The 16th Clouds and the Earth's Radiant Energy System (CERES) Science Team meeting was held at Oregon State University (OSU) in Corvallis, OR, on September 16-18, 1997. The focus of the meeting was on the status of the Earth Observing System (EOS) program and the EOS Data and Information System (EOSDIS), and the CERES instrument, data management system, algorithms, and validation plans. The first CERES launch is scheduled for November 1997 in Japan on the Tropical Rainfall Measuring Mission (TRMM) spacecraft. The Science Team guides the definition of the CERES instrument and science studies to provide a climate data set suitable for examining the role of clouds in the radiative heat balance of the climate system.
Bruce Wielicki, CERES Co-Principal Investigator, opened the meeting with an EOS program status report. The EOS AM-1 and EOS PM-1 missions are on schedule. Wielicki, Bruce Barkstrom, and the Science Team officially welcomed three new CERES Co-Investigators: Leo J. Donner of the NOAA Geophysical Fluid Dynamics Laboratory (GFDL), who has cirrus/convective global and regional cloud-modeling expertise; David P. Kratz of the NASA Langley Research Center (LaRC), an expert in longwave radiative transfer; and David F. Young of LaRC, who will lead the Time Interpolation and Spatial Averaging Working Group. The addition of Leo Donner to the team is in response to a recommendation at the last peer review to strengthen the modeling component.
CERES Instrument Status: TRMM and EOS AM-1
Jack Cooper (LaRC) presented the instrument status report. The CERES Proto-Flight Model (PFM) instrument on the TRMM spacecraft successfully completed spacecraft-level testing. The TRMM pre-shipment review was successfully completed in early August, and the spacecraft was shipped to Japan. The primary issue of concern was the azimuth stall anomaly, but the data verified proper bearing operation, and a recovery sequence was successfully demonstrated. The flight model (FM1 and FM2) instruments were successfully integrated on the EOS AM-1 spacecraft, and comprehensive functional tests were completed. Opto-isolator units that were reworked to correct improper tinning of gold-coated parts were subjected to accelerated life-cycle tests to verify integrity of solder joints. One failure occurred after the equivalent of 13 years; no other failures were noted after the equivalent of 50 years. The FM3 sensors for EOS PM-1 were delivered to TRW in late April 1997. Spectral characterization tests completed in early September indicated extremely small air/vacuum shift and drift characteristics. The FM4 sensors were scheduled for delivery in late September. Fabrication of FM3 and FM4 electronic and mechanical assemblies is proceeding on schedule. Based on the units produced thus far, the CERES instrument is designated as a 50-W, 50-kg-class instrument.
Data Systems: EOSDIS, LaRC DAAC, and CERES DMS
Bruce Barkstrom (LaRC) reviewed EOSDIS progress and problems. A demonstration of a prototype system was successful, but reviewers noted concerns over operational aspects of expanding the prototype to full operations. Richard McGinnis (LaRC) presented the status of the Langley TRMM Information System (LaTIS) for CERES data processing at the Langley Distributed Active Archive Center (DAAC). LaTIS development was undertaken because the EOSDIS will not be ready by the time of the TRMM launch. In addition, the Science Working Group for the AM Platform (SWAMP) insisted on an emergency backup system to cover AM-1 processing in case of additional EOSDIS delays. The CERES Science Team remains concerned about the delivery and viability of the EOSDIS Core System. The Science Team strongly preferred to evolve the LaTIS data system from TRMM to EOS-AM/PM, and suggested that such a proposal be submitted to the EOSDIS Project if this could be shown to be cost neutral or cost saving.
Jim Kibler (LaRC) presented the CERES Data Management System (DMS) status. The CERES Release 2 DMS deliveries to the DAAC are proceeding on schedule. End-to-end TRMM mission simulations resolved numerous technical and interface issues and demonstrated the successful operation of the data systems. Near-term plans include completion of LaTIS testing with a month of simulated CERES data prior to the TRMM launch, and maintaining compatibility with both LaTIS and EOSDIS for CERES processing.
CERES Validation
The Science Team reviewed the CERES validation plans. No major problems or issues were identified. Updated validation plans for all CERES subsystems are nearly complete. Bruce Wielicki led a general discussion of the overall validation plan, its shortcomings, and strengths. Wielicki remarked on the lack of tropical land and mid-latitude ocean corroborative data. Michael King of the NASA Goddard Space Flight Center (GSFC) suggested that data taken in a tropical field experiment planned for August-September 1999 (with Moderate Resolution Imaging Spectroradiometer [MODIS] participation) will be quite helpful. Jim Coakley (OSU) updated the group on the Indian Ocean Experiment (INDOEX) planned for February-March 1999. Tom Charlock (LaRC) presented plans for the Workshop for Atmospheric Validation in EOS AM-1 and SAGE III (WAVES) in October, 1997. WAVES will coordinate new EOS validation investigations with EOS instrument science teams.
Invited Presentations
Dennis Hartmann (University of Washington) briefed the team on "Climate Processes Over the Oceans," an EOS Interdisciplinary Science Investigation. Their goal is to construct an integrated view of atmospheric climate over the oceans. The physical processes considered involve boundary-layer dynamics and resulting fluxes, cloud-scale and mesoscale dynamics, cloud physics, and global-scale circulations. The important phenomena of interest in this investigation include: boundary-layer fluxes of heat, momentum, and moisture; low-level clouds; tropical convective clouds; and midlatitude synoptic systems.
Bob Cess (State University of New York at Stony Brook, SUNY-SB) showed evidence to demonstrate that with the current state of General Circulation Models (GCMs), the seasonal cycle cannot be used to validate a model's ability to predict climate change.
Dave Young (LaRC) gave a status report on reprocessing data from the Earth Radiation Budget Experiment (ERBE). Numerous calibration and modeling improvements are being incorporated and tested in the ERBE algorithm. The final archived ERBE data and the CERES ERBE-like data will use the same algorithm.
Working Group Reports
Instrument Working Group: Robert B. Lee III (LaRC) led the Instrument Working Group (WG) meeting. Lee completed a draft of a paper on pre-launch calibrations for the PFM, FM1, and FM2. D. K. Pandey (SAIC) discussed spectral characterization of the CERES sensors and showed results from TRW's Fourier Transform Spectrometer (FTS) Vacuum Spectral Characterization Facility. Kory Priestley (LaRC) presented current operational scan-dependent offsets for the PFM instrument, and preliminary scan-dependent offsets for FM1 and FM2. The offsets will be verified on-orbit while viewing cold space during planned spacecraft pitch-over maneuvers. Priestley also presented a comparison between measured and numerically predicted point response functions for the PFM instrument. Models and data verify that a time offset of 24 ms should be implemented for the PFM instrument. Lou Smith (Virginia Tech) designed a numerical filter to reduce a slow, single thermal mode transient in the PFM total-channel signal to a negligible level.
Cloud Working Group: Bruce Wielicki led discussions on the status of cloud-related algorithm and validation activities. He cautioned the group to be very careful in analyzing Visible InfraRed Scanner (VIRS) data from TRMM because the cloud algorithm may be the first real test of calibration accuracies. Jim Coakley showed close comparisons between cloud layers identified using the spatial coherence method and Lidar In-Space Technology Experiment (LITE) data. The LITE data had almost no 100% cloud-free scenes, demonstrating that it is capable of detecting optically-thin clouds. Bryan Baum (LaRC) highlighted some special problem areas with the CERES cloud mask (e.g., smoke over highly-reflective land surfaces), but concentrated on results from discrete ordinate calculations that can be used to establish clear/cloud temperature thresholds in the 10.8-and 3.7-µm channels. Patrick Heck (Analytical Services & Materials, Inc., AS&M) presented new strategies for updating the clear-sky radiance history (CRH) database, including a method for updating albedo based on long-term changes rather than day-to-day variations. Presentations on the construction of Angular Distribution Models (ADMs) by Larry DiGirolamo (University of Arizona) and Yongxiang Hu (Hampton University) stimulated lively discussions on the use of reciprocity in the construction of ADMs for cloudy skies, but more research is needed in this area before firm conclusions can be reached. A related study by Steven Dewitte (Royal Meteorological Institute of Belgium) showed that ADMs derived using the Radiance Pairs Method (RPM) model performed better than four other models in removing the angular dependency of ERBE or ScaRaB (Scanner for Radiation Budget) measurements for clear-sky desert areas.
Surface and Atmospheric Radiation Budget (SARB) Working Group: Thomas Charlock (LaRC) led the SARB WG in presenting the status of the Release 2 SARB code and discussing issues regarding the validation of SARB and Surface-only data products. Shi-Keng Yang (NOAA National Centers for Environmental Prediction, NCEP) reviewed the definitions of tropopause height used by various institutions, and the WG selected 200 hPa and 70 hPa as the upper-troposphere and lower-stratosphere reference levels for archiving SARB-derived heating/cooling rates. Fred Rose (AS&M) incorporated a simplified parameterization of the CKD (Clough, Kneizys, and Davies) water-vapor continuum into the Release 2 version of the SARB code. He also demonstrated the benefit of using a high-resolution surface-elevation map in SARB computations at instrument footprint resolution. Shi-Keng Yang presented details of the Stratospheric Monitoring-Group Ozone Blended Analysis (SMOBA) ozone product that has been adopted as the primary source of ozone for CERES processing. David Rutan (AS&M) presented a database that provides surface optical properties, a digital elevation map, and a land-water percentage map. Tim Alberta (AS&M) presented a brief summary of the CAGEX (CERES/ARM [Atmospheric Radiation Measurement]/GEWEX [Global Energy and Water Cycle Experiment]) Version 2.0.0 data set that was recently released to the science community. The data set consists of multiple sets of meteorological parameters, cloud properties, and aerosol optical properties and distributions taken at the ARM Southern Great Plains (SGP) site during the ARM Enhanced Shortwave Experiment (ARESE) in October 1995. He showed that atmospheric absorption computed with the Fu-Liou radiative transfer model was always lower than the observed values regardless of which input data set was used. Bob Wheeler (AS&M) described the CERES Airborne Radiometer Scanner (ARS) effort to characterize bidirectional reflectance distribution functions for various surface types as a part of CERES validation activities. Wheeler also described the Chesapeake Light Ocean site being readied for radiation and aerosol measurements, also for CERES validation.
Time Interpolation and Spatial Averaging (TISA) Working Group: David Young led discussions of software development, validation plans, and ongoing temporal and spatial averaging studies. All TISA subsystems are on-target for delivery to meet TRMM schedules. Lou Smith (Virginia Tech) showed that results from an analytical method for estimating temporal averaging errors for monthly mean LW and albedo compared well with rms errors derived from ERBE data. Dave Doelling (AS&M) presented results of studies to estimate spatial sampling errors for CERES radiation and cloud parameters. Young developed new SW broadband directional models for ERBE and CERES using 60 months of ERBS data.
Investigator Presentation Highlights
Tom Charlock (LaRC) calculated the absorption of SW by a cloudy atmosphere using the Fu-Liou delta 4-stream code. The column atmospheric absorption inferred by differencing GOES-8 and surface radiometers was about 100Wm2 more than the computed absorption for October 30, 1995, apparently demonstrating anomalous absorption. Calculations included a wide array of remote-sensing inputs for cloud properties. Computed absorption approached measured values only when questionably large aerosol particles were placed in the cloud.
Dominique Crommelynck (Royal Meteorological Institute of Belgium) reported on the status of the Solar Constant Experiment (SOLCON) Hitchhiker mission. SOLCON will make periodic observations of the absolute value of the solar constant using a differential absolute radiometer and a digital processor unit. The next SOLCON Hitchhiker flight is planned for October 1998.
Xiquan Dong of AS&M (representing Patrick Minnis) showed good agreement between boundary layer cloud properties derived from surface measurements during SUCCESS (SUbsonic aircraft: Contrail and Cloud Effects Special Study) and satellite-based retrievals from AVHRR (Advanced Very High Resolution Radiometer) and GOES (Geostationary Operational Environmental Satellite). Cloud reflection and optical depth from GOES were lower than the surface-derived values due to instrument calibration, differences in spectra, view angles, and retrieval methods. The comparison was better for lower values of cloud liquid water path (LWP) than for higher cloud LWPs.
Leo Donner (GFDL) addressed the use of high-resolution, 3D cloud system models in the development of GCM cumulus parameterizations. The modeled cloud system undergoes a convective-to-stratiform life cycle. The modeled heat source and moisture sink are in broad agreement with observations, but upper-troposphere moisture exceeds observations and cloud/radiation interactions are not well understood. CERES data will be used to evaluate 3D distributions of clouds and radiative fluxes, and to examine synoptic evolution of cloud systems.
Qingyuan Han of University of Alabama - Huntsville (representing Ron Welch) showed that POLDER (Polarization and Directionality of Earth Reflectances) data can be used to identify sun glint areas and to discriminate ice-water cloud, thin cirrus over water cloud, and aerosol above cloud. Analysis of cloud morphology effects showed good agreement between POLDER results and plane-parallel predictions for large cloud fields, but revealed substantial differences due to 3-D effects for mesoscale cumulus clouds.
Martial Haeffelin of Virginia Tech (representing David Young) presented methods for reducing temporal sampling errors in CERES monthly mean atmospheric and surface fluxes using regular or solar-zenith-angle-weighted linear interpolation schemes. Temporal sampling errors of CERES monthly mean radiative fluxes were reduced to less than 0.5% and 1% in the SW and LW domains, respectively (on the order of 2Wm2 in both spectral domains).
Anand Inamdar of Scripps (representing V. Ramanathan) reported on a continuing analysis of water vapor radiative feedback. They analyzed a new data set encompassing a global domain including both the continents and oceans as well as both the ascending and descending branches of the Walker and Hadley cells. They derived the greenhouse effect for oceans and land separately and combined, and mapped observed changes in the atmospheric greenhouse effect and total precipitable water. The coupling between the surface temperature, the atmospheric greenhouse effect, and the vertical distribution of atmospheric water vapor and temperature was examined from the annual cycle.
Bing Lin of Hampton University (representing Bruce Wielicki and Patrick Minnis) analyzed daytime and nighttime scenes containing overlapped clouds over the tropical western Pacific warm pool regions using the combined Special Sensor Microwave/ Imager (SSM/I) and Geostationary Meteorological Satellite (GMS) data. Infrared (IR) and visible (VIS) measurements from GMS are used to detect upper-layer cirrus/anvil decks, while cloud liquid-water temperature and path for lower layers are retrieved from SSM/I microwave (MW) data. Results show that MW, IR, and VIS satellite images provide the signals for overlapping clouds very well.
Norman Loeb of OSU (representing Jim Coakley) reported on the influence of horizontal and vertical cloud inhomogeneities in satellite retrievals of cloud optical depth and liquid water path. At pixel scales ranging from 50 m to 2 km, retrieved cloud properties from predominantly overcast marine stratus show very little sensitivity to pixel resolution. LWP estimates based on radiative transfer schemes which ignore vertical variations in cloud properties overestimate the true LWP. The LWP bias tends to increase with cloud optical depth and with solar and viewing zenith angles.
David Randall (Colorado State University, CSU) reported on EauliqNG, the next generation model for predicting cloud water, cloud ice, rain, and snow. EauliqNG prognosticates cloud amount and the temperature and water vapor mixing ratio of the clear and cloudy portions of each grid cell. EauliqNG fully couples the stratiform cloud and convection parameterizations, allows a mesoscale circulation with different vertical velocities in the clear and cloudy regions, and determines the strength of the mesoscale circulation by requiring that the virtual temperature be the same in the clear and cloudy regions. EauliqNG is currently being tested in the CSU GCM.
David Rutan of AS&M (representing Lou Smith, Virginia Tech) applied principal components analysis to 5 years of Earth Radiation Budget Experiment data to study the diurnal cycle of albedo. Determined that there is very little intraseasonal variability in the first Empirical Orthogonal Function (EOF); over 90% of the seasonal variance is explained by the first three EOFs; the first EOF for land/ocean explains the magnitude of the diurnal variation; and the second EOF accounts for the morning-to-afternoon albedo variability due to clouds.
Si-Chee Tsay of GSFC (representing Michael King) reported on cloud mask and cloud property retrieval in the Arctic. The MODIS cloud mask does a good job except for very thin/low-level stratus clouds over sea
ice in the Arctic and low-level cloud shadow cast from upper-level clouds. Cloud properties retrieval in the Arctic is very sensitive to surface spectral reflectance. Using the 1.24-µm window channel as a non-absorbing channel can reduce the probability of getting multivalued solutions and suppress the effect of multiple reflection between cloud and surface.
Yaping Zhou of SUNY-SB (representing Bob Cess) described an algorithm for calculating the relation between surface downward LW and outgoing LW radiation. Limited observational data and model sensitivity studies showed the correlation between the LW flux ratios and column water vapor to be remarkably invariant to both the presence of clouds and, evidently, also to the vertical distributions of water vapor and temperature.
CERES Educational Outreach
Lin Chambers (LaRC) briefed the team on the CERES S'COOL (Students' Cloud Observations On-Line) Project. Students observe and report clouds at the time of CERES instrument overpass, compare their observations to CERES cloud retrievals, and provide feedback on up- versus down-looking results. The program has already been tested locally and in Europe. CERES also supported preparation of a series of NASA fact sheets for public distribution.
Science Team Logistics
The next CERES Science Team meeting is scheduled for April 21-23, 1998 at the NASA Langley Research Center in Hampton, VA. The major topic will be a review of progress on the validation for each subsystem, with a focus on instrument and ERBE-like data products. Other items will include an assessment of CERES and VIRS data from TRMM, and CERES launch readiness for EOS AM-1 in late June or early July of 1998. Another Science Team meeting will be held in California to coincide with the EOS AM-1 launch.