scale climate variability. Although there have been a num- ber of ad hoc efforts to project regional climate change by a variety 

210 APPENDIX E of groups (e.g., Hay et al., 2002; NCAR, 2005), along with con- tinuing improvements in the resolution of global models, these efforts have been limited by available computing resources (see priority 14, below) and by the lack of a well-developed, agreed- upon framework for regional downscaling, among other factors. There are also a number of important scientific and technical ques- tions to be answered, for example: • What are the potential benefits and drawbacks associated with statistical downscaling, nested models, stretched-grid global climate models, and other regional modeling approaches? • What are the computational, observational, data assimila- tion, and other demands associated with each approach? • What challenges can be expected when going to ultra-high resolution as a means of dealing with subgridscale parameteriza- tions? Given the importance of local-to-regional climate projections to decision makers, the CCSP should, in our view, make improved local and regional climate change prediction and projection a top priority. As with our first overarching priority on observations, regional modeling is a crosscutting effort that would facilitate pro- gress in many of the CCSP goal areas, and it should be noted that these two overarching priorities are complementary and interde- pendent (e.g., observations are needed for model initialization, assimilation, forcing, and validation). The absence of adequate emphasis on regional climate change was also a concern expressed in the reviews of the current CCSP strategic plan (NRC, 2003b, 2004b), as was the coordination of modeling and observations at both regional and larger scales. Thus, we think it is imperative that the next CCSP strategic plan include regional climate modeling as a top priority, and that it contain a viable, integrated strategy for improving both these models and the observations on which they depend to ensure that local and regional decision makers have ac- cess to the information they need to plan for and respond to climate change. 

APPENDIX E 211 EXISTING PRIORITIES 3. Atmospheric distributions and effects of aerosols. The 2003 CCSP strategic plan listed “advance[ing] the understanding of the distribution of all major types of aerosols and their variabil- ity through time, the different contributions of aerosols from human activities, and the processes by which the different contri- butions are linked to global distributions of aerosols” as one of the top three priorities for the Climate Change Research Initiative (CCRI), which was launched in 2001 to “leverage existing U.S. Global Change Research Program (USGCRP) research to address major gaps in understanding climate change.” Over the past 5 years, significant progress has been made in measuring and charac- terizing the abundances and radiative impacts of certain kinds of aerosols, for example, the effects of black soot on snow and ice surfaces. However, the radiative forcing associated with aerosols, especially the so-called indirect effect of aerosols on cloud radia- tive properties, remains the single largest uncertainty in the total global radiative forcing associated with human activities (IPCC, 2007). Aerosols also have an important and complex influence on regional climate forcing, the implications of which are only begin- ning to be appreciated (NRC, 2005b). In light of these remaining uncertainties and the rate of pro- gress made to date, additional investments in the measurement, characterization, and modeling of aerosols could be expected to yield additional progress in understanding how aerosols impact both global and local climate. Encouragingly, the most recent up- date to Our Changing Planet (CCSP, 2007a) lists “understanding aerosol radiative forcing and interactions with clouds” as the first of eight interagency implementation priorities for FY 2008; the key objectives of this priority are “to quantify the effects of atmos- pheric aerosols (tiny airborne particles) on radiation and on clouds, to quantify the modification of the radiation balance by non-CO2 greenhouse gases, and to quantify the influence of the chemistry of the lower atmosphere on both aerosols and non-CO2 greenhouse gases.” All of these objectives are, in our view, important and worthwhile. However, as with all of the other priorities listed be- low, continued progress in understanding aerosols depends on the two overarching priorities above, and will also require both fo- 

212 APPENDIX E cused and deliberate implementation planning and sustained fund- ing commitments. 4. Climate feedbacks and sensitivity. Climate feedbacks, ini- tially focusing on polar feedbacks, and the overall sensitivity of the climate system were the second research priority identified for the CCRI (CCSP, 2003). As with the aerosol priority, considerable progress has been made over the past 5 years in improving under- standing of several key individual feedback processes, including those associated with sea ice and water vapor (e.g., IPCC, 2007). However, the scope of climate processes and research activities that could be considered to fall into the general category of “cli- mate feedbacks and sensitivity” is so broad that it is not clear that this actually constitutes a priority (e.g., many if not most of the activities that fall under CCSP goals 1 and 3 could be considered to fall under this heading). In addition, recent work (Roe and Baker, 2007) suggests that it will remain difficult to rule out the upper end of the probability distribution for the overall sensitivity of the climate system—that is, the possibility that Earth may warm much faster than the current midrange projections for a given future emissions scenario—even with continued progress in understanding climate feedbacks. In our view, the CCSP would be better served to focus on a few key feedback processes that have both global and regional im- portance and that may be amenable to rapid progress. For example, in addition to the interaction between clouds and aerosols dis- cussed in priority 3, above, there is also considerable uncertainty associated with how clouds will respond to rising temperatures (e.g., NRC, 2003c). Current climate models predict, rather than diagnose, cloud properties, and these predictions are now being tested against new data on cloud water content and other new global cloud observations. This should, over time, lead to greater realism in global cloud simulations in climate models. Other im- portant feedback processes that may be amenable to progress include ice sheets and the carbon cycle, both of which are dis- cussed below. 5. Carbon sources, sinks, and feedbacks. Improving under- standing of carbon sources and sinks, with a focus on North America, was the third “science” priority included in the CCRI. The U.S. Carbon Cycle Science Program, which is specifically 

APPENDIX E 213 tasked with making progress in this area, has made significant pro- gress in “clarifying the changes, magnitudes, and distributions of carbon sources and sinks; the fluxes between the major terrestrial, oceanic, and atmospheric carbon reservoirs; and the underlying mechanisms involved including human activities, fossil-fuel emis- sions, land use, and climate” (CCSP, 2007a). Many climate models are also now starting to include an explicit carbon cycle, improv- ing the realism of climate change projections. Further details of progress in this area can be found in CCSP Synthesis and Assess- ment Product 2.2 The First State of the Carbon Cycle Report (CCSP, 2007c). This recent progress does not mean, however, that there are not still a number of important research questions to be answered re- garding the carbon cycle. For example, it will be critical to improve our understanding of the net carbon balance of high- latitude ecosystems, such as tundra and permafrost, as the climate continues warming, since this may be a key feedback on future warming due to the release of methane to the atmosphere (ACIA, 2004). Continued work on developing global carbon cycle models, and improving the linkages between these and other components in Earth systems models, would also be expected to improve our understanding of the interactions between carbon, ecosystems, ag- riculture, and hydrology, in addition to allowing better estimates of overall climate sensitivity and improved realism in future climate change projections. Thus, in our view, the CCSP should continue to include progress in this area as a program priority. 6. Synthesis and assessment products. The CCSP strategic plan (2003) called for the creation of a series of synthesis and as- sessment products that were intended to respond to “the top- priority research, observation, and decision support needs.” While these documents have provided useful summaries of research pro- gress and gaps in several important areas, to date only 3 of the 21 planned products have been released, and there have been concerns raised about the coordination of these activities, the significant time and resources needed to complete each assessment (especially given the time and resources recently dedicated to complete the IPCC assessment process), and the relevance of the synthesis and assessment products to decision makers, including those involved in the CCSP’s own planning and budgeting process (e.g., NRC, 

214 APPENDIX E 2003b, 2007a, d). In addition, the courts have ruled (Center for Biological Diversity v. Brennan, Court Order No. C 06-7062 BNA N.D. Cal. August 21, 2007) that the collection of 21 Synthesis and Assessment products does not fulfill the requirement of the 1990 Global Change Research Act for periodic national assessments of climate impacts. Hence, the emphasis on individual, topical syn- thesis and assessment products may be a priority that could be deemphasized or eliminated in the next CCSP strategic plan. In our view, a true national assessment of current physical changes in climate and the impact of these changes on ecosystems and human systems, at both the regional and national levels, would also be much more valuable than a series of isolated reports on different topics. Without such an assessment, it is difficult for de- cision makers to evaluate the vulnerability of any particular region or sector to climate change, an important first step in improving resiliency to climate variability and change; when compounded by the lack of reliable regional model projections of future climate change (see priority 2, above), the dearth of knowledge about cur- rent and future climate change impacts on different sectors and regions also makes it difficult for decision makers to develop ef- fective plans for responding to climate change through adaptation planning, mitigation efforts, and other activities. However, before a national assessment can be conducted, the United States first needs to develop a comprehensive, inclusive, well-thought-out, and carefully vetted strategy to produce it. Developing such a strategy should, in our view, be a top CCSP priority. EMERGING PRIORITIES 7. Integrated Earth system analysis (IESA). Our first emerg- ing priority (i.e., area that has surfaced or increased in importance during the past 5 years) for the CCSP is IESA. Analysis (or re- analysis, when done retrospectively) refers to the synthesis of numerous, disparate observations, typically using a model, to pro- vide a comprehensive, continuous, and physically consistent quantitative depiction of the temporal evolution of a climate sys- tem component; one notable example is the National Centers for Environmental Prediction (NCEP)/National Center for Atmospheric 

APPENDIX E 215 Research atmospheric reanalysis (Kalnay et al., 1996). There has been good progress over the past decade in reanalyzing several individual system components, including the atmosphere and oceans (e.g., IPCC, 2007). Progress has been somewhat slower for other system components such as the cryosphere, land surface, and eco- systems, and efforts to link these separate analyses together are still in their infancy. In our view, an emerging priority for the CCSP should be the development of an IESA capability that as- similates the full range of Earth system observations, with the objective of creating an accurate, internally consistent, synthesized description of the evolving Earth system that is greater than the sum of its parts. Since analysis is the bridge between observations and model- ing, IESA would build on and expand our two overarching priorities. Many potential benefits of IESA have been identified, including the support of practical applications (e.g., agriculture, energy, and other economic sectors) and contributions to the Global Earth Observing System of Systems (GEOSS). IESA would also help address a previous criticism of the CCSP strategic plan (NRC, 2003b), namely, that observations and modeling ef- forts should be tied together more closely, for example, by using the observational record to critically assess model performance and using model simulations to assess the quality and adequacy of ob- servations. The scientific and technical challenges associated with building an IESA capability include identifying the criteria for op- timizing assimilation techniques for different purposes, estimating uncertainties, and meeting user demands for higher spatial resolu- tion in analysis products. 8. Ice sheets. A number of recent studies have suggested that the Greenland and West Antarctic ice sheets may be less stable than previously thought, raising the possibility that global sea level rise could accelerate. For example, the downward percolation of surface meltwater can lubricate ice sheet movement, speeding ice flow, and destabilize floating ice shelves, as illustrated by the dra- matic breakup of the Larson B ice shelf in 2002. The Larson B collapse also demonstrated that the removal of ice shelves or ice sheets can accelerate ice loss from surrounding glaciers, a process also now observed in Greenland. In addition, the Greenland ice sheet may be more vulnerable than previously thought to seawater 

216 APPENDIX E invasion at its margins. All of these concerns raise the possibility that ice sheets will discharge their ice volume to the sea more quickly, leading to more rapid sea level rise than the 1 to 2 feet projected for the late twenty-first century by IPCC (2007). There are signs that both the Antarctic and especially the Greenland ice sheet are already experiencing accelerating mass loss, but the error bars are wide. Despite these recent advances in understanding, the response of ice sheets to global warming remains one of the largest uncer- tainties in projections of future climate change (IPCC, 2007). Ice sheet dynamics are in general poorly understood and poorly ob- served, and the dynamical response of ice sheets is completely unresolved in the current generation of climate models, and so the vulnerability of the Greenland and West Antarctic ice sheets to accelerated melting or collapse is thus poorly constrained. These gaps in understanding make it extremely difficult to place an upper bound on sea level rise during the twenty-first century and beyond. Rapid ice sheet disintegration, and the accompanying sharp in- crease in sea level rise, is also one of the most alarming potential mechanisms for future abrupt climate change, since even small changes in sea level are expected to have significant impacts on coastal communities and ecosystems around the globe. In our view, the CCSP should make the development of an ice sheet modeling capability in U.S. climate models a priority, including both the provision of computation resources and a plan to draw talent into the field, and should also continue to support ice sheet observations (including those from space; e.g., see NRC, 2007b) and monitoring during and after the International Polar Year. 9. Decadal variability and abrupt climate change. Our next suggested priority is improving our understanding of climate vari- ability on decadal timescales and the nature and likelihood of possible abrupt climate changes during the twenty-first century. The problem of decadal climate variability is a major obstacle to understanding and predicting the El Niño/Southern Oscillation (ENSO) and also a major obstacle to understanding and predicting the regional effects of climate change; however, sorting out de- cadal variability from forced trends plus feedbacks is a major challenge. NRC (1998) identified the importance of the problem and made a single recommendation—that a national program in 

APPENDIX E 217 decadal variability should be established. Unfortunately, such a program has not been established; although some progress has been made through CLIVAR and other domestic and international efforts, progress on this topic has generally been somewhat slow. Understanding abrupt climate change is an important and closely related topic. Changes in the climate system are considered abrupt if they occur more rapidly than the time needed by ecosys- tems and society to adapt to them (NRC, 2002). In addition to rapid ice sheet collapse, described in the priority 8, other major mechanisms for abrupt climate change may include changes in radiative forcing (e.g., methane or carbon dioxide feedbacks), sea ice, the meridional ocean circulation, precipitation regimes (drought), and/or atmospheric circulation regimes (including tropical-extratropical feedbacks). Pos- sible impacts include rapid sea level rise, severe and sustained droughts, or systematic changes in weather patterns over broad re- gions that may result from changes in ocean circulation (CCSP, 2007b). In our view, an important priority for the CCSP is to reduce the remaining knowledge gaps surrounding both decadal variability and abrupt climate change, and obtaining a better understanding of decadal/abrupt processes should be a prerequisite to attempting to de- velop an “early warning system” for detecting abrupt climate change. 10. Nonstationary climate variability and seasonal-to-interannual prediction. Currently, efforts to forecast how the climate may vary based on current conditions (climate predictions) and efforts to model long-term climate changes induced by changes in the natu- ral and anthropogenic forcings (climate projections) are performed on separate research and organizational tracks in the United States. This separation impedes progress on critical questions at the inter- section of climate prediction and projection, such as how the major modes of climate variability (e.g., ENSO, Pacific Decadal Oscilla- tion, North Atlantic Oscillation, monsoons) will respond to global warming, and what the implications of these changes are for the skill of seasonal-to-interannual climate forecasts. Such questions are especially important in the context of our second overarching priority above (regional climate modeling), since one of the pri- mary agents of regional climate change is these same regionally focused modes of climate variability. As a result, we lack a cohe- sive, integrated understanding of climate variability and predictability in the context of changing climate (IPCC, 2007). 

218 APPENDIX E The NRC Review of the Final U.S. Climate Change Science Program Strategic Plan (NRC, 2004b) states that the plan “pre- sents a strategy for producing climate change projections through two modeling centers, but fails to present a national strategy for the seasonal to interannual climate predictions so important to many stakeholders” and that “without a fundamental change in approach to fully support seasonal to interannual climate predic- tion, the United States will be unsuccessful in the delivery of climate services.” In our view, it is imperative that the next CCSP strategic plan include a strategy for moving toward a seamless and integrated suite of forecasting tools that spans the full range of timescales and modes of variability needed to make accurate pre- dictions and projections. A necessary subelement of the above is a critical assessment of the present predictive skill of seasonal-to- interannual climate forecasts, as well as an evaluation of the poten- tial predictability of the climate system on these and longer timescales. Without such an assessment, it will be difficult to de- velop a national strategy that integrates basic research, applied research, and application of climate forecasts on regional scales in direct support of the delivery of climate services (see also priority 14 below). 11. Earth system predictability. A number of federal agencies, departments, and programs produce forecasts and other environ- mental predictions. For instance, NOAA’s NCEP delivers climate, weather, and ocean prediction products to a range of users, and is at the leading edge of efforts to improve predictive capabilities for these and other elements of the physical climate system. However, as distinct from 100-year projections, there does not appear to be a national strategy to expand, integrate, and improve these capabili- ties into a comprehensive environmental prediction system that combines weather forecasts, short-term climate predictions, and other components of the Earth system such as air quality, water quality, and terrestrial and ocean ecosystems. Ideally, such a strat- egy would ultimately lead to a robust operational infrastructure for an “end-to-end” process that accesses a huge array of global ob- servations; assimilates them into interactive and coupled global atmospheric, land, and ocean models; runs weather and climate forecast models; and delivers forecast products to a diverse array of users. 

APPENDIX E 219 The European Centre for Medium Range Weather Forecasting is actively moving toward a program for Earth system prediction that includes assimilation of the global carbon cycle, prediction of infectious disease outbreaks such as malaria, and seasonal fore- casts for a range of agricultural crops. Similarly, in Japan the Earth Simulator supercomputer has served as a national focus for the development of a comprehensive Earth system model. In our view, the United States should also make Earth system prediction a prior- ity. NOAA, NASA, and other agencies have already demonstrated international leadership with concepts such as GEOSS. A key to the success of GEOSS in the long term will be the sustained use and demand for Earth system observations in support of operational prediction across a broad range of sectors and Earth system com- ponents. This priority also has close connections to several of the other priorities suggested in this document, including the two overarching priorities on observations and regional modeling and the emerging priority on nonstationary prediction. Furthermore, the development of a predictive capability for the Earth system has unique policy relevance at both the national and international levels with respect to agriculture, ocean resources, energy, transportation, commerce, health, and homeland security. However, important questions will need to be answered before true Earth system predic- tions will be possible, such as the optimal pathway for developing predictive coupled physical–biological–chemical models of the Earth system, and the limits of predictability for different system components. 12. Geoengineering. Scientists and policymakers are begin- ning to appreciate that responding to climate change will require a portfolio of different responses (e.g., Pacala and Socolow, 2004). One possible, but controversial, response is geoengineering, or direct human intervention in the climate system intended to offset some aspects of climate change (NRC, 1992). Before policy mak- ers can decide if geoengineering should play a role along with adaptation and mitigation efforts (for instance, if global warming occurs even more rapidly than the high end of the IPCC scenarios), a major research effort is needed to understand the efficacy, costs, and potential consequences and risks of the various geoengineering strategies that have been proposed, and to identify other potential alternative strategies. While there is a danger that some may inter- 

220 APPENDIX E pret geoengineering research as a “quick fix” to the climate prob- lem that obviates critical adaptation and mitigation efforts, a failure to conduct careful research into different alternatives would be an even bigger risk. Articles on geoengineering by well- respected researchers are beginning to appear in the literature, but a more extensive research program in this area is needed. Such a program would need to involve both the CCSP and the Climate Change Technology Program, but the responsibilities, cost, deliv- erables, and form that such a program should take still need to be determined. CROSSCUTTING (INSTITUTIONAL AND ORGANIZATIONAL) PRIORITIES1 13. Computing (and storage). In our view, most if not all the priorities listed above are dependent on, and have to date been lim- ited by, inadequate computational resources and the associated technical personnel. For example, regional climate modeling, cli- mate reanalyses, and Earth system modeling have all been constrained by the lack of access to petascale computing, despite the current CCSP strategic plan (CCSP, 2003) listing “develop- ment of state-of-the-art climate modeling” as a CCSP priority. Storage issues have also gained urgency in light of the data man- agement challenges associated with archiving and providing access to large quantities of data from new satellite missions and other high-volume observational streams (NRC, 2007c). In addition, the United States appears to lack a comprehensive strategy to marshal the considerable technical talent currently available in the private sector to address computing issues in climate research. Thus, in our view, a critical crosscutting priority for the CCSP is a realistic assessment of current and future computational requirements fol- lowed by the development of a comprehensive strategy for providing the requisite computational resources to support pro- gram activities. First and foremost among these is the issue of CPU speed and availability of overall computational horsepower/cycles, 1 Our two overarching priorities (observations and regional modeling), could also be considered crosscutting priorities; however, we consider them sufficiently important to warrant elevating them to a higher level. 

APPENDIX E 221 but other important issues include dealing with heterogeneous plat- forms, distributed/grid computing, and data archiving. Interagency coordination across the National Science Foundation, NASA, NOAA, and the Department of Energy and related engagement of the computer science and information technology communities will be essential to success. 14. Climate services. As a result of the progress made to date within the CCSP and its predecessor USGCRP, the nation is poised to benefit in a routine manner from the transition from basic research to applied research to the provision of climate services—a mechanism to connect climate science to decision-relevant ques- tions and support building capacity to anticipate, plan for, and adapt to climate fluctuations (Miles et al., 2006). Climate services, by definition, are “mission-oriented and driven by societal needs to enhance economic vitality, maintain and improve environmental quality, limit and decrease threats to life and property, and strengthen fundamental understanding of the earth” (NRC, 2001). While climate services may have some aspects in common with the mission and products of the National Weather Service, climate products are far more diverse and have unique requirements that go far beyond those associated with the day-to-day prediction of atmospheric conditions (e.g., Visbeck, 2008). However, there is currently not a single lead agency with the mandate and resources to operationally deliver climate services in response to stakeholder needs. What is needed, in our view, is a National Climate Service with many of the elements envisioned by Miles et al. (2006), for instance, organization at the federal level but taking advantage of the substantial regional-level expertise and experience (for in- stance, through the Regional Integrated Sciences and Assessments) needed to connect scientific results to individual stakeholders. Also, as called for in NRC (2001), the research enterprise dealing with environmental change and environment–society interactions should be enhanced in order to address the consequences of cli- mate change and better serve the nation’s decision makers, including “support of (a) interdisciplinary research that couples physical, chemical, biological, and human systems; (b) improved capability to integrate scientific knowledge, including its uncer- tainty, into effective decision support systems; and (c) an ability to conduct research at the regional or sectoral level that promotes 

222 APPENDIX E analysis of the response of human and natural systems to multiple stresses.” 15. Integrated assessment. Integrated assessment, our final crosscutting priority area, refers to the integrated analysis and modeling of the human activities and natural processes that give rise to greenhouse gas emissions and other climate forcings, the changes in the climate system caused by these forcings, the vulner- ability and adaptive capacity of both human systems and natural systems to these changes in climate, and the estimated costs, bene- fits, and limitations of various mitigation and adaptation measures (e.g., Parson and Fisher-Vanden, 1995). By developing compre- hensive models to address these issues, we could begin to provide decision makers with the information needed to make climate pol- icy decisions that are both environmentally effective and economically efficient. The U.S. research community has a great deal of experience in the area of environmental assessments (e.g., NRC, 2007a), but integrated assessment represents a truly cross- cutting activity that straddles even the traditional boundary between the natural and social sciences, which poses additional challenges and barriers. In our view, the CCSP should become more involved in leading and coordinating integrated assessment efforts—which should not be one-time events but a process that bal- ances the needs of policy makers and the flow of information—to ensure that decision makers have access to the full range of informa- tion they need to respond to the many challenges of climate change. REFERENCES ACIA (Arctic Climate Impact Assessment), 2004, Impacts of a Warming Arctic, Synthesis Report, Cambridge University Press, Cambridge, 139 pp. CCSP (Climate Change Science Program), 2003, Strategic Plan for the U.S. Climate Change Science Program, Climate Change Sci- ence Program and Subcommittee on Global Change Research, Washington, D.C., 202 pp. CCSP, 2007a, Our Changing Planet: The U.S. Climate Change Science Program for Fiscal Year 2008, Climate Change Sci- 

APPENDIX E 223 ence Program and Subcommittee on Global Change Research, Washington, D.C., 212 pp. CCSP, 2007b, Summary of Revised Research Plan, December 27, 2007, 12 pp., available at http://www.climatescience.gov/Library/ stratplan2008/summary/. CCSP, 2007c, The First State of the Carbon Cycle Report (SOCCR): The North American Carbon Budget and Implications, A.W. King, L. Dilling, G.P. Zimmerman, D.M. Fairman, R.A. Hough- ton, G. Marland, A.Z. Rose, and T.J. Wilbanks, eds., Synthesis and Assessment Product 2.2, Climate Change Science Program and Subcommittee on Global Change Research, Asheville, NC, 242 pp. Hay, L.E., M.P. Clark, R.L. Wilby, W.J. Gutowski, G.H. Leavesley, Z. Pan, R.W. Arritt, and E.S. Takle, 2002, Use of regional climate model output for hydrologic simulations, Journal of Hydrometeorogy, 3, 571–590. IPCC (Intergovernmental Panel on Climate Change), 2007, Climate Change 2007: Synthesis Report, Contribution of Working Groups I, II, and III to the Fourth Assessment Report of the Intergovernmen- tal Panel on Climate Change, Core Writing Team, R.K. Pachauri and A. Reisinger, eds., Geneva, 104 pp. Kalnay, E., M. Kanamitsu, R. Kistler, W. Collins, D. Deaven, L. Gandin, M. Iredell, S. Saha, G. White, J. Woollen, Y. Zhu, A. Leetmaa, B. Reynolds, M. Chelliah, W. Ebisuzaki, W. Hig- gins, J. Janowiak, K. Mo, C. Ropelewski, J. Wang, R. Jenne, and D. Joseph, 1996, The NCEP/NCAR 40-year reanalysis project, Bulletin of the American Meteorological Society, 77, 437–471. Miles, E.L, A.K. Snover, L.C. Whitely Binder, E.S. Sarachik, P.W. Mote, and N. Mantua, 2006, An approach to designing a na- tional climate service, Proceedings of the National Academy of Sciences, 103, 19,616–19,623. NCAR (National Center for Atmospheric Research), 2005, Mesoscale and Microscale Meteorology (MMM) Division Science Plan: Five Years and Beyond, available at http://www.mmm.ucar.edu/ about_mmm/stratplan/index.php. NRC (National Research Council), 1992, Policy Implications of Greenhouse Warming: Adaptation, Mitigation, and the Science Basis, National Academy Press, Washington, D.C., 944 pp. 

224 APPENDIX E NRC, 1995, A Review of the U.S. Global Change Research Pro- gram and NASA’s Mission to Planet Earth/Earth Observing System, National Academy Press, Washington, D.C., 96 pp. NRC, 1997, Letter to OSTP and Department of State on Global Observations of Climate, National Research Council, Wash- ington, D.C., 2 pp. plus attachments. NRC, 1998, Decade-to-Century-Scale Climate Variability and Change: A Science Strategy, National Academy Press, Wash- ington, D.C., 160 pp. NRC, 1999, Adequacy of Climate Observing Systems, National Academy Press, Washington, D.C., 50 pp. NRC, 2001, A Climate Services Vision: First Steps Towards the Future, National Academy Press, Washington, D.C., 96 pp. NRC, 2002, Abrupt Climate Change: Inevitable Surprises, Na- tional Academy Press, Washington, D.C., 244 pp. NRC, 2003a, Estimating Climate Sensitivity: Report of a Work- shop, National Academies Press, Washington, D.C., 62 pp. NRC, 2003b, Planning Climate and Global Change Research: A Review of the Draft U.S. Climate Change Science Program Strategic Plan, National Academies Press, Washington, D.C., 85 pp. NRC, 2003c, Understanding Climate Change Feedbacks, National Academies Press, Washington, D.C., 166 pp. NRC, 2004a, Climate Data Records from Environmental Satellites, National Academies Press, Washington, D.C., 136 pp. NRC, 2004b, Implementing Climate and Global Change Research: A Review of the Final U.S. Climate Change Science Program Strategic Plan, National Academies Press, Washington, D.C., 96 pp. NRC, 2005a, Earth Sciences and Applications from Space: Urgent Needs and Opportunities to Serve the Nation, National Acad- emies Press, Washington, D.C., 45 pp. NRC, 2005b, Radiative Forcing of Climate Change: Expanding the Concept and Addressing Uncertainties, National Acad- emies Press, Washington, D.C., 224 pp. NRC, 2007a, Analysis of Global Change Assessments: Lessons Learned, National Academies Press, Washington, D.C., 196 pp. 

APPENDIX E 225 NRC, 2007b, Earth Sciences and Applications from Space: Na- tional Imperatives for the Next Decade and Beyond, National Academies Press, Washington, D.C., 456 pp. NRC, 2007c, Environmental Data Management at NOAA: Archiv- ing, Stewardship, and Access, National Academies Press, Washington, D.C., 130 pp. NRC, 2007d, Evaluating Progress of the U.S. Climate Change Science Program: Methods and Preliminary Results, National Academies Press, Washington, D.C., 170 pp. NRC, 2007e, NOAA’s Role in Space-Based Global Precipitation Estimation and Application, National Academies Press, Wash- ington, D.C., 142 pp. NRC, 2007f, Options to Ensure the Climate Record from the NPOESS and GOES-R Spacecraft: A Workshop Report, Na- tional Academies Press, Washington, D.C., 88 pp. NRC, 2008, Earth Observations from Space: The First 50 Years of Scientific Achievements, National Academies Press, Washing- ton, D.C., 144 pp. Pacala, S., and R. Socolow, 2004, Stabilization wedges: Solving the climate problem for the next 50 years with current tech- nologies, Science, 305, 968–972. Parson, E.A., and K. Fisher-Vanden, 1995, Searching for Inte- grated Assessment: A Preliminary Investigation of Methods, Models, and Projects in the Integrated Assessment of Global Climatic Change, Consortium for International Earth Science Information Network, University Center, Michigan. Roe, G.H., and M.B. Baker, 2007, Why is climate sensitivity so unpredictable? Science, 318, 629–632. Visbeck, M., 2008, From climate assessment to climate services, Nature Geoscience, 1, 2–3.