History and Status of the ARM Program
Introduction
More than 13 years after the dedication of the first research site of the U.S. Department of Energy's Atmospheric Radiation Measurement (ARM) Program, the primary tenet of the program has remained unchanged: to improve the performance of the general circulation models used for climate research and predication by improving how those models deal with radiative energy transfer and the impact of clouds.
To this end, the ARM Program has made a significant contribution in improving climate prediction models: radiative heat transfer, radiation absorption, and cirrus cloud properties. ARM scientists are using data gathered from ARM’s three continuously sampling observatories worldwide to address these issues and compare the observations to their models. As the quality of the data has improved over time, the scientists have been able to generally compute the radiation transport to the accuracy required for atmospheric modeling, although there remain unresolved issues relating to the modeling of complex cloud geometries. The use of the ARM sites by other programs such as NASA and NOAA and other countries such as Japan and several European countries has exceeded expectations. This worldwide interest bodes well for the ARM Program and for the future of ground-based remote sensing for climate modeling and weather forecasting.
Initial Concept
The initial concept for ARM came out of a series of studies that fell under the auspices of the Intercomparison of Radiation Codes in Climate Models (ICRCCM). ICRCCM pointed to several key issues that are now central to the ARM approach and strategy. First, ICRCCM was based on an assertion that one must understand the quality of the physics inside a climate model if one is to understand the quality of the climate model itself. Second, it is possible, and in fact necessary, to understand the relatively coarse representations of the physical processes in a climate model in terms of higher resolution process models. Finally, the hierarchy of models that leads to needed parameterizations must be built on a sound base of experimental verification.
Concurrently, with the release of the ICRCCM results, it was becoming clear that the radiative transfer of energy in the atmosphere and the impact of clouds was, and remains, one of the greatest sources of error and uncertainty in the current generation of general circulation models used for climate research and prediction. With this starting point, DOE proposed a major program targeted at improving the understanding of the role and representation of atmospheric radiative processes and clouds in models of the earth's climate. Initially, the DOE program focused on the radiative aspects of the climate problem. As the scientific issue was studied in more detail, however, it was obvious that a study of radiative processes associated with clouds could not be decoupled from the problem of representing the processes by which clouds form, are maintained, and dissipate in climate models. As a result, the ARM Program was proposed to the then Committee on Earth Sciences of the Federal Coordinating Council on Science Engineering and Technology with two basic objectives:
- to improve the treatment of radiative transfer in climate models under all relevant conditions
- to improve the treatment of clouds in climate models, including the representation of the cloud life cycle and the prognosis of cloud radiative properties.
The Approved Plan
The ARM Program Plan was subjected to peer review in the fall of 1989. The key element of the proposed ARM effort was to be the Cloud and Radiation Testbed (CART). This user facility was proposed to consist of four to six semi-permanent observational facilities designed to allow detailed investigations of processes represented in climate research models. These semi-permanent facilities were to be supplemented with a mobile facility that would allow related measurements to be made at other locations on a campaign-oriented basis. The CART facility would include a data management and communications system capable of acquiring and quality-controlling site data; acceptability to acquire data from sources outside the program; and to communicate that data to a Science Team. The Science Team would be selected through a peer review process open to all investigators nationally and internationally.
Based on the peer review, the subcommittee on Global Change Research of the Committee on Earth Sciences approved the Plan, noting several key things about how it should be carried out. First, the scope was broadened beyond radiative transfer to include clouds and cloud processes represented in general circulation models, a change deemed necessary to adequately address those atmospheric properties important to radiative transfer in the atmosphere and the atmosphere's radiation balance. Next, the Committee recommended that the DOE implementation of this program involve the talents of other federal agencies to the extent possible and that an interagency steering group be formed to assist in that process. Finally, the relevance of ARM to several other climate programs was noted, and DOE agreed to coordinate its deployment of facilities with the schedules of other national and international programs.
The Early Implementation
The implementation of the ARM Program began in January 1990, proceeding on two coupled but parallel tracks. First, a multi-laboratory team was formed to plan the detailed implementation of the ARM facilities. The second track involved the formation of the Science Team. Because the science drivers were important to the design of the ARM facilities, a series of scientific workshops were held in the spring and summer of 1990 to clarify the scientific foundations of the program. In parallel, a solicitation process was initiated to establish the Science Team.
As these two tracks moved forward, features of the program emerged. One of the most significant was a pattern of collaboration with other programs. This collaboration was characterized on one hand by a series of joint field campaigns and, on the other, by involvement in program planning for other major research efforts. In the field collaborations, ARM attempted to bring a value-added contribution to another agency's or group's planned effort, while at the same time trying to gain operational experience necessary to guide its own field deployment.
This strategy resulted in collaborations with the Federal Aviation Administration's Winter Icing and Storms Program (WISP) and First ISCCP (International Satellite Cloud Climatology Project) Regional Experiment (FIRE) activities in Coffeyville, Kansas, and the Azores. In Coffeyville, early ARM concepts were tested in the Spectral Radiance Experiment (SPECTRE), jointly funded by NASA and DOE. It also led to ARM-fostered projects such as the Boardman-ARM Regional Flux Experiment, which tested key aspects of surface and surface flux characterization.
From the standpoint of planning, ARM attempted to gain early involvement in the program planning of other programs that would be evolving in parallel with it. Most notable among these planning collaborations was the GEWEX. One of these joint planning activities culminated in the field deployment of the Pilot Radiation Observation Experiment (PROBE) to Kavieng, Papua New Guinea, as part of the Tropical Ocean Global Atmosphere-Coupled Ocean Atmosphere Response Experiment (TOGA-COARE), in the winter of 1992-1993. Again, experience gained during TOGA-COARE has been a crucial influence in ARM planning.
A key convergence between science and facility planning tracks was the selection of a siting strategy for the ARM facilities. This process resulted in the identification of five locales in which ARM should locate its semi-permanent facilities and a comparable number of secondary locales in which the program should consider shorter, campaign-like activities. The primary locales in the order of their intended occupation were the Southern Great Plains (SGP) of the United States, the Tropical Western Pacific (TWP), the North Slope of Alaska (NSA), the marine stratus zones of either the Atlantic or Pacific Ocean, and the Gulf Stream. Deployment of the first instrumentation to the SGP site occurred in the spring of 1992, just 24 months after the program was approved. The site was dedicated in November 1992. Additional instrumentation and data processing capabilities have been incrementally added in the succeeding years. The TWP locale was the second facilities site implemented by ARM. TWP began phased operations in 1996 at its first facility on Manus Island. The second facility on Nauru Island was implemented in 1998. In 2002, a third facility in Darwin, Australia was established in collaboration with the Australian Bureau of Meteorology to support the other two sites. The NSA site at Barrow was dedicated in July 1997. Barrow is located at the northernmost point in the United States, 330 miles north of the Arctic Circle. ARM chose to place one of its three sites in the Arctic because the Arctic is particularly sensitive to climate changes. In the summer of 1999, ARM relocated instrumentation previously used for the SHEBA experiment to a new facility at Atqasuk, Alaska. At the same time, a new data collection system was installed at that location.
The ARM Observatory Today
The Lamont site is the central facility of the ARM observatory. It is designed to continuously sample all of the components of the radiation budget at the Earth’s surface and all the relevant constituents in the atmosphere above the site. To understand the size of this site, the site is composed of 29 facilities spread over an area of approximately 55,000 square miles (200 e-w, 215 n-s) and, nominally, 240 instrument systems representing over 930 separate sensors and instruments, producing 280 distinct data streams. Substantial efforts have been made or are under way to improve instrument performance. The radiometric calibration facility completed two additional cycles of broadband radiometer calibrations, calibrating a total of 152 pyranometers and pyreheliometers. Symptomatic of the distinctly different focuses of the scientific working groups of the science team, yet highlighting the complementary nature of the data needs of each of those groups, the largest event at the SGP site so far was the 1997 execution of a single IOP comprised of seven distinct areas of research involving between 70 to 100 scientists and colleagues of the science team and collaborating programs.
Improvement of instrument performance has been a major concern and the focus of a number of instrument related evaluation efforts. The delivery and installation of Vaisala ceilometers and atmospheric emitted radiance interferometers (AERIs) at the four boundary facilities completed the boundary facilities and permit studies to compare driving single-column model (SCM) research using remote sensors and satellite data in comparison to driving these models solely with radiosonde data.
Two shortwave spectrometers were installed in 1998. A reconfiguration of the broadband radiometric suite has proven to make that suite of instruments more robust and amenable to real-time monitoring and management. A zenith pointing narrow-field-of-view radiometer, based on the multi-filter rotating shadowband radiometer (MFRSR) sensor body, was developed and installed. The Raman lidar and 50-MHz radar wind profiler (RWP) radio acoustic sounding system (RASS) both suffered hardware and software problems that affected data quality. The RWP RASS problems were successfully corrected with a substantial performance improvement in its temperature profiling capability. The Raman lidar was repaired in the spring of 1998 before the Spring 1998 IOP. Additional instruments introduced during the year included a sky video for time lapse documentation of sky condition and two radiometer suites using radiometers developed by Francisco Valero. These radiometers were intended to operate continuously alongside the epply radiometers used generally by ARM, but it was anticipated that some components would not be durable enough for this style of operation. Once established, these instruments proved to be a valuable checkpoint for radiometric measurements made at all sites using epply radiometers.
A new 35-GHz cloud radar was tested and evaluated during intercomparisons to NOAA's Ka-band radar and the University of Massachusetts' 35- to 95-GHz dual channel radar. The new radar has proven to be as sensitive as the design desired, and validation activity has corrected several problems detected from data analyses and the intercomparison activity. Since its installation, the radar has proven to be extremely reliable, checking in at about 98% of the potential operating time. The radar was built by NOAA's Environmental Technology Laboratory (ETL) for ARM. Additional radars were built and installed at the NSA and TWP sites.
The balloon-borne sounding system (BBSS) continued to be the focus of efforts to evaluate apparent differences in water vapor measurements between the BBSS and other observational tools such as the AERI, the Raman lidar, and the microwave radiometer (MWR). Data intercomparisons and analysis following the two water vapor IOPs at the SGP suggested lot-to-lot variability in the calibration of sondes purchased from Vaisala. Since this was the first problem detected following the 1996 IOP, ARM has worked closely with Vaisala to ensure the highest quality sonde data.
ARM's cloud observational capability was substantially strengthened with the reconfiguration of the whole sky imager (WSI) to achieve both improved data transfer and availability as well as more reliable instrument operation. Adding CIMEL sunphotometers to the instrument suite as extensions of the CIMEL network supported by NASA is strengthening aerosol data taken at ARM sites.
In the TWP, the first ARCS, installed on Manus Island, Papua New Guinea, proved to be remarkably robust. From November 1996 through early 1998, operations were continuous with only occasional instrument problems and/or failures. The facility itself had few problems. The ARCS facility is a rugged, semi-autonomous system designed for use in such remote locations. Health-of-station data are transmitted daily to the TWP program office enabling the instruments and data system to be managed remotely. The bulk of the data is stored on tape and returned by mail periodically. The first ARCS installation was complete with the exception of two major instruments, the whole sky imager and the cloud radar.
The Nauru site was home to an ARM-sponsored international research collaboration effort in 1999. The NOAA research ship Ron Brown and the Japanese Marine Science and Technology (JAMSTEC) research ship Mirai measured surface and radiation fluxes at sea, for comparison with the land-based instruments and the TAO buoy array. Research resulting from this intensive month-long campaign is being used understand spatial variability and island effects. Data returned on tape is quality controlled by the site scientist before release and transmission to Science Team members.
At the NSA site, two major activities were undertaken in 1998. The first was completion of cold weather testing of the instruments for SHEBA and deploying those instruments to a Canadian icebreaker for deployment on the ice in the Beaufort Sea as part of the collaborative project. This was successfully done and by the time of the Science Team meeting, the ice camp had been operating for several months and had experienced many of the vagaries of working on the ice, including unannounced lead openings and polar bears. The deployment, overall, was being highly successful, and valuable data were acquired. For ARM, the data are returned on tape and quality controlled by the NSA Site Scientist prior to release and transmission to ARM Science Team members. At Barrow, the shelter and platforms had been put in place and instrumentation and data system installed. Planning was proceeding for full operational capability by the time of a planned collaboration with the NASA FIRE field program starting in April 1998. Significantly, the extended range AERI (ER-AERI) was installed and operating at both the SHEBA ice camp and at Barrow. The ER-AERI was built specifically for the high latitudes where low water vapor concentrations are common.
In addition to its longer-term undertakings, the ARM observatory provides opportunities for special intensive operational periods (IOPs), directed at specific science problems. For example, IOPs have been directed at understanding the accuracy of routine atmospheric measurements, and at specific cloud problems such as anomalous absorption of solar radiation. The IOPs have generally required expanded surface instrumentation and aircraft or unmanned aerospace vehicles. For many of the IOP efforts, the availability of the continuous ARM measurements is a unique asset.
Collaborations
Collaboration with the agencies and programs is a fundamental philosophy of ARM. Major collaborations included participation in the SHEBA effort, but also included efforts to support the NASA-USDA summer hydrology experiment at the SGP site. In December 1997, DOE committed ARM to support the GEWEX Water Vapor Program (GVap) by coordinating the establishment of the GVap ground-based validation network. This network was comprised of up to 20 sites worldwide that acquired water vapor profile data using advanced radiosondes. Another major international collaboration took place in the area of Nauru in the summer of 1999, following the installation of the Nauru ARCS. Collaboration also continue with the "Schools of the Pacific Rainfall Climate Experiment" and the "South Pacific Regional Environmental Program." SPREP has been a valuable source of help in talking to the governments of the island nations as well as planning and coordinating logistical support in the area. The use of the ARM observatory by these and other programs such as NASA and NOAA and other countries such as Japan and several European countries has exceeded expectations.
Science Team Strategies
Science Team research efforts largely fall into the two fundamental strategies through which ARM seeks to achieve its programmatic objectives and to focus its scientific efforts. These strategies are also the basic organizing principle behind defining the requirements for individual IOPs and determining what additional measurement capabilities are required. The first strategy, and the one that was at the heart of the priorities that led to the initial focus on the implementation of the SGP central facility, is the "instantaneous radiative flux" measurement and modeling effort. The second is single-column modeling to evaluate the cloud and radiative process models either used in, or being developed for, general circulation models being used for climate studies. A third focused area of activity, related to establishing the lower boundary condition for both SCM evaluations and instantaneous radiative flux (IRF) calculations, is the effort to characterize surface fluxes, surface radiative properties, and planetary boundary layer behavior on scales appropriate to general circulation models.
In the IRF strategy, the effort consists of collecting data on radiative transfer, the distribution of radiatively active constituents, and the radiative properties of the lower boundary. The radiative properties of the atmosphere and the lower boundary are used as input to radiative transfer models, including both detailed models with high spectral and angular resolution and simplified models suitable for use as parameterizations in climate models. The results produced by the models can then be compared with the radiation measurements as depicted in Figure 1.
The IRF approach is crucial to ARM, but it is not sufficient. Specifically, it does not address the large-scale processes that lead to cloud distribution and structure and the resultant cloud radiative properties that are important to understanding the instantaneous radiative fluxes. Using an SCM approach allows the testing of models and parameterizations intended to represent cloud property life cycles in general circulation model grid cells. Thus the fundamental idea of the SCM is to measure the external forces at work on a column of the atmosphere that corresponds to a single general circulation model grid column, to use transfer processes inside the column, and to evaluate the results produced by the models by comparing them with additional observations.
Figure 1. Experiment-based radiative model test scheme.
Science Team efforts focusing on the evolving nature and use of the measurement and modeling capabilities represented by the sites, their instruments, and the routinely running algorithms is largely centered in the activities of various working groups.
Accomplishments
As the quality of ARM data has improved over time, the focus of research has shifted as well. Much of the initial effort was focused on instrument development and basic radiative transfer theory. This gradually gave way to studies of atmospheric processes. This has continued to progress into comparing models with observations and evaluating parameterizations.
The ARM Program has formed some surprising relationships along the way. For one, the ARM Program works in close cooperation with weather forecasting operational centers. The weather centers have found that ARM data is extremely helpful for evaluation and improving their own models. And, in turn, they have supplied research support and ideas for the ARM community. In general, the use of the ARM observatory has exceeded initial expectations, and resulted in some great and unexpected partnerships.
The ARM observatory has become an integral part of international collaborations and of U.S. government research programs sponsored by agencies such as NASA and the NOAA. Although SGP remains the primary facility, data from the remote ARM sites at Alaska and in the Pacific are used to study polar and tropical environments. Japan, Australia, and several other European countries have expressed interest in creating observatories like ARM.
This worldwide interest speaks well for the future of ground-based remote sensing for climate modeling and weather forecasting. Along with continued observation at the existing ARM facilities, the program is building a mobile facility that will extend the types of measurements the fixed sites currently perform. The remote facility will be able to perform anywhere on Earth for periods of up to a year.
References
Ackerman, T.P. and Stokes, G.M., 2003: "The Atmospheric Radiation Measurement Program." Physics Today 56(1): 38-44.
U.S. Department of Energy, 1996: Science Plan for the Atmospheric Radiation Measurement (ARM) Program DOE/ER -0670T, Washington, D.C.
