The Earth Observer, November/December, 1995

Clouds and the Earth's Radiant Energy System (CERES)

Bruce R. Bardstrom (brb@ceres.larc.nasa.gov), Principal Investigator, and
Gary G. Gibson, NASA langely Research Center

The 12th Clouds and the Earth's Radiant Energy System (CERES) Science Team meeting was held September 20-22, 1995, at the NASA Langley Research Center (LaRC) in Hampton, VA. The focus of the meeting was CERES instrument status, algorithm development, and validation. The CERES instrument is designed to provide a climate data set suitable for examining the role of clouds in the radiative heat balance of the climate system. The CERES Science Team blends expertise in broadband radiometry, cloud and radiation remote sensing, and climate modeling. The Science Team guides the definition of the CERES instrument and science studies.

Robert B. Lee III and Leonard Kopia (LaRC) presented the instrument status report. The CERES instrument weight and power are forecast to be very close to allocations for both the Tropical Rainfall Measuring Mission (TRMM) and Earth Observing System (EOS) satellites. Instrument assembly for the Proto-Flight Model (PFM) was completed in May, and instrument performance during initial vacuum testing was excellent. Modifications to reduce Electro-Magnetic Compatibility (EMC) exceedances to acceptable levels were completed. Vibration testing and thermal balance/thermal vacuum testing were also successfully completed.

Final calibration and comprehensive functional tests were on target for a mid-October delivery to the Goddard Space Flight Center. Integration on the TRMM spacecraft is scheduled to begin in January. EOS AM flight model (FM-1 and FM-2) detectors and mirror coatings passed inspection. The FM-1 main contamination cover (MCC) experienced several anomalies as a result of vibration testing, but operated successfully. The MCC caging device is being reworked and will be qualified on FM2.

The CERES absolute radiometric calibration facility provides calibration of each instrument over its full spectral range, field of view, and dynamic range. CERES PFM calibration results showed that: (i) sensor offsets are stable and vary less than 3 counts with elevation angle, (ii) mirror attenuator mosaic (MAM) calibration mechanisms should yield solar calibrations at the 0.5% precision level, and (iii) the PFM sensors have a second thermal time constant which, if uncorrected, may increase the instantaneous measurement errors above 0.5% for longwave (LW) and 1.0% for shortwave (SW). A correction algorithm is being developed.

Jim Kibler gave a detailed status report on each subsystem including a summary of data products and code development. Code development is generally on schedule with some modules now undergoing testing. Challenging areas include the validation of submitted science code for cloud analysis, the size of the meteorological, ozone, and aerosol product (25 MB per hour), and memory requirements for executing the Surface and Atmospheric Radiation Budget (SARB) prototype code. Kibler also reported on the development of graphics tools to assist software development and support production processing and validation. Near-term plans include the implementation and testing of Release 1 science algorithms, delivery of the Release 1 code and test data sets to the LaRC Distributed Active Archive Center (DAAC) for integration and testing, and definition of requirements for Release 2 (TRMM flight processing system).

Bob Lutz, Quality Assurance (QA) Scientist in the Earth Science Data and Information System (ESDIS) Science Office at GSFC, presented a science QA procedure for EOS products. He provided a framework for understanding the operational QA methodology utilized by the Instrument Teams, identifying the QA requirements of the users of the data products, and ensuring that the EOS Data and Information System (EOSDIS) satisfies the requirements of both of these communities.

The Instrument Working Group was led by Robert B. Lee III (LaRC). The Release 1 algorithm is set for test runs in November and will be delivered to the DAAC in January 1996. Procedures were established for flight count conversion coefficient instrument gain and offset determinations. Ground-derived coefficients are used as preliminary flight coefficients. The Release 2 (flight) version is currently being defined as a refined Version 1 including special data handling features for in-flight calibrations and validation. A preliminary validation plan draft was prepared, which includes flight calibration analyses (internal calibration module and MAM), multi-channel comparisons (Inversion Working Group), multi-satellite intercomparisons (Time Interpolation and Spatial Averaging [TISA] Working Group), single spacecraft cross-track and rotating azimuth plane instrument comparisons, and geolocation/coastline detection studies.

John Chapman (LaRC) reported on a CERES instrument simulator that is being developed to allow flight operational familiarity with the instrument prior to launch. The simulator development is a joint effort of the Data Management Office and the Langley Summer Scholars Program, which is coordinated through the LaRC University Affairs Office. The PC-based CERES simulator consists of circuit cards functionally identical to the flight items, but using low-cost commercial microcircuits and components. One simulator application is for use as a testbed for functionality checking of atypical memory uploads and for anomaly investigations.

The Working Group was led by Bruce Wielicki (CERES Interdisciplanary Science PI). Bryan Baum (LaRC) reported that the Version 1 prototype code is up and running. A global data processing strategy has been developed and implemented that can be used with other imagers. All submitted Co-I algorithms have been integrated and the code exercised using 3 hours of Advanced Very High Resolution Radiometer (AVHRR) Global Area Coverage (GAC) data. The stage is now set for Science Team involvement to guide the application of algorithms. CERES cloud retrieval follows these steps: (i) apply cloud mask, (ii) update clear-sky map, (iii) classify clouds and detect cloud layers, (iv) locate cloud-top heights for each layer, (v) derive cloud microphysical and optical properties, and (vi) convolve imager results with the CERES field of view. Short-term goals are to make Satellite Image Visualization System (SIVIS) software and CERES cloud output available to the team, develop production code, and process the CERES cloud algorithm at the LaRC DAAC under Pathfinder. Long-term goals are to implement nighttime algorithms, implement strategies for smoke/fire-covered areas, sunglint, mountains, etc., and improve quality control (QC)/exception handling capabilities.

Michael King, EOS Senior Project Scientist, gave an overview of the lessons learned during the Arctic Radiation Measurements in Column Atmosphere-surface System (ARMCAS) experiment. ARMCAS involved satellite remote sensing, two high-altitude and boundary-layer aircraft measurements, and surface remote sensing with ground truth observations to better understand radiative processes in the Arctic. King also summarized the current Moderate-resolution Imaging Spectroradiometer (MODIS) status, which is of particular interest since MODIS data will be used to derive cloud properties for CERES.

Several science studies were reported which aim at enhancing cloud estimation/analysis capabilities and could affect validation planning. Bing Lin (LaRC) presented a method for estimating multi-level cloud liquid water path (LWP) and height (ą1 km) from satellite passive microwave and optical measurements in oceanic environments. Lin Chambers (LaRC) compared 2D and plane-parallel methods for optical depth retrieval for ocean boundary layer clouds. The rms optical depth retrieval error was 10.5%, with a maximum error of 40%. Applying a cloud-aspect-ratio-based correction reduces the rms and maximum errors to 4.5% and 18%, respectively. The retrieval of cloud aspect ratio is under investigation. James Coakley (Oregon State) showed the effect of spatial subsampling of MODIS data on cloud retrievals.

V. Ramanathan (Scripps) submitted a new clear-sky LW top-of-atmosphere (TOA)-to-surface flux parameterization method. Larry Stowe (NESDIS) presented an improved algorithm for aerosol remote sensing, which he is testing with AVHRR data. Ronald Welch (South Dakota) completed his new cloud masking algorithm and Qingyuan Han (South Dakota) presented some results of recent studies relating cloud microphysics and albedo. Coakley showed that Spatial Coherence inferences of layered cloud structures correlated closely with lidar returns taken during the Lidar In-space Technology Experiment (LITE) mission. He also compared reflectivities predicted using plane-parallel radiative transfer theory to observed data for single layer clouds and found significant discrepancies. This is particularly important since plane-parallel radiative transfer theory is the mainstay of cloud retrieval algorithms and radiation schemes in climate models. This is one of the primary reasons CERES will develop new empirical models of anisotropy to convert radiance measurements to estimates of radiative flux. Richard Green (LaRC) developed new SW angular distribution models (ADMs) from Nimbus-7 ERB data using the Radiance Pairs Method (RPM). Green also presented the CERES-proposed EOS grid and results of a regridding error analysis.

Patrick Minnis (LaRC) presented a comparison of cloud property retrievals using the daytime (visible, 10.7, and 3.9 ľm) and nighttime (10.7, 12, and 3.9 ľm) algorithms. Minnis also showed initial results of an error analysis of the daytime cloud microphysical property retrieval algorithm which has been applied globally to determine cloud phase, effective particle size, and optical depth. Dual solutions and limited dynamic range contribute to difficulties in retrieving water droplet radius in backscatter viewing conditions.

The Working Group identified several key areas of research and development that are being examined for Release 2 cloud and TOA flux algorithms: Particle size retrieval algorithms (day and night), vertical dependence of cloud particle size, TRMM Visible Infrared Scanner (VIRS) ocean aerosol algorithm, nighttime polar cloud mask, 2-D and 3-D effects on derived cloud properties, addition of sounder cloud heights, multi-layer cloud mask and properties algorithms, microwave cloud property retrievals, beam filling for 2km VIRS footprints, ADM simulations, and RPM versions of Earth Radiation Budget Experiment (ERBE) ADMs.

Thomas Charlock (LaRC) led SARB Working Group discussions on analysis methods, algorithms, and validation. One objective of the CERES investigation is to better estimate broadband shortwave and longwave fluxes at the surface and within the atmospheric column. Because obtaining surface fluxes is much more difficult than measuring TOA fluxes, CERES is pursuing two independent approaches. First, simple parameterization methods are used to directly determine surface LW and SW fluxes from TOA data. Second, cloud physical and narrowband radiative properties derived from cloud imager data are used along with atmospheric temperature and humidity profiles in a radiative transfer model to calculate broadband radiative fluxes at the surface of the Earth, through the atmosphere, and up to the TOA. The Science Team has decided to provide flux divergence calculations initially at the tropopause and at several levels in the stratosphere. Later work will add 500 mb and additional tropospheric levels as warranted by validation studies.

Release 1 SARB software modules have all been developed and are being integrated for the January code run-through. Most meteorological, ozone, and aerosol auxiliary input data for October 1986 have been acquired.

V. Ramanathan suggested that SARB tie the full within-atmosphere code to the results of the parameterized surface-only retrieval for consistency between the two methods. It was concluded that this constraint would be applied if validation showed that the parameterized surface fluxes were more accurate than the model calculations constrained to TOA fluxes. Robert Cess (Stony Brook) presented new results to support the theory of anomalously high atmospheric SW absorption. David Randall (Colorado State) discussed a recent statement from 30 international scientists at climate modeling and numerical weather prediction centers stressing the importance of continuing calibrated broadband global measurements of the Earth's radiation budget. He and his colleagues concluded that such measurements are fundamental and essential for monitoring, understanding, and predicting the state of the climate system and should continue without interruption into the indefinite future.

Shashi Gupta (LaRC) developed new LW surface emissivity maps for Release 1. The new maps incorporate spatial variability based on surface/vegetation classification maps along with emissivity measurements. Temporal variability is included by superimposing seasonally-dependent snow/ice maps on the underlying surface maps. David Rutan (representing Louis Smith) presented a methodology for determining surface spectral reflectance for various scene types. The spectral reflectivities are averaged over the CERES footprint and used as surface boundary conditions for the Fu and Liou radiative transfer model.

Shi-Keng Yang (representing Jim Miller) discussed the global surface reflectance and surface albedo data at the National Centers for Environmental Prediction (NCEP) as well as Reanalysis Project validation results using ERBE data. Maurice Blackmon (NCAR) compared a 4D data assimilation technique and NCEP/NCAR (National Center for Atmospheric Research) Reanalysis Project results and showed that the two methods produce substantially different atmospheric heating rates. The ongoing Reanalysis Project provides a unique opportunity for studying and evaluating the cloud and radiation fields generated from a state-of-the-art global data assimilation system.

One of the most difficult SARB problems is the case of determining surface downward longwave fluxes when there is a middle-level or upper-level optically thick cloud present over a lower-level cloud. Charlock showed that the largest uncertainty in the LW fluxes at the surface is not in estimating the thickness of an observed cloud to get cloud base from cloud top, but in knowing the amount of cloud overlap, i.e., the presence of multi-layer clouds. The Cloud Working Group is investigating the use of a passive microwave instrument (or a more sophisticated imager/microwave combined multi-channel retrieval) to detect the lower cloud and measure LWP over oceans.

David Kratz (LaRC) presented a validation plan for the TOA-to-surface parameterization approach that depends on the availability of simultaneous TOA and surface measured LW and SW net fluxes. A limited validation data set has been produced from measurements taken at the Atmospheric Radiation Measurement (ARM) Cloud and Radiation Testbed (CART) site in Oklahoma in 1994. Cess provided information on SW narrowband instruments to be used during the ARM Enhanced Shortwave Experiment (ARESE) in September 1995 and at other ARM sites in the future.

Charlock and Charles Whitlock (LaRC) presented a SARB validation plan that relies on the CERES/ARM/GEWEX (Global Energy and Water Cycle Experiment) Experiment (CAGEX) and several surface networks. The Working Group strongly endorsed the CAGEX activity. Whitlock discussed surface fluxes from a number of measurement networks including ARM, NOAA Integrated Surface Irradiance Study (ISIS) sites, World Climate Research Program (WCRP) Baseline Surface Radiation Network (BSRN) at selected sites around the world, and an operational instrument tower (Walker Site) in Virginia. He also showed initial albedo and bidirectional reflectance results from the CERES helicopter tests. The Working Group suggested that some field experiments should be delayed, if possible, to have overlap with CERES for validation.

Takmeng Wong and David Young (LaRC) led the TISA Working Group where the agenda encompassed satellite sampling, temporal interpolation, algorithm development, code generation, and validation planning. Wong used cloud parameters over the ARM site derived using the Hybrid Bispectral Threshold Method (HBTM) to evaluate methods for interpolating cloud amount, visible optical depth, infrared emissivity, cloud top height, cloud base height, and cloud effective height. Results showed that cloud temporal interpolation errors are large for the CERES single satellite product and decrease as the number of satellites increases. He showed that the use of cloud products from geostationary satellites could significantly reduce CERES cloud interpolation errors. Claudia Stubenrauch (representing Robert Kandel) showed that ERBE TOA SW flux diurnal interpolations could be improved by using 3-hourly International Satellite Cloud Climatology Project (ISCCP) cloud cover and cloud visible reflectance data.

The development of algorithms for TISA Release 1 Subsystems is on schedule. Subsystems were sized and coding/testing is in progress and on schedule for transmittal to the DAAC in January. The issues of module duplication in different Subsystems and of using local time versus GMT are being reviewed. QC formats and validation graphics were finalized.

Minnis presented a TISA validation plan outline. Validation is required for LW and SW TOA total-sky flux, LW and SW TOA clear-sky flux, window radiance, LW surface flux, atmospheric flux, cloud amount (total and levels), cloud particle size, cloud liquid and ice water path, cloud emittance and optical depth (daytime only), and cloud height and thickness. Pre-launch validation involves applying CERES algorithms to ERBE data and comparing the results to řtruthÓ data from other satellite and surface observations. When CERES data are available, qualitative evaluations will be made based on comparisons to previous ERBE data, ERBE wide-field-of-view data (if available), Scanner for Radiation Budget (ScaRaB) data (if available), GOES cloud properties, ARM data, CAGEX, and operational ground station observations. EOS AM results will also be validated by comparison with TRMM data. A tentative list of CERES validation regions representing a range of surface types and climate regimes was identified.

Validation plans are due by March 31, 1996. The next CERES Science Team meeting is scheduled for March 13-15, 1996 at the Goddard Space Flight Center. Major topics will include approval of the CERES validation plan as well as discussions of Release 1 algorithm tests.