A Plan for a Research Program on Aerosol Radiative Forcing and Climate Change (1996)

It is the panel's judgment that (1) airborne particles can reduce solar radiation reaching the Earth's surface; (2) anthropogenic aerosol provides a negative climate forcing for large regions, and sulfate particles produce forcing in the Northern Hemisphere comparable to that from anthropogenic greenhouse gases (but opposite in sign); and (3) there is substantial uncertainty about the magnitude and spatial distribution of this radiative forcing of the climate by aerosols. Reduction in this uncertainty requires an integrated program that will pull together the research capabilities of the aerosol and atmospheric science communities in this country. The proposed Interagency Climate-Aerosol Radiative Uncertainties and Sensitivities (ICARUS) Program represents such a strategy.

At the outset it should be stressed that the guiding principle of the proposed overall ICARUS research program is integration, within programs sponsored by each agency and among programs sponsored by separate agencies. This latter integration might be termed horizontal integration, that is, to promote interagency coordination and cooperation. Integration across modeling, monitoring, process, and laboratory studies is also essential, whether these programs are sponsored by a single agency or more than one agency; this might be called vertical integration. The vertical integration of modeling and experimental studies is needed to guide experimental programs via sensitivity analysis and, in turn, to provide data to improve and evaluate models of atmospheric aerosol concentrations, composition, and behavior, including radiative and cloud interactions.

The goal of the ICARUS program is to provide climate models with sufficiently reliable estimates of radiative forcing by airborne particles so that remaining uncertainties in forcing no longer limit quantitative evaluation of climate change. A prime responsibility of the ICARUS organizational structure would be to ensure integration among field and laboratory measurements, model development, and model evaluation.

Prior to the present report, many studies have been recommended to reduce uncertainties about direct radiative effects of airborne particles and to permit responsible estimates of their indirect effects. Appendix A gives an incomplete, but illustrative, list of such recommendations, generated during the past quarter century. It is our opinion that currently the greatest need is for administrative leadership in executing a program on aerosols and climate. Thus, we recommend both an administrative strategy and a research strategy, because it is our firm conclusion that developing and applying an effective administrative strategy is equally critical to specifying needed research. To accomplish this goal and achieve the needed integration, the following actions are recommended.

Past failure to undertake needed climate-aerosol research has been caused, mainly, by inadequate coordination, stimulation, recognition, and encouragement. What is needed, as a start, is to stimulate U.S. aerosol and climate scientists (generally affiliated with separate agencies) to pull together as a team. Therefore, as a first step in Action Item 1, we recommend that the four federal program managers who requested this report define leadership for U.S. climate-aerosol research by forming a Science Team and an Executive Committee for an ICARUS program.

We expect that these four managers can immediately create this ICARUS program, but, to empower it fully, ICARUS must be organized as a component of the U.S. Global Change Research Program (USGCRP). We therefore recommend that, as soon as possible, ICARUS be recognized as an integral part of the USGCRP. Thereby, no doubt, ICARUS would receive USGCRP guidance for agency representation, Science Team, and Executive Committee membership.

It is appropriate that (1) the ICARUS Executive Committee include government representatives from at least the National Aeronautics and Space Administration (NASA), National Oceanic and Atmospheric Administration

(NOAA), National Science Foundation (NSF), and Department of Energy (DOE), and the director of the USGCRP; (2) the ICARUS Science Team include representatives from the principal climate-aerosol research groups in the United States, as well as selected international researchers; (3) at least annually, a national ICARUS forum be convened for interdisciplinary, interagency, and international discussions about aerosol-climate research progress; and (4) the prime responsibilities of the ICARUS Executive Committee be to set research priorities (see Action Item 2) and then coordinate, evaluate, fund, and steer U.S. research on the climate-aerosol problem toward measurable achievements (see Action Item 3).

The procedure for setting research priorities starts with evaluations of the sensitivities of climate predictions to uncertainties in a host of processes (cloud cover, ocean circulation, air-surface fluxes, snow and ice cover, etc., including changes in atmospheric aerosols); the result is the familiar USGCRP "priority framework."

The USGCRP should reevaluate its science priorities to establish the importance of aerosols to climate sensitivity and uncertainty, relative to all other sources of uncertainties in climate predictions, and then define and adopt an appropriate funding priority for the ICARUS program. It is our expectation that, when this action by the USGCRP has been completed, ICARUS research priorities will appear prominently in the USGCRP's priority framework.

At the present time, the role of clouds in the climate system is listed as the most important research priority for the USGCRP because the degree of response of clouds and water vapor to climate change controls in large measure the overall sensitivity of climate to greenhouse gases. This sensitivity can be measured (and therefore quantified) only by measuring climate change as well as forcing. In view of the importance of knowing the correct forcing for quantifying this process, we find that the quantification of aerosol forcing should be considered of equal priority to the role of clouds in the climate system within USGCRP.

In light of the fundamental complexity of the aerosol-climate problem, discerning how to proceed with the ICARUS program will require wise

leadership, guided by much-improved mathematical models that are based on accurate and representative field and laboratory data. To advance, there is no scientifically defensible alternative other than to apply the obvious iterative approach: understand the relevant atmospheric processes, improve models, apply sensitivity analyses to design new observational and experimental programs, and use results from these measurement programs to improve models further. Simultaneously, through this iterative process, ICARUS can promote the fundamentally important need to strengthen working relationships between theoreticians and experimentalists—forcing integration between measurement and modeling programs.

Once the USGCRP has defined funding levels for ICARUS and the ICARUS Executive Committee has adopted research priorities, then as Action Item 3, Phase 1, the Executive Committee should issue a Request for Proposals (RFP) and thereby invite all interested scientists, federal agencies, and all relevant international research programs to submit detailed plans for proposed climate-aerosol research projects for evaluation by ICARUS Science Team groups. In particular, we encourage NASA and DOE to progress from the available ''strategic documents" describing CLIMSAT and an aerosol component of the Atmospheric Radiation Measurement (ARM) Program to "implementation plans," as are now available for the Tropospheric Aerosol Radiative Forcing Observation Experiment (TARFOX) and the North Atlantic Regional Aerosol Characterization Experiment (ACE-2). Thus, in summary, we suggest Action Item 3, Phase 1: Issue a single RFP for research plans to be submitted to ICARUS for evaluation.

The next phase is Action Item 3, Phase 2: ICARUS Science Team groups should evaluate submitted research plans. We have distinguished this phase in part because conducting reviews is an important, difficult, and time-consuming task; in part because we wanted to mention that ICARUS Science Team groups should perform the reviews; and in part because we felt it appropriate to add the parenthetic remark that we are unconcerned about potential conflicts of interest associated with these reviews. The potential for individual scientists (or groups of scientists) essentially reviewing their own research plans can be avoided by following standard procedures, including requirements for testimonials about any conflicts of interest and appropriate self-regulating disqualification.

An important detail about Action Item 3 is our recommendation that a number of models be available to address the climate-aerosol issue. No doubt each model will appropriately emphasize different aspects of greatest concern to the developer. Using models to perform sensitivity analyses, each agency should then develop plans for future observational and experimental projects and submit these plans for comment by the ICARUS Science Team. For the next decade or so, models are likely to continue to be, at once, so complicated and yet sufficiently inadequate that derived program plans will almost

certainly be model dependent. Consequently, the only procedure available for ICARUS Science Team groups to evaluate project plans independently (if not totally objectively) will be to subject them to analyses by more than one model, developed relatively independently.

The third, continuing phase of maintaining effective leadership is Action Item 3, Phase 3: ICARUS Executive Committee should maintain an integrated program. As emphasized earlier, integration must be a cornerstone of an effective program. This integration must occur both horizontally (across agencies and overall projects) and vertically, across the four classes of activities:

1. chemical, climate, and radiative transfer modeling;

2. in situ measurements;

3. remote measurements; and

4. laboratory and field process studies.

Several conclusions emerge from these three proposed action items that should be a central consideration in allocating resources to research on climate forcing by aerosols. These can be listed as a set of questions to be asked regarding any proposed project:

• Is the research part of the integrated interdisciplinary program? Is it part of the systematic approach?

• Is it focused on defining uncertainties and reducing them?

• Is it practical and achievable?

• Does it relate to relevant effects on regional to global scales?

The suggested organizational structure of the ICARUS program is shown in Figure 4.1. The ICARUS Executive Committee would be composed of the program managers of the four federal agencies and selected members from the ICARUS Science Team. An ICARUS Review Panel, appointed by the USGCR Program Office (PO) and comprised of global change scientists outside the immediate ICARUS program, would periodically review ICARUS and report progress and problems to the USGCRPO.

In the preceding, we have presented a vision of an integrated program to reduce uncertainties about climate forcing by aerosols. We recommend that, to launch ICARUS, the four agency program managers who requested this report (and, as appropriate, their choices of lead ICARUS scientists and representatives from other agencies that commit to funding ICARUS research) proceed to prepare a joint RFP, in coordination with the USGCRP. In the next section, we suggest some broad elements of this first RFP for

FIGURE 4.1 Organizational structure of the ICARUS program.

NOTE: CENR = Committee on Environment and Natural Resources.

ICARUS. When general features of this joint RFP are agreed upon (including identification of lead agencies for its various components, and after proposals have been reviewed by a panel created by the ICARUS Executive Committee), contracts should be let by each agency for its identified ICARUS components, according to each agency's standard practices.

In Chapter 2 we identified the components of research required to increase our understanding of aerosol forcing of climate. We stressed that this should be an integrated program involving process studies, model development and evaluation, field measurements, and technology development. In this final section of the report, we recommend specific research tasks, together with estimated costs and durations.

From the point of view of cost, four categories of research projects emerge from the analysis in Chapter 2:

Each category represents an important component of the overall plan, yet the panel recognizes that the manner in which funding decisions are made will likely vary according to the category of project. In spite of the substantial cost of new satellite systems for remote sensing, this type of development is likely to be undertaken by a single agency such as NASA. On the other hand, large, multiplatform field studies are generally funded by a consortium of agencies. Individual process studies are normally funded by a single agency. The goal of the ICARUS management structure is to achieve communication among the different agencies supporting the research, so that the various projects constitute an integrated attack on the problem of aerosol forcing of climate.

It is important to note that the costs itemized above and below assume the existence of an atmospheric science research infrastructure. This includes active research groups in universities and the national laboratories, ships (the University-National Oceanographic Laboratory System and NOAA fleets), and aircraft (e.g., those at the National Center for Atmospheric Research and NASA) that are configured to conduct research of this type, and supercomputing facilities that can run the complex models. If budgetary constraints were to necessitate a substantial reduction in the current infrastructure, it would become much more costly to redevelop the capability for answering these questions.

Using existing global climate and global chemistry models, determine relative sensitivities of both global and regional climate predictions to uncertainties in aerosol forcing versus uncertainties in other factors, such as cloud parameterizations. Using global chemistry models examine the sensitivities of aerosol forcing predictions [at scales from global circulation model (GCM)

grid scales to the global scale] to uncertainties in microphysical properties (aerosol size, aerosol composition, relative humidity) and processes (sulfate production pathways and rates, and vertical convective transport). We recommend up to four such projects to enable assessment of the influence of model platform and assumptions employed in individual models.

Develop and evaluate the next-generation models of aerosol radiative forcing at scales from regional to global, incorporating all chemical, physical, and meteorological processes that are important in determining the concentrations, radiative and cloud-nucleating properties, and direct and indirect radiative forcing from natural and anthropogenic aerosols. We recommend two such projects, employing different GCM and global chemical model platforms. Note that aerosol submodel development will require integration with research on aerosol process models.

The goal is to ascertain to what extent homogeneous nucleation is occurring in the atmosphere, where (e.g., marine boundary layer, free troposphere) it is occurring, under what conditions, what chemical species are involved, and to what extent theoretical treatments of nucleation can represent observed new particle formation in the atmosphere. This project will consist of several individual, highly integrated, projects, including

1a. elucidation of the linkage between dimethyl sulfide (DMS) SO₂ emissions and nss (non-sea salt)-sulfate formation in the marine and continental boundary layer based on evaluation of boundary-layer aerosol data;

1b. evaluation of nucleation as a source of new particles in the free troposphere based on evaluation of free tropospheric aerosol data;

1c. evaluation of the applicability of nucleation theory to describe the atmospherically relevant systems of H₂SO₄-H₂O, NH₃-H₂SO₄-H₂O;

1d. evaluation, through laboratory experiments, of the conversion of low vapor pressure organic species to aerosols through nucleation processes; and

1e. elucidation of tropospheric aerosol growth processes involving gas-to-particle conversion both in clear air and in-cloud, based on analysis of ambient size distribution data and chemical routes of gas-to-particle conversion.

Note that each of the projects must be closely integrated with field measurement programs.

The goal is to determine the extent to which the theoretical treatments of aerosol and cloud optical properties that are used in aerosol forcing models are applicable to ambient aerosols and clouds. This includes laboratory experiments on pure and mixed-component aerosols, together with in situ measurements of ambient particles for important classes of tropospheric aerosols. Also included are theoretical treatments and in situ and satellite measurements of cloud optical properties and albedo to determine the accuracy of the treatment of cloud optical properties in global models.

For fundamental evaluation of the effect of anthropogenic emission changes on cloud optical properties (the indirect effect), it is necessary to develop a first-principles understanding of the relation between changes in aerosol number concentration (as a function of aerosol molecular composition) and cloud condensation nuclei (CCN) behavior; between CCN and resulting cloud drop number concentrations (CDNC); and between CDNC and cloud albedo. This project includes detailed aerosol process modeling, coupled with cloud microphysical modeling. It also includes laboratory and field studies of the CCN properties of aerosols in both clean and anthropogenically influenced regions of the atmosphere. Note the integration with technological development of new CCN spectrometers. Aerosol model parameterizations of the aerosol-CCN-CDNC-cloud albedo linkage should be developed for GCMs and ACTMs (atmospheric chemical transport models).

The prime goal of the dry deposition studies is to obtain reliable, particle size-specific data in the field (i.e., not just from wind tunnels) for dry deposition to "real-world" collectors—from forests in inhomogeneous terrain to the oceans under a variety of conditions. To achieve the essential goal of obtaining particle size-specific dry deposition velocities, developments in technology (e.g., for eddy-flux measurements) and techniques (e.g., for measuring monodisperse particles actually deposited) almost certainly will be necessary. The prime emphasis of the precipitation scavenging field studies will be to develop parameterizations for storm venting for use in global-scale models for all relevant species (especially for particles as a function of their sizes), for all climatologically important cloud and storm types, and for representative ranges of pollution loadings and storm microphysical and dynamical variables. Analysis of field data must be performed with appropriate mesoscale models of precipitation formation, efficiencies, and scavenging.

One multiplatform field campaign will take place every two years for the next decade. The goals are to integrate satellite radiation measurements and surface-based column-integrated radiation measurements with in situ (aircraft) chemical, physical, and optical measurements in both clear-sky and cloud environments, and in both clean and anthropogenically influenced air masses. In situ measurement of CCN, CDNC, aerosol size and composition distribution, optical depth, and cloud albedo will be needed. The goal is to perform closure studies to assess the accuracy of treatment of aerosol and cloud radiative processes in GCMs and ACTMs.

Conduct a research measurement effort involving regularly scheduled flights by a suitably instrumented aircraft in a nearly continuous circuit over selected surface aerosol monitoring sites. A similar routine measurement program from ships of opportunity is also suggested as is a program of balloon-borne measurements. All of these observations are needed to provide ground truth for satellite observations and the information necessary to determine chemical composition for source identification and quantification.

Establish a dual-density network of surface-based stations for continuous monitoring of aerosols. A high-density network of about 30 stations spanning North America will characterize the spatial distribution, seasonal variability, and trends of aerosol optical depth with a spatial coverage suitable for testing chemical transport and radiative transfer models. A companion network of about 10 stations will provide detailed information on means, variability, and trends of key aerosol radiative, chemical, and microphysical properties, for different aerosol types, that are used in chemical transport and radiative transport models. These networks will be supplemented with a systematic program of ship-and airborne surveys to characterize the horizontal and vertical distributions, over the global oceans and North America, of the same aerosol properties studied in the low-density network.

It is recommended that data from current (e.g., Advanced Very High Resolution Radiometer) and future [e.g., Multiangle Imaging Spectroradiometer (MISR)] satellite instruments be integrated with laboratory, theoretical and in situ optical, chemical, and microphysical studies in order to quantify the uncertainty inherent in passive radiometry and optimize the information provided. For improving retrievals of tropospheric aerosol optical depth, especially over land, and for determining the vertical distribution of tropospheric aerosols globally, we recommend that a spaceborne aerosol backscatter measuring lidar operating at a minimum of 2 wavelengths with the capability of measuring depolarization be flown as soon as possible. On the same spacecraft, a carefully designed set of ancillary instruments boresighted with the lidar should be flown, e.g., an imager to aid in horizontal extrapolation. It is necessary to have in place a network of supporting ground-based and airborne in situ measurements to constrain the remotely sensed retrievals of optical depth to meet the required measurement accuracies for both tropospheric and stratospheric aerosols. Further, we recommend that the Stratospheric Aerosol and Gas Experiment (SAGE) series of satellite remote sensors be continued for the measurement of aerosol optical depth in the stratosphere, and that global coverage be achieved. Presently, NASA's Earth Observing System (EOS) program is developing SAGE III instruments for flights beginning in 1998. However, global coverage requires a spacecraft in both polar and mid-inclined orbits, simultaneously.

The goal is to advance the state of the art of instrumentation needed to assess aerosol forcing. Particular focus is on the development of compact and lightweight sensors for aircraft use.

The goal is to develop the capability to monitor extinction profiles (i.e., vertically resolved aerosol optical depths) on a global scale from an Earth-orbiting spacecraft. Lidars appear to be the strongest candidates for producing unambiguous vertically resolved measurements in the troposphere with the accuracies needed, especially when coupled to local in situ measurements that constrain the inverted solutions. In the stratosphere, where optical depths are low and aerosols more homogeneous, limb occultation passive instruments provide the necessary aerosol information. All passive sensor data, however, especially those from NASA's EOS program, which begins its flights in 1998 (SAGE III, Moderate-Resolution Imaging Spectroradiometer, MISR, etc.), will be integrated into this program.

We recognize that this integrated program of research on climate forcing by aerosols will require the commitment of resources for a sustained period of time. We also recognize that acquiring new resources during a period of fiscal constraint is likely to be difficult. We urge the agencies that are involved to recognize the strength of the evidence for the existence of a substantial aerosol forcing and base the priority for this proposed research accordingly.

The foregoing projects will be useful for reducing uncertainties in climate forcing by aerosol only if three further activities are undertaken. It is

essential that the ICARUS management structure formally organizes projects on coordination, integration, and assessment.

The goal is to maintain communication among projects regarding technical details of the ongoing efforts. Such coordination is necessary, for example to ensure that models produce results with averaging times that are relevant to in situ measurements. Ground-based, aircraft, and satellite observations must coincide in space and time if closure exercises are to be conducted. Two specific activities are envisaged:

1. a small coordination project that actively tracks all the projects and seeks to develop logical interfaces between them ($100,000 per year), and

2. annual Science Team meetings for reporting research results and exchanging data ($100,000 per year).

The final product of all of the projects is supposed to be information on climate forcing by aerosols that is tailored for use in studies of climate response. Thus, there is a need for active interaction with the meteorological and climate modeling community as well as the meteorological data analysis community. In addition, there is a need for delivery of these research results to the Intergovernmental Panel on Climate Change (IPCC). Presently, U.S. contributions to IPCC regarding aerosols are arranged on an ad hoc basis, and there is no means for coordinating U.S. scientific input other than by actions of individual scientists.

The needs here are for time and travel for representatives of ICARUS to deliver results to the user community ($50,000 per year).

The dominant role of ICARUS will be to integrate U.S. research on aerosol forcing of climate. Some of this integration will be funded indirectly, through each agency's representatives on ICARUS's Executive Committee and through sponsorship of principal investigators who will be members of the ICARUS' Science Team. Other integration tasks, however, will require separate and sometimes substantial funding, including establishing and maintaining quality assured data bases; providing coordination and support services for field campaigns; funding model and measurement intercomparison studies; and performing a variety of management tasks, such as hosting workshops

and annual meetings, and publishing proceedings. Consequently, ICARUS's first RFP should solicit proposals for providing integration services, possibly through a management contract.