Estimated radiative forcings between 1850 and the present. Hansen, J., M. Sato, A. Lacis, R. Ruedy, I. Tegan and E. Matthews. Climate forcing in the industrial era. Proc. Nat. Acad. Sci., 95, 12753-12758, 1998
-- Renny Greenstone (rgreenst@pop900.gsfc.nasa.gov), Raytheon ITSS
Note: Direct access to many of the presentations made at this meeting may be achieved by accessing the World Wide Web at the following URL: http://eospso.gsfc.nasa.gov/eos_homepage/misc_html/iwg_durham_98.html
The 14th meeting of the Investigators Working Group (IWG) of the Earth Science Enterprise/Earth Observing System (ESE/EOS) took place at the New England Conference Center in Durham, New Hampshire (the home of the University of New Hampshire), October 19-21, 1998.
Morning Plenary Session (Earth Science Enterprise/EOS/EOSDIS Status)
Byron Tapley (University of Texas, Austin), chair of the IWG Science Executive Committee and session chair, opened the meeting and then gave the floor to Ghassem Asrar, NASA's Associate Administrator for Earth Science.
Ghassem Asrar (NASA Headquarters) reviewed the "long path from the beginnings of EOS to where we are now." He pointed to a major change from SeaWinds to QuikSCAT, calling it a "fantastic recovery," and then he went on to discuss two sets of challenges that face the EOS program for the next few years: 1) we need to establish the benefits of EOS science and educational activities that will come from the first series of EOS satellites, and 2) we need to establish the follow-on program for the post-2000 time frame. He pointed out that by accomplishing challenge number 1 we will build confidence in the achievements to be expected in meeting challenge number 2.
In any event we have learned that we cannot repeat the first EOS series as originally planned. Consequently, we have initiated a process on how to do the next series. The process began with a Request For Information (RFI) and a resulting workshop held recently in Easton, Maryland.
The first EOS series cannot be repeated because of the budget process. We had assumed that there would be 30% efficiencies thus setting aside funds for the second series, but that did not materialize. We wouldn't want to repeat the first series because we've learned new science, which has to be factored in. The old plan was too rigidtechnology has moved on. Another factor is that we have an expanded community of Earth scientists with greater needs for exchange of Earth science data and information. Still another factor is the need to deal with practical problemsnew customers are very vocal, acting through the Congress, in asking us to show that we are meeting their needs.
Our general philosophy calls for us to adopt two sets of principles for the next decade of EOS: 1) We must move from an observational paradigm to a science hypothesis/question paradigm. Instead of our current measurement strategy where observations are made in support of general scientific questions, we now think of NASA's responsibility as answering key and specific scientific questions. Our approach must not be just space-based but must also include suborbital flights and end-to-end modeling. 2) We need to achieve and maintain a balanced NASA Earth Science Program.
We will now have four program elements and we need to achieve a balanced investment portfolio among them:
Asrar pointed out that NASA's Research and Analysis (R&A) program has been the foundation of its science activities. Lately, however, there has been concern over the health of R&A. Now, Headquarters is restoring support to the R&A program to rebuild it back to its 1994 level by the year 2000. At this time 18-20% of the $1.4 B ESE program goes to the R&A program. The intent is to build up R&A support to 25% of the ESE budget in the next 7-to-10 years. Along with this, the ESE Science Implementation Plan process will continue.
(This program element is now the basis for a new Division.) This Division consults users of ESE services and works with private practitioners as well as representatives of the States. Its mission is to bring out the practical benefits of ESE science, particularly in light of the current budget constraints. At this time 5% of the ESE budget goes into this area but there is a lack of focus. (Asrar wants to see it grow to 10% in the next several years.)
In this program, ESE is trying to decrease mission development times to 2-to-3-year periods and at the same time to lower risk. We expect an overall cost reduction of at least 20% over the first EOS series. This involves improvements in spacecraft technology calling for industry investments. Also, there is a program to develop smaller/more efficient instruments. The "instrument incubator" program now has seven candidates under development. Two of the new instrument concepts are soil moisture and ocean salinity. We intend to increase our budget in this area from 6% to about 10% over the next several years, focusing on short-to-mid-term science objectives, with NASA-wide technology supporting long-term developments.
This program deals with the use of EOSDIS on the ground and various elements of the space segment. First, Asrar focused on the DIS (ground) elements of the program.
ESE has a commitment to full use of the EOSDIS investmentthere is no alternativebut possibly not making use of all the planned functionality, perhaps using about 80-90% of what had been planned. The first increment of EOSDIS is working to support the Landsat and AM-1 missions, but may not accommodate the further missions, PM-1 and CHEM. The Office of Earth Science (OES) is working to identify resources to support PM-1 and CHEM. There is also an internal NASA team studying what is to be done for the decade to come. (The PIs are expected to play a more-active role in the decision process for future developments.)
Asrar said that the EOSDIS Core System (ECS) will continue to perform the functions of data capture and initial processing to Level 0, and Level 1 processing would be performed either by the DAACs or by the PIs associated with the EOS instruments, this to be decided on a case-by-case basis. Generation of higher level data products will be performed principally by the PIs.
The space segment will undergo the most changes in order to avoid crises and the possible loss of support that might follow. Asrar noted that the first EOS series was general purpose, with technology innovation not a particular concern. Maintaining stability of observations between the first series and what follows is an obstacle to innovation. Asrar feels that we must separate our goals and presented five dimensions to be taken into account:
1) Determine what data are really needed. The initial program had a 15-year data-accumulation effort, but suffered from unnecessary constraints. (At the RFI review there was an attempt to identify what was really needed for 15 years and, possibly, beyond.) Systematic measurements in this vein must be continued. They may continue measurements from the first EOS series or they may begin systematic measurements of other variables, based on the growth of our scientific understanding, e.g., precipitation.
2) Experimental objectives must include short-term demonstrations of new capabilities. These may be carried out through the Earth Probes program.
3) There must be a transition from research experimental to operational. (At present there is no formal process to make this transition; we tend to follow an ad hoc approach.) The OES is now working with the Integrated Program Office (IPO) of NPOESS to define a bridge between AM, CHEM, and NPOESS for the period 15-to-20 years from now.
Asrar pointed to the development of the next Landsat instrument (ALI), which has already been built and is now in thermal vacuum test. It is 3-4-times better than ETM+ and is slated to be launched late next year. For ocean-related systems there is planning to effect the transition from experimental to operational going on right now, but at this time there is no mature system available. TOPEX, NSCAT, and SeaWinds should be transitioning to operational systems. NASA can't continue to support systems like these for the long run and, to this end, there are dialogues with the Navy, NOAA, and international organizations to have them take over operational support. If this doesn't happen, NASA may have to drop out of these activities. Asrar pointed out that NPOESS can serve as the "venue" for transitioning ocean studies into the operational world.
4) There must be major support to University-developed missions. Asrar referred to a University-based Education Earth System Science (UNESS) micro/small-satellite program intended to engage graduate students and their advisors in NASA's Earth Science program.
5) The New Millennium small-satellite program is intended to demonstrate key enabling technologies and their associated risks, prior to their integration in science missions.
In what he called "side remarks," Asrar congratulated TRMM and TOPEX/Poseidon for their remarkable successes in articulating the benefits of NASA missions to society at large. They have been of tremendous help in overcoming budget cycle difficulties relating to EOSDIS and AM-1 problems. In contrast, he said that the EOS community is falling behind in its outreach activities.
On December 6 at the AGU meeting, Dan Goldin is to articulate his vision statement for Earth Science. This will be the first official NASA commitment to the Earth Science program. Asrar invited all the participants in EOS to help in guiding the EOS process for the next decade.
There is an issue with uncosted commitments. Congress says that this problem must be fixed or "they will do it for us." Right now, $550 M is being held as uncosted commitments. EOS stands to lose $150 M if the money is not spent by September 1999. It is urgent that university billing be on time.
In regard to the new Commercial Space Act, we will show clearly that we will not do something that industry states it can do.
Other comments made by Asrar were these:
1) The PI-led mission will be the preferred approach for mission management.
2) There is no national strategy for transitioning experimental scientific observations into the operational weather-monitoring system, and, in the absence of such a strategy, we are working very closely with the IPO to see that NPOESS flies at least the following instruments: an advanced Earth surface imager, an atmospheric temperature and humidity sounding package, an ozone column profile monitor, and a total solar irradiance monitor. We are also interested in seeing that ocean altimetry and scatterometry are transitioned into the converged operational system.
3) We are investing in the area of ocean data assimilation, seasonal-to-interannual prediction modeling, and state-of-the art supercomputer capability toward meeting this objective.
Asrar ended his talk by saying that he is committed to making this new Enterprise strategy happen, and that he is looking forward to working with all those present to ensure our continuing success.
Berrien Moore (University of New Hampshire) described the current status of the "National Academy of Sciences (NAS) Pathways Report." The report is the product of the NAS Committee on Global Change Research, Board on Sustainable Development, Policy Division. Moore said that the "overview" document for the report is now available from National Academy Press. (The overview carries the title: "Global Environmental Change: Research Pathways for the Next Decade." The full report is still not ready for distribution.) He urged "careful" reading of the Pathways Reportevery word in the report was debated. He said that we must examine the "lessons learned" and added that it stressed the criticality of observations. In the science arena the report presents six areas of focused research: biology, two climate studies, paleoclimate, chemistry, and human dimensions. Then there are five "themes" for each of the six focused research areas. Finally, each of the five themes has five science questions. There is a recommendation to focus efforts on key science questions.
Overarching in the report were recommendations to the USGCRP that it should work on the carbon cycle and the hydrologic cycle to understand and document them; that it should work on modeling the climate system (there is a need to understand 30-to-50-year changesthe climate system transient response, and not just the equilibrium response, must be understood); and that it should work on radiative forcing and chemical changes, particularly involving the roles of aerosols and methane. Another recommendation was to strengthen the technology for in situ observations, which are falling apart around the world.
The following recommendations cited by Moore were not in the short "Overview" report , but will appear in the main document, due to be released in an on-line version in early November:
1) There should be a long-term (more than 15 years) coherent measuring system, e.g., the Keeling records of CO2. NASA cannot do these long-term measurements, which must transition to NOAA.
2) The Data and Information System (DIS) needs to be more flexible in its response to new technology. The Committee says that it will be watching the progress of NASA's "federated" approach.
3) The report notes that the U.S. is not the lead country in modeling, but it is good that other countries are taking the lead. The report also takes note of the progress made at the Hadley Center in modeling the atmosphere with full chemistry.
Moore cautioned that the overview report has very little to say about the critical issue of clouds and radiation, but that the full report has extensive information on this.
Moore stated that the Academy will be trying to establish its role in seeing that the Pathways report recommendations are carried out. Asrar added that there will be a new committee to push implementation of the recommendations, and that there may be Congressional hearings related to this.
Pierre Morel(NASA Headquarters) presented the latest information on the "EOS Second Series: Status and Plans." He said that NASA wants to operate with all "due process." His talk would focus on flight mission planning. There is a need to re-validate the linkage between Earth Science questions and research priorities and measurement priorities that will lead to the most essential measurements. He referred to the RFI review session held in Easton, Maryland (August 24-26) that was previously mentioned by Ghassem Asrar as the occasion where the plans for the second series were reviewed.
The RFI had been issued in May 1998. Six discipline panels reviewed the proposals that had been received in response to the RFI. 23 flight mission concepts came from the review, and they can be found on the World Wide Web. The second step in the process was to review the mission concepts using an industrial "cost model." A "nominal mission scenario" was developed and reviewed by a panel of 33 "distinguished scientists" headed by Charles Kennel. (The report of the RFI review is available at URL: http://eospso.gsfc.nasa.gov/eos_homepage/misc_html/intro_kennel.html.
The outcome of the RFI process was the conclusion that there is a need to identify the key systematic measurements required for the next decade. Three specialized mission types were identified: 1) EOS systematic follow-on measurements leading to operational systems as appropriate; 2) Earth Probes (also referred to as Discovery missions); and 3) New Millennium Program missions bringing in new technology. It was also stressed that it is essential to transition measurements to NPOESS, and that NPOESS should develop research-grade measurements programs.
Morel gave what he called "Priorities at a Glance":
1) Atmospheric Chemistry
2) Atmospheric Physics and Radiation
3) Hydrology and Mesoscale Weather Processes
4) Oceans and Ice
5) Land Cover and Terrestrial Ecosystems
6) Geodynamics and Natural Hazards
Morel showed the response to the panel recommendations in the form of "Timelines for Systematic Measurements" for the period 1999 to 2011. In his charts he showed the flow from EOS measurements into NPOESS via "bridging" measurements. He noted that performing stratospheric chemistry measurements will bring the International Space Station into play. There is a hope to be able to buy radar imaging commercially. TRMM's measurements will become part of the International Precipitation Mission in 2003 under Japanese leadership and then become global in 2009.
There are to be eight Discovery missions (Earth Probes) starting in 2001, with the possibility of supporting three such missions in the course of two-year intervals. Morel listed seven such missions saying, however, that there is no commitment to carry them out:
Skip Reber expressed concern that through the Discovery missions with their single objectives we are losing out on the simultaneity that is often wanted by the IDS investigators. Morel responded that the models can serve the function of integrating non-simultaneous measurements.
Berrien Moore called attention to the ill-defined geostationary observing missions. Morel responded that we are aware of this and working on it.
Following the morning break, Chris Scolese (Goddard Space Flight Center) reported on the "Status of the GSFC EOS Missions." He said that recovery activities are under way to deal with the one-year delay of the AM-1 launch that has been imposed by difficulties with the Flight Operations System (FOS). The spacecraft hardware will basically be ready in January, and it is expected that a launch date will be set in two months. CERES may replace its voltage converters, depending on the outcome of the TRMM instrument study. MODIS is going to fix the crosstalk between detectors in the readout electronics and is deciding whether to go back and do thermal vacuum testing and re-calibrate with the new electronics. There have been budget impacts affecting both ICESat and EOSDIS.
Scolese listed major issues, which include developing a control center for AM-1, providing the necessary funding for ICESat and EOSDIS, and establishing working agreements for Landsat 7 operations and data processing. He referred to reorganizing activities taking place at both GSFC and OES. Both algorithms and instruments are the responsibility of GSFC.
Landsat 7 has been assigned a launch date of April 15, 1999, and this is being tracked. AM-1 is to be launched in the third quarter of 1999, but this awaits finalizing of the FOS recovery plan.
Scolese showed pictures of the AM-1 spacecraft and the instruments and showed the launch vehicle on its stand at the launch site. There are still some problems affecting the status of the AM-1 launch.
The PM-1 launch is affected by the lack of instrument contingency funds. The spacecraft I & T is to begin in June of next year.
Regarding ICESat, Scolese said that the GLAS instrument is in good shape. GSFC has total responsibility for the mission. The launch vehicle has yet to be selectedas a matter of preference it could be either Athena II or Taurus XL. Unfortunately, there is a $15 M cut in the FY 99 budget threatening the ICESat schedule.
On a happy note, he said that CHEM-1 is "moving along."
Summarizing, Scolese said that considerable progress has been made overall in the last year. The FOS is still the pacing item for the AM-1 launch date. An agreement on a specific launch date is to be worked out in the next few months. The spacecraft must be ready for launch early in 1999 to maintain its place on the launch schedule.
Dolly Perkins (Goddard Space Flight Center) reported on EOSDIS. She referred to the EOS Polar Ground Network, the EOS Data and Operations System (EDOS), and the associated networks.
Concerning the Science Data Processing Segment, she said that, by launch date, all Landsat 7 and AM-1 data will be provided though Level 1. The Emergency Back-up System (EBS ) is ready for use as needed.
Since 1997 the ECS contractor (formerly Hughes and now Raytheon) has been having problems, partly due to heavy attrition (greater than 35 percent). Some of the ECS functions are now being carried out by other elements of EOSDIS. The instrument teams are now taking on more of the instrument data processing, and budget problems await us in the out years.
Since March there has continued to be slippage of the Flight Operations System (FOS). As a result the command execution system is being replaced by a new commercial system called ECLIPSE. Facing a budgetary challenge beyond the AM-1 launch time, ESDIS has considered four alternatives. The selected "option A+" minimizes the 1999 budget impact and relaxes some of the requirements. It calls for adding capabilities incrementally over four releases occurring in FY 1999 and 2000. The requirements that were dropped can be restored at a later date, provided that the budget permits this. Generally, the system is becoming more adaptable over time, permitting the use of new technologies and processing approaches.
Three times processing of the data is now being considered. "One-stop shopping" is still in place.
In answer to a question, Michael King replied that there has been some full-load testing with MODIS, and full end-to-end testing will start in November.
Asrar said, if the PIs can do the data processing job within their budgets, they will be permitted to do it. Jim Dodge commented that archiving and distribution of the data are still to be carried out by the DAACs.
Byron Tapley was concerned that network bandwidths might be constricting. Perkins replied that we are considering bandwidth for data entering the system and have yet to look at the bandwidth for data coming out of the system. Asrar said that we never intended to provide full data access to the outside community. He stressed that any requirements from the outside community with budgetary impacts must be sent to Headquarters for resolution.
Afternoon Plenary Session (TRMM Early Results and Lessons Learned). The session was chaired by Chris Kummerow (Goddard Space Flight Center)
John Wilding (Goddard Space Flight Center) was the leadoff speaker. Wilding's topic was "SeaWiFS Data System Experiences and Lessons Learned." Chuck McClain made a few introductory remarks before Wilding gave his presentation. McClain said that mid-September was the first anniversary of routine SeaWiFS operations, and they are now able to provide real-time products. There are four tightly coupled elements to the SeaWiFS ability to deliver products: data capture, mission operations, calibration/validation, and core processing. McClain suggested that this mission approach might serve as a prototype for other missions to come. The system is adaptable to multi-systems. He explained that Wilding has been in charge of keeping the processing system going, from capture through to the DAAC.
Wilding explained that the system requirements were to have high automation, high flexibility, and be able to do reprocessing readily. The main elements of the system are the scheduler and the visual database cookbook (VDC). They use a relational database.
Among the key system features listed by Wilding were its ability to serve as a MODIS emergency back-up. The system offers high adaptability, scalable architectureit was adapted to OCTS in just two weeks, and it was possible to create a fully functional prototype for MODIS in less than a month.
The primary lesson learned was the value of having all project elements in a collaborative working environment. In summary, Wilding stated that SeaWiFS is now a complete end-to-end system. McClain added that the great asset of SeaWiFS is its ability to provide a system for users of other platforms. Users of the SeaWiFS system can specify run corrections. They just submit them as "special recipes."
Chris Kummerow (Goddard Space Flight Center, TRMM Project Scientist) was the enthusiastic lead-off speaker for "TRMM Overview of Early Science Results: Hot Gems." He explained that the three precipitation sensors were the TRMM Microwave Imager (TMI), the Precipitation Radar (PR), and the Visible and Infrared Sensor (VIRS, like AVHRR). LIS and CERES are also on the spacecraft. The first TRMM images were available on-line on December 8, 1997. He showed some images from VIRS over Florida and a comparison of PR images with ground radar over a squall line. Using TMI, brought out the amount of graupel to be found in oceanic storms. The 10-GHz TMI channel was used to study sea-surface temperatures during the recent El Niño. The greatest differences among the TRMM algorithms occur in the deep tropics, and were found notably in the February 1998 data.
There have already been tremendous improvements in the algorithms since the first TRMM data became available. Now the Science Team is working very hard to build strong consensus algorithms to ensure the TRMM legacy. The Science Team is hoping to get down to 10% anomalies between rain estimates, but further research will be needed to do better than that. Planning for TRMM science communications has really paid off, making data readily available not only from the regular data stream, but from a new "real-time" stream as well.
The Science Team has been checking out the algorithms that were applied to warm rain over land and found no glaring errors. The topographic errors occurring in mountainous areas have been "cleaned out."
Bruce Wielicki (Langley Research Center) discussed "CERES on TRMM Early Science Results." He noted that the first CERES data from the TRMM spacecraft became available on December 27, 1997. He said that the calibration accuracy and stability of the CERES radiometers exceed the specifications. Navigation accuracy is about 1 km, and CERES calibration is 2-to-10 times better than the ERBE calibration. Over the first 8 months, no change in instrument gain was found within 95% confidence limits of between 0.1 to 0.3%, depending on the channel, an unprecedented level of stability for Earth-viewing radiometer data. Level 1b CERES radiance data (calibrated and navigated) have been in the archives since July 1998, roughly a factor of 4 faster than was done for ERBE.
The CERES-archived "ERBE-like" TOA flux data currently appear to have better accuracies than ERBE. The data for January through August 1998 are available in the Langley DAAC.
It appears that there could be a real climate change of 4 W m-2 in outgoing longwave (LW) flux for the +/- 20° latitude zone as measured by ERBE and CERES, from the ERBS period (1985-1989) to the present. Unfortunately, there is an observational gap of 8 years between the current CERES data and the previous ERBE scanning radiometer data from NOAA and ERBS missions, but the ERBE wide-field-of-view (WFOV) instrument on the ERBS spacecraft is still functioning after 14 years. The ERBS WFOV data indicate that this change occurred in early 1991, before the Pinatubo eruption. The SCARAB data from 1994 are also consistent with such a change.
Early CERES data have shown that the maximum albedo of deep convective clouds is roughly 0.74, and the albedo of precipitating deep convective clouds (identified by the TRMM radar) is about 0.66. Theory for small ice particles predicts a maximum of about 0.79. The drop in albedo for precipitating clouds may be due to an increase in particle size for the
precipitating clouds or may be due to lightning production of SO2 in the upper cloud.
As of September 1 there was a CERES voltage-converter anomaly which caused
one of the voltage converters to be sensitive to temperature. The CERES instrument has been shut down for an investigation of the impact of this sensitivity on instrument lifetime. Results of the investigation should be available by late November, with a decision then on future operations of the instrument.
Hugh Christian (Marshall Space Flight Center) discussed "LIS on TRMM Early Science Results." He explained the physics of lightning-causing clouds. In order to have lightning there must be graupel particles plus ice crystals plus supercooled water and 7-8 m/s updraft velocities in the cloud. Also, there most be some mass above the zero-degree isotherm. In the updraft the ice particles become positively charged and the graupel become negatively charged. As they separate the potential difference leads to lightning. Thus, lightning is well correlated with updraft intensity and with cold-process rain. Lightning can be seen to be a measure of how energetic cloud cells may be.
The Optical Transient Detector (OTD) is a precursor of LIS and is still flying after three years in orbit. LIS is a more-sensitive instrument than OTD.
Using OTD and LIS it has been shown that equatorial Africa is the hottest lightning spot on Earth. OTD observations have shown lightning occurrences at latitudes higher than 70 degrees north. The preponderance of lightning occurs over northern hemisphere land. The global flash rate has been found to be about 40 flashes per second, and this is just about half of the old estimate. This lower flash rate thus implies lower NOx and thus lower ozone production. Intracloud lightning dominates. An interesting finding is that continuing-current discharges may be a warning signal for forest fires.
After an afternoon break the TRMM session continued with a presentation by Arthur Hou (Goddard Space Flight Center) on the "Impact of TRMM Precipitation and Moisture Observations on DAO's Assimilated Global Data Sets." He stated that the Data Assimilation Office (DAO) is the first among the global data producers to attain rainfall and total precipitable water (TPW) assimilation capabilities. Assimilating TMI (from TRMM) and SSM/I rain rates and TPW data is effective in improving the hydrological cycle and atmospheric energetics in the Goddard Earth Observing System (GEOS) model analysis. Errors in the outgoing longwave radiation (OLR) in the GEOS analysis are reduced by approximately 20% in bias and 25% in rms error between 20°S and 20°N. As a result of adding TRMM data, the GEOS Data Assimilation System (DAS) now gives better tropical precipitation forecasts beyond one day.
The last speaker for the first day of the IWG meeting was Erich Stocker (Goddard Space Flight Center) presenting "TSDIS Experiences and Lessons Learned." Like SeaWiFS, the TRMM Science Data and Information System (TSDIS) is database driven. Even more than in SeaWiFS, commercial software systems are used. Errors are viewed through a system called OVERVIEW. TSDIS has used a commercial database called SYBASE and has found some problems with it in the form of page "lockups." TSDIS has no responsibility to provide data to the general public.
Among management lessons brought out by Stocker were these:
Stocker said that in the world of TRMM "real time" refers to data that are less than three hours old at the time the user gets ftp access. In fact, at least 80 percent of the time the data are less than two hours old at the ftp site.
TSDIS now includes four government people and 26 full-time equivalent (FTE) contractor people.
Morning Plenary Session (Anomalous Absorption of Clouds and Regional Aspects of Global Change). The session was chaired by Michael D. King (Goddard Space Flight Center).
Robert Cahalan (Goddard Space Flight Center) dealt with the question: "Is the Sun's Heating of Earth's Atmosphere Excessive?" He began by referring to an FY 96 Report to Congress which referred to three papers in a 1995 issue of Science, by Cess et al., Ramanathan et al., and Pilewskie and Valero, dealing with atmospheric absorption of solar radiation. He then defined such terms as total absorption, cloud forcing, cloud forcing ratio, and "excessive."
Various other studies in the period 1975 to 1997 found solar absorption by the atmosphere to be typically near 65 W m-2, but Cahalan showed a second group of four previous studies that found results closer to 85 W m-2. So there was a bimodal distribution of results, with most finding
approximately 20-percent absorption by the atmosphere (~65/340), but a second group 20 W m-2 higher, or about 25-percent absorption (~85/340).
When plane-parallel theory was applied to absorption by clouds, the results appeared to be insensitive to cloud amount or aerosol, and there was a negligible excess absorption in the visible. Cahalan said that there had been studies by Li and colleagues showing variations of absorption with latitude, with a large excess near the equator, little or no excess at mid-latitudes, and a deficit in absorptance at high latitudes. Work by King and colleagues, using the Cloud Absorption Radiometer (CAR), did show some excess absorptance in water vapor window regions in the near infrared, but the amount was not comparable to the very large numbers reported in the 1995 Science articles. These CAR measurements apply to absorptance deep inside thick clouds, in the so-called "diffusion regime" of radiative transfer. While these in-cloud diffuse measurements found no large excess, significant excesses in diffuse radiation at the surface have been reported. The explanation for the excess diffuse radiation at the surface is still being debated.
Cahalan went on to mention a two-aircraft experiment that was conducted at the DoE/ARM Southern Great Plains site to check directly on the excess absorptance problem in clouds. Two aircraft flew simultaneously, one above and the other below the cloud. There was a problem in keeping the two aircraft and their radiometers level. This was important because the flux determinations are very sensitive to the radiometer view angle. However corrections were made to determine what the flux would have been in level flight, and the results show an excess that increases with cloud amount, with the largest absorptance exceeding 35% during a flight on October 30, 1995.
Cahalan also mentioned the 3D errors that affect theoretical calculations. These are large enough to often make local absorptances appear to be negative, due to net downward radiation from cloud sides. However, Cahalan said that these effects have been carefully simulated and can be
effectively removed. After 3D errors are removed from the observations, there remains a real 3D correction to absorptance, but such corrections are typically only on the order of 1%. Also, 3D effects have characteristic signatures which depend on cloud type and sun angle, and in some cases 3D effects can decrease the all-sky absorption when compared to clear sky, as well as increase it. By contrast, the 1995 Science studies found no spatial or spectral signatures, and some later satellite studies also find no dependence of excess absorptance on latitude or season. Cahalan reported that no physical hypothesis is known to explain this lack of spatial or spectral signatures found in many of the empirical studies. Cahalan feels that there is a need to quantify expected signatures of absorption due to various 3D cloud effects, aerosols, drizzle, and the radiative continuum, as well as to examine a wide variety of 3D cloud situations, in combination with aerosols and drizzle.
Cahalan's conclusion was that there is a likely excess absorptance of about 5% (10-to-15 W m-2 ) with an unknown impact on climate sensitivity, and that future studies will need accuracies of 1% (~3 W m-2) in order to distinguish the various physical mechanisms which might be responsible for the solar absorption excess in the cloudy atmosphere.
Stephen Schwartz (Brookhaven National Laboratory) reviewed experimental work on "Atmospheric Absorption Inferred from Sun and Sky Photometry." He began by saying that he was concerned with possible aerosol effects on the solar irradiance budget. Working mainly with surface measurements at the DOE Southern Great Plains ARM site in North Central Oklahoma, his group found that direct normal solar irradiance (DNSI) at the surface is most sensitive to aerosol optical thickness, t. Sun photometry with "Langley plots" was used to determine values of t, permitting comparisons of measured and modeled DNSI. The findings were that it is necessary to know aerosol optical depth with great accuracy in order to model DNSI accurately, and that there is no unrecognized atmospheric absorption.
In part II of his talk Schwartz presented work on the diffuse downwelling irradiance (DDI), measured with a shaded pyranometer. The finding was that "measured" equals "modeled" at high-altitude sites but not at low-level sites such as the Southern Great Plains ARM site or Saskatchewan. At low altitudes the model results were systematically higher than the measured results. He stated that "something" seems to remove the power from the diffuse radiation! He suggested that this absorption masquerades as an "enhanced" aerosol optical depth, and further that such absorption is inconsistent with typical properties of tropospheric aerosols and with measurements of aerosol single scattering albedo made at the Oklahoma site. Something other than aerosols is responsible for this atmospheric absorption, of magnitude about 0.02 optical depth. AERONET results from 80,000 measurements indicate that such a magnitude of minimum extinction is widespread.
[An overview kindly provided by Bruce Wielicki gave these insights into the Schwartz talk: Diffuse radiation in clear skies appears to be off by an equivalent aerosol optical depth of 0.02, or an equivalent global mean clear-sky flux of 17 W m-2. The difference does not appear to correlate with aerosol optical depth, as it should, if it is explainable by an uncertainty in aerosol single scatter albedo. The difference also did not appear to correlate with column water vapor.]
Jim Hansen suggested that perhaps the single scattering albedo value could be changed, and Schwartz agreed that this should be considered, especially in instances of high aerosol optical depth.
Tom Charlock (Langley Research Center) addressed a "Comparison of Computed Shortwave Fluxes with Observations from Surface Radiometers, CERES, and GOES over ARM SGP (Oklahoma and Kansas)."
Charlock began by saying that we lack information on shortwave forcing of anthropogenic aerosols and changes in land use. The CERES ARM GEWEX (CAGEX) experiment was a web-based exercise conducted "pre-launch" with access to input data, computed fluxes, and measurements. Charlock showed data from October 11, 1995 on surface insolation vs. time. The diffuse radiation exhibited large errors (it was not well calibrated). The measured diffuse radiation was low whereas the calculated radiation was high. The measured diffuse radiation did not correlate well with the measured precipitable water. Clear-sky measurements and calculations for October 11 showed no discrepancies, but huge discrepancies were found for cloudy-sky conditions.
The CAGEX ARM Enhanced SW Experiment (ARESE) was conducted in the period September 25 to November 1, 1995.
Generally, large cloud absorption was found in this experiment. October 30, 1995 was a cloudy day. Good agreement was found between the measured and the calculated radiation at the top of the atmosphere (TOA). However, the mean radiation calculated for the surface was much too high ~100 W m-2.
Cloud profiles were obtained using radar, microwave radiometers, and lidars for different cloud conditions, but still discrepancies of ~100 W m-2 were found.
Charlock reported on CERES data from January to June of 1998 and compared the data with results from the CERES ARM Validation Experiment (CAVE) with the conclusion that ERBE data have an uncertainty of about 5-to-10 W m-2 . The surface measurements are generally not consistent with the model calculations.
Charlock pointed out that the sorts of disagreements found at the ARM site were apparently not applicable to data taken at high altitude at Mauna Loa. There seems to be the need for better calibration of surface solar insolation measurements. Direct-beam solar insolation measurements seem to be well calibrated, but diffuse radiation measurements are still a concern.
Steve Wofsy (Harvard University) asked "Are Forests Really Important Sinks for Atmospheric CO2?" He showed the trends for CO2, for 1998 to 1990, and pointed out that the atmospheric anomaly was low. In the period 1980 to 1990 there has been a source of 6.3 to 7.3 Gt C/yr from fuel consumption and deforestation, which, taking atmospheric uptake into account, leads to a missing sink of 1-to-2 Gt C/yr. Keeling and coworkers have been measuring oxygen concentrations over the Pacific and find that oxygen concentrations are changing more rapidly than CO2 concentrations are changing. They find that land and ocean uptakes are about equal.
In the period between 1800 and 1950 the biospheric addition of CO2 to the atmosphere has been significant. (This finding is based on an analysis of the amount of 13C present in the atmosphere.) In more-recent years it appears that the biosphere has been taking up carbon.
Wofsy referred to a Science article by Tans (the next speaker) in which there is this quote: "North America sops up 1.7 petagrams C/yr." He then quoted from the "Birdsey Forest Inventory Plots," which showed that in 1950 U.S. forests had an increase in carbon. He commented that there are widely differing numbers as to the rate of increase of forest carbon intake.
Wofsy has looked at data from two very different sites, one is in Manitoba (a boreal site with trees about 75-90 years old) and the other is the Harvard Forest (with trees about 60 years old). Some comparisons between the two tree stands were these:
Factors affecting the sequestration of carbon differ at the two sites because of their different surfaces. Warming over the last 30 years has affected the two sites differently.
Wofsy concluded that terrestrial land-use history is most significant in determining the role of forests in carbon sequestration. Climate warming is a very important factor for midlatitudes. It is important to get more information on the ecosystems.
Steve Running commented that this work shows the importance of vegetation disturbances. We need more satellite data to learn more about this. He pointed out that the tower sites are giving reasonable magnitudes for carbon release. Bob Dickinson added that land-use change should be put in models along with trace-gas histories for the past century.
Pieter Tans (NOAA) had for his subject "Keeping Track of Atmospheric Trace Gas Budgets." His group (CMDL) has the assignment to keep track of the greenhouse gas budgets. He displayed a chart giving the current CMDL flask sites and towers along with the aircraft used for monitoring of trace gases. They have enhanced coverage for marine boundaries and are now branching out to get more land measurements. The CO2 program goes back to 1976.
A recent Nature article reports on methane measurements for the period 1984-1986 and shows a declining growth rate. New global measurements of fossil fuel-derived CO2 in 1992 and 1993 show large uptake in midlatitudes, mostly terrestrial. It is very difficult to distinguish biospheric from oceanic uptake of CO2 using the 13C/12C ratio. (An accuracy of 0.01/mil would be required.)
CMDL uses a 2D transport model to go from concentrations to identification of sources and sinks. In another chart he showed that, omitting fossil fuels, there is a net uptake of CO2 in the northern hemisphere for mid- to high-latitudes. CMDL has also used a 3D transport model and applied it to 14 "source regions." They run the model in monthly time steps. Using the model brings out the sparsity of data. They have used an estimated marine boundary layer time series to fill in data missing from other records, but there are problems with this method in some parts of the world. Tans has concluded that column-averaged CO2 should be used in CMDL calculations.
Tans has now formulated a desired data-gathering program. There should be 3 x 3-degree grid cells with monthly means. Accuracy per column should be 0.1 ppm. Aircraft should be used to obtain in situ profiles along with satellite laser measurements at 1.58 micrometers using DIAL. Steve Wofsy commented that diurnal measurements are needed from the satellites, and Tans replied that perhaps the satellites would only provide night measurements but that aircraft could be used to determine the bias associated with missing daytime observations.
Dave Skole (Michigan State University) undertook a review of "Pattern to Process: Linking Satellite Observations and Socioeconomic Variables for Regional Land Cover Analysis." Skole said that the key question was the nature of the carbon cycle on an interannual basis. The human dimension is hard to incorporate in models with the land-use/land-cover change information that is needed in order to establish a true understanding of interannual effects. His work begins with Amazonia where he has used data at ten-year intervals. Regrowth rates there are not well quantified. There is an "asynchrony" between deforestation and regrowth. 1995 was a peak year for deforestation in Amazonia.
Research issues he posed were: distinguishing between interannual and decadal rates of change and how to use socioeconomic data and satellite data to better understand the dynamics of deforestation. We must understand how farmers make decisions. "Big" farmers are not a big factor in deforestation but they are important in establishing the regrowth rate. Skole has used a Markov approach in what he called short -run models. This work involved interviews with individual farmers. Skole presented a hypothesis of his work which is: Deforestation events depend on site accessibility, life cycle stage of a family, and family wealth. He noted the regular grid pattern that prevails for farm plots.
Eric Davidson (Woods Hole Research Center) presented "A Conceptual and Empirical Basis for Estimating Nitrogen Oxide Emissions from Soils at Regional and Global Scales." Davidson explained that he uses nitrogen oxide to mean both NO and N2 O since the same organisms, under different circumstances, produce both. He said that nitrifying bacteria, de-nitrifying bacteria , and biological reactions must all be taken into account. The supply of nitrogen is importantit comes from both fertilizers and from nitric acid-acid rain deposition.
Soil moisture affects the N2O/NO ratio, with dry soil causing an increase in NO and wet soil favoring N2O. Davidson has studied NO emissions from soils. For the southeast U.S. he has found that heavy use of fertilizers has led to high NO emissions. Soils are a moderate source of NOx in general, but may dominate in some areas leading to local increases in ozone. For four tropical locations he has found a positive correlation between the amount of NOx flux and the amount of water-filled space in the soil. He found the same sort of correlation with the Tragnet database, which includes sites all over the world. Litter fall also leads to increases in NOx flux.
In his final chart Davidson showed predicted NOx based on litter fall for six sites around the world. Two sites with high acid rain were way off the curve. The presence of leguminous trees was another reason for falling off the curve.
Afternoon Plenary Session, Tuesday (Results from Recently Launched Spacecraft and Analysis Projects). This session was chaired by Michael Freilich (Oregon State University).
Dixon Butler (The GLOBE Program) introduced GLOBE as a project involving children from all over the world. There now over 6100 participating schools and 72 partner countries. GLOBE requires that there be at least one teacher trained by GLOBE in each participating school, and there are more than 75 U.S. GLOBE "training franchisees."
Butler listed some of the types of measurements being taken by the GLOBE students. Student data have been validated against "agency" sources, and some significant differences have been found in some instances. Elissa Levine of Goddard has reported that the soil moisture measurements are very good and helpful in her work.
Butler introduced Russell Congalton (University of New Hampshire) as the GLOBE PI for land cover and related biological measurements. Congalton said that the objective of the program is to obtain data of genuine value. In one instance, wetlands validation was needed. The work entails using the Modified UNESCO Classification system (MUC). MUC has over 150 land-cover classes. Each school involved was given Landsat TM imagery for its area. Then the students classified areas on the map, recorded biometry, made land-cover determinations, and made accuracy assessments. The next step will be to establish protocols to detect future changes. This activity was found to very acceptable when it was applied to state parklands. High schools have to buy a $600 instrument kit to particiate in the program.
Congalton said that this program is an incredible resource for environmental science and education.
Mick Follows (Massachusetts Institute of Technology) presented "Models of Seasonal and Interannual Variability in North Atlantic Biogeochemical Cycles." Follows uses SeaWiFS colors to study interannual variability. He begins with process-oriented models. He has found that some biogeochemical variations correlate with meteorological indicesbut the observations are local and not basin wide. He is looking for underlying mechanisms that might explain the variability and listed three causes: 1) meteorological forcing leading to modulation of the nutrient supply, 2) biophysical interactions, and 3) internal variability of the ecosystem, e.g., predator/prey relations.
Follows has been researching the North Atlantic Oscillation (NAO). The NAO has been strong in the period from 1987 to 1997 leading to tight gradients. His model shows that the convective index for the Labrador Sea has been high during this same NAO period leading to transport of nitrogen through the north Atlantic basin.
Follows wants to model chlorophyll a and compare his results with measurements from SeaWiFS. He showed his model results for chlorophyll concentration for the spring periods of 1988 (low concentration), 1992 (high concentration), and 1994 (high concentration) in the period that the NAO went from low to high.
Summarizing, Follows said that the simplest nutrient cycle shows interannual variability in export production. The variability is dominated by the convective contribution.
Mark Abbott pointed out that as the NAO changes, the wind stress curl will also change. He asked if Follows will use QuikSCAT wind field data. Follows replied that winds will be important, but his results are not strongly sensitive to them. He is doing large-time-step averaging.
John Walsh (University of South Florida) gave some insights into "Regional Models of Carbon/Nitrogen Cycling: Minimum Levels of Ecological Complexity." He started his remarks with a comment that grid models tend to miss near-coastal waters, and these are the places where most of the production takes place.
Walsh gave his findings for three regions: 1) Southern Caribbean Sea: Cariaco Basin and Barbados, 2) Antarctic waters: circumpolar current and Bransfield Strait spring blooms, and 3) Gulf of Mexico: Loop Current and the West Florida Shelf annual cycles.
Southern Caribbean Sea: Cariaco Basin: Walsh said that on a yearly basis the Basin may be a weak source of CO2 to the atmosphere. He said that his springtime case matches observations quite well, but the summertime case was not as good. Some problems are due to missing trichodesmium, missing iron, and missing phosphorous in the model. He pointed out that the source of iron is the Saharan desert plume.
Concerning production in the Antarctic he said that three phytoplankton variables are needed, and, in his model, they were subjected to four types of herbivores to explore the relative importance of grazing pressure and light limitation on annual carbon cycling within both iron-replete waters around the Antarctic Peninsula and putative iron-poor offshore regimes.
Gulf of Mexico: Loop Current and the West Florida Shelf Annual Cycles: A model of the food web affecting red tides, inherent optical properties, and biogeochemical cycles on the West Florida shelf is now being constructed for five types of phytoplankton. All the parameters in his model are measured except for the ciliates. Silicon must be added as an explicit state variable to effect competitive disadvantages for the fast-growing diatoms.
Chuck McClain (Goddard Space Flight Center) discussed "Ocean Applications of the SeaWiFS Sensor." He began with some highlights of the first SeaWiFS year ending with September 18, 1998:
McClain showed some SeaWiFS seasonal composites and pointed out that this was the first time that such data were available. Polar projections show the large contrasts between opposite polar regions. This is the effect of having more iron in the North. The Northern Hemisphere also has more seasonal variations.
McClain also showed some high pigment concentrations that appeared in the equatorial region by August 1998, as we entered the La Niña phase. In the Indian Ocean SeaWiFS observed seasonal variations and phytoplankton blooms in July and August as the summer monsoon regime began to take over. Large coccolithoporid blooms were observed in the Bering Sea during the summer and fall of 1997 and in April 1998. SeaWiFS provided some of the observations of this phenomenon in the Bering Sea. This bloom will impact fisheries such as salmon because the salmon are unable to swim through the bloom to spawn in the Alaskan rivers. SeaWiFS also provided observations of "red tide" in the Bahai Sea off the China Coast in September 1998.
McClain said that SeaWiFS was the first ocean color project with an organized calibration/validation (cal/val) program. They do their cal/val in coordination with the MODIS science team. An on-orbit vicarious calibration methodology for the NIR bands of SeaWiFS is needed even though the sensor was well calibrated before launch. Thus, the prelaunch calibration of these two bands is being used even though the visible bands are being adjusted based on simultaneous in situ SeaWiFS comparisons with MOBY. The solar and lunar calibrations only provide calibration stability relative to the initial on-orbit solar and lunar calibration data. The SeaWiFS project is providing real-time data to several field experiments at any given time.
McClain ended his talk by showing global NDVI results, which they produce routinely, observations of fires in 1998 for Mexico and Florida, and cloud optical thickness measurements.
Following the afternoon break the first speaker was Eric Vermote on the subject of "Prototype Land Surface Reflectance, Vegetation Index, and Aerosol Products from SeaWiFS Data." Vermote addressed four products: 1) Level 3 monthly surface reflectance (minimum blue composite), 2) seasonal variations in RGB, 3) aerosols over land/ocean based on AVHRR/SeaWiFS fusion, and 4) MODIS prototype vegetation indices.
The SeaWiFS Level 3 monthly surface reflectance uses minimum blue that has the advantage of minimizing atmosphere and cloud perturbations. Using the seasonal variation in RGB it was possible to compare April 1998 to September 1998 and note the disappearance of snow cover and the development of vegetation for North America. Over Africa the development of vegetation in the sahelian zone was observed in the northern part, and the seasonal variation was observed in the southern part.
Data from AVHRR were used to detect targets of low reflectance in conjunction with SeaWiFS, thus enabling the derivation of aerosols over land in the SeaWiFS blue and red channels. The Enhanced Vegetation Index (EVI) is being examined as a prototype for MODIS products. The new EVI algorithm tends to minimize aerosol contamination. EVI does better than NDVI for broadleaf vegetation in Brazil.
Vermote concluded that his work with SeaWiFS has generally been successful but that cloud screening needs some improvement.
David Long (Brigham Young University) reviewed "Tropical and Cryosphere Applications of High Resolution Scatterometer Data." He was addressing Ku-band observations at 14 Ghz. While SAR systems can achieve very high resolution, they have limited temporal and spatial coverage. Scatterometers have low (50 km) resolution, but rapid, global coverage. Long used the Scatterometer Image Reconstructive Filtering (SIRF) algorithm to enhance the scatterometer resolution. Referring to tropical rainforest mapping, he said that classification is performed according to values of the A (sigma-0 at 40 degrees incidence angle) and B (sigma-0 slope vs. incidence angle) coefficients. Accuracy of 94 percent has been achieved using the A and B coefficients along with polarization data. Comparisons of Seasat-A Satellite Scatterometer (SASS) data from 1978 with NSCAT data from 1996 have permitted identifying new highways and new settlements in Brazil from the change images. Scatterometer images are particularly useful in the polar regions. Scatterometer images of the polar regions are useful in monitoring seasonal melt cycles and tracking ice motion.
Long summarized the advantages of scatterometers as follows: They are able to provide wide-area coverage; they offer frequent observations; they have high radiometric accuracy; they provide multiple incidence angles; and they also have multiple azimuthal angles. As such, they complement both high-resolution SAR sensors and low-resolution passive radiometers.
Mark Drinkwater (Jet Propulsion Laboratory) described "Land Ice and Sea Ice Studies with Radar Data." The European scatterometers have now been operating from 1991 to the present day, giving us valuable C-band microwave radar time series data. Scatterometers have high sensitivity to the presence of liquid water along with ice and are less sensitive to meteorological conditions than passive microwave instruments. Drinkwater showed images of surface melting in Greenland for the period 1992 to 1996.
Knowledge from Greenland observations has been applied to the study of Antarctic ice.
Drinkwater mentioned that sea-surface scatterometry goes back to the Seasat-A Satellite Scatterometer (SASS) flown in 1978. ERS Scatterometer data for Palmer and Marie Bird Land showed strong trends in snow accumulation over land, whereas SSM/I appears less sensitive to the changes. Differences between NSCAT and SASS images indicate significant regional decadal variability, which reinforces the ERS Scatterometer observations.
Drinkwater made some concluding observations about scatterometry and sea ice: Sea-ice snow-surface melt causes large reductions in the backscatter signal. Surface salt-water flooding, which is also particularly prevalent in Antarctica in summer, can be distinguished from surface melt on snow-covered sea ice. Observations indicate considerable interannual variability in surface melting. Scatterometry is able to distinguish the onset of Antarctic sea-ice melt, and how widespread the effects of summer flooding are in the Antarctic.
Melting indicates situations in which atmospherically delivered heat contributes to removal of the sea-ice cover, whereas flooding only occurs under special conditions in which the ice is buffered from the atmosphere by snow cover. Flooding is significant, as it results in upward growth of meteoric ice, and a summer survival mechanism for Antarctic ice. Meteoric ice (possibly better referred to as snow ice) comprises a significant component of the mass of Southern Ocean sea ice.
Antony Liu (Goddard Space Flight Center) reviewed "Sea Ice Motion from Wavelet Analysis of Satellite Data." Liu began with a presentation of the mathematics of the approach he used. The technique involves a 2D gaussian waveletthe Mexican hat wavelet.
He described an analysis of SSM/I, SMMR, and NSCAT images merged with buoy data. He then illustrated the international arctic buoy program and showed a good match between buoy drift speeds and NSCAT and SSM/I wind speeds and angles. He noted that ice-motion streamlines tended to match atmospheric pressure fields. He was able to discern a reversal of the ice motion that occurred in just four days.
Liu concluded his presentation with the statement that he has shown that it is possible to merge sea-ice motion data from NSCAT, SSM/I, and buoys, but that there is still a problem to do this tracking in the summer melt season.
Morning Plenary Session (Climate Change and Public Policy). John Melack (University of California Santa Barbara) chaired this final session.
Jim Hansen (Goddard Institute for Space Studies) led off with a review of "Climate Forcings and Climate Change: Relevance of Global Observations to Kyoto Protocol." The Kyoto protocol attempts to minimize greenhouse gas emissions, but there are two problems: 1) the opponents assert the likelihood of significant economic damage from constraints on greenhouse emissions; and 2) the proponents think that scientific understanding is so good that no more research is needed!
Hansen said that we need better understanding in order to adopt the most cost-effective policies. He referred to a Science paper (August 21, 1998) on energy resources. He noted that it now looks as if global production of oil and gas will peak during the next few decades, raising the issue of what sources of energy should be developed to follow oil and gas.
He went on to review the estimated climate forcings for the roughly 100 years between 1850 and the present (see the figure). He divided up the types of forcing into three classes: greenhouse gases, other anthropogenic forcings, and natural forcings. Well-mixed greenhouse gases were the largest source of positive greenhouse forcings, and aerosols and related "forced cloud changes" led to the greatest negative forcing. The imputed tropospheric aerosol cooling is quite uncertain. Three types of aerosols were included in the calculation: sulfates, organics, and soil dust. The much larger uncertainty in forced cloud changes is due to the indirect effect of aerosols on clouds. A microphysics model was used to determine this. The net forcing due to all the factors considered in the figure is + 1 W m-2.
Hansen said that the rate of increase of greenhouse gas forcing has been decreasing over the last 20 years. Apparently, the CO2 sink has been increasing for the last 20 years along with the increases in emissions.
The present growth rate of the greenhouse climate forcing is only half of that in the most common IPCC scenarios, mainly because of reduced growth of methane, chlorofluorocarbons, and carbon dioxide. A key issue is to understand better the sources and sinks for methane and carbon dioxide so that we can estimate future climate forcings and the effectiveness of different policy options.
Frank Wentz (Remote Sensing Systems) reported on an analysis of "Decadal Time Series for Tropospheric Climate State Variables." He reviewed the recent findings of tropospheric cooling that have been reported, based on the Microwave Sounding Unit (MSU), and said that when an orbit-decay correction was made, the reported cooling became warming. The debate has to do with the global climate record for the period 1979 to 1995. The accepted surface thermometer value has shown warming of +0.15 K/decade, whereas the MSU results were -0.05 K/ decade. In the same period the accepted tropical SST trend was +0.10 K/decade whereas MSU results were -0.11 K/decade. After applying the orbit-decay correction, the MSU trend becomes +0.07 K/decade, in better agreement with the surface data.
Another comparison for the period 1987 to 1997 showed warming for both SST as measured by AVHRR and water vapor as measured by SSM/I. Wentz explained some of the differences between MSU determinations and ground-based as due to the channel 2 weighting functions of the MSU sensor. The MSU weighting function peaks in the mid-troposphere. It is important to realize that the shape of the function is dependent on scan angle. The cooling that is computed is an artifact of the procedure that is used to get the near-surface temperature by taking into account Tnadir and Tlimb. The orbital decay affects the value of Tlimb. When appropriate corrections are made, MSU gives reasonable temperature profiles.
Wentz compared MSU, AVHRR, and SSM/I results and concluded that we must look at data from many different approaches based on the work of many different groups. It is important to have better-calibrated radiometers with better-known responses to instrument temperatures. We need plenty of overlap between successive space platforms, and they should have fixed ascending nodes. There should be a focus on measuring climate with low-cost systems that will be good for establishing the long-term record.
Wentz feels that NPOESS requirements are getting too complex for long-term sustainability of the mission.
Mark Abbott (Oregon State University) discussed "The Oceans and Climate Change: Impacts and Feedbacks." He said that the oceans play a critical role in Earth's climate. There are links between the large time- and space-scale patterns of ocean circulation and very small-scale responses in the ocean ecosystem. Broecker has assumed that nitrogen is the key variable in limiting net primary productivity in the ocean. John Martin sees not nitrogen but iron abundance as the limiting factor for ocean productivity. He is quoted as saying "give me enough iron and I can create an ice age."
Abbott described the relation between the microbiological loop and iron. Low iron favors small phytoplankton, and they are grazed by very small grazers, leading to rapid and efficient recycling of iron and nitrogen. This is the baseline ecosystem for much of the world ocean. There is a tight balance between grazers and phytoplankton. The grazers keep nutrients in the upper layer.
Abbott then discussed "non-equilibrium strategies" and asked what would happen if we added iron in the Southern Ocean. Landry has said that adding iron to the ocean would change the types of phytoplankton that were present. Pulses of iron favor diatom blooms (large species). Small forms don't grow. Non-equilibrium communities dominate ocean fluxes, thus supporting many fisheries.
Abbott showed the weekly position of the Antarctic polar front. He said that drifters in a polar front meander indicate the patterns of iron increase and decrease. He showed observations of intense spring blooms of chlorophyll obtained from moorings located around Antarctica. Strong blooms are associated with topography and iron present in the meanders. The appearance of fluorescence is an indication that the phytoplankton are not using nutrients efficiently.
Reviewing the record for the Central North Pacific, Abbott commented that this is the locale of one of the largest, most ancient ecosystems. ENSO events led to more-stable stratification and more phosphorous-limited communities. Changes in variability are being observed in the northern Pacific. The paleologic record could provide evidence for a strong drawdown of atmospheric carbon dioxide, but the record is incomplete and contradictory. More ENSO events could mean more shifts in oceanic regime, but the number of possible oceanic states is limited.
In summary, Abbott said that a long time series of consistent observations is criticalboth satellite and in situ observations are needed. Better models of mesoscale phenomena are needed. Present ecosystem models are too limited, and multiple "species" and multiple nutrients must be included in the models. Physiological processes such as nitrogen and iron fixation need to be parameterized. Response to environmental transients must be included in the models.
Ken Hawker (MITRE Corporation), as final speaker, addressed "The Roles of Uncertainty and Error in the Functioning of the Kyoto Protocol." Hawker called attention to the polarization that stands between the public's policy views and the Kyoto implications. He raised questions such as: What about inadvertent errors in achieving compliance? Can we stick to mean values?
He said that the treaty is a "greenhouse gas" treaty intended to control the rate of use of certain stocks. (The treaty sets a target of cutting emissions worldwide by an average of 5.2 percent from 1990 levels, between 2008 and 2012. The U.S. target, as a developed nation, is 7 percent, but the U.S. Congress is not likely put this into law.) There are direct economic stakes in emissions trades. The treaty fails to define what is meant by "emission reduction units." "Transparency" is the key feature of the treatyeach nation has to report what it's doing.
There are complexities in the treaty that must be dealt with. The use of a "market basket" of gases requires the use of Global Warming Potentials (GWPs). The basic equation used for oil, gas, coal, deforestation, rice farming (methane production) is:
Resource Use ¥ Emissions Factor ¥ GWP = Equivalent CO2 emissions.
Net emissions must account for tropical deforestation. The treaty is not a self-contained instrument. The protocol incorporates the 1996 IPCC methodologies, e.g., the IPCC prescribes the method to be used to calculate methane emissions from rice paddies.
Nations are responsible for treaty-defined quantities. Every nation could be in compliance, and still the overall objectives of the treaty would not be achieved.
Hawker commented on errors and uncertainties. We don't know the absolute errors in national fossil fuel consumption estimates, not even for the Organization for Economic Cooperation and Development (OECD) countries. There are great differences between the UN and the International Energy Agency compilations of emissions. The greatest uncertainties lie in the estimates of land use change forestry (LUCF).
The estimated uncertainty for fossil fuel emissions is still too large: two different current estimates differ by 12 percent for the 48 largest emitting countries, 26 percent for the medium emitters, and 60 percent for the smallest emitters. Combining uncertainties leads to a total uncertainty in CO2 of 10 percent, this despite the fact that the global goal is a reduction of five percent! There is also uncertainty resulting from the permissible trading of CO2 emissions between countries.
The contrast between the Kyoto treaty and the Montreal protocol is that the latter forbids things. This is easier to enforce than setting limits that must be adhered to.
Estimated radiative forcings between 1850 and the present.
Hansen, J., M. Sato, A. Lacis, R. Ruedy, I. Tegan and E. Matthews.
Climate forcing in the industrial era. Proc. Nat. Acad.
Sci., 95, 12753-12758, 1998