NEWS NEWSWorld Climate Research Programme__WCRP

Vol. 12, No. 4 November 2002

Global Energy and Water Cycle Experiment

NEW 18-YEAR ISCCP PRODUCTSHOWS WIDE SWINGS

IN SURFACE ANDATMOSPHERIC FLUXES,BUT LITTLE CHANGE IN

TOP-OF-ATMOSPHERE(see article on page 7)

CROSS-COMPARISON OF GEWEX20+ YEAR GLOBAL DATA SETS

WITH NEW EOS DATA ISBEGINNING WITH TRMM AND GPCP

Comparison provides a basis for improving GEWEXdata sets and a 20-year context for new Earth Observ-ing System (EOS) data. Mean rainfall in mm/day. GPCP(top panel), TRMM (middle panel) and GPCP-TRMMdifference (bottom panel). See article on page 5.

What's New

• Up to 6-month daily GCM rainfall predictions show success in streamflow forecasting in Brazil.

• Bob Schiffer begins serving as IGPO Director.

• Piers Sellers soars into space on Atlantis.

• CEOP focusing on EOP-1 (seasonal data set) for early dissemination.

November 20022

COMMENTARY

ContentsPAGE

Commentary – Robert A. Schiffer to Serveas Director Of IGPO 2

IAA Elects Soroosh Sorooshian as Member 2Up To 6-Month Daily GCM Rainfall PredictionsShow Success In Streamflow Forecasting InBrazil 3

Former GEWEX Scientist Walks In Space 5TRMM And GPCP Initial Cross-Comparison 5Moustafa Chahine Awarded WilliamNordberg Medal 6

New ISCCP Global Radiative Flux DataProducts 7

Workshop/Meeting SummariesGEWEX Radiation Panel Meeting 8GEWEX Hydrometeorology Panel Meeting 9GSWP-2 Kickoff Workshop 11CEOP Status Session at GHP-8 13

GAME Results Available on CD-ROM 14CEOP Plans Presented At International Forums 15New Web Location for ISLSCP Initiative II Data 15Meetings Calendar 15

Join me in welcomingDr. Robert A. Schiffer as theinterim Director of the Inter-national GEWEX Project Office(IGPO) for the next year whilethe process to select a perma-nent director is underway.Dr. Paul D. Try, founder ofthe IGPO and its Director forthe past 12 years, is stepping

down but will continue to support IGPO activities ina part-time capacity for the next year. The inter-national visibility and success of the GEWEX Programis due to the tireless efforts of Paul Try. Thecommunity is grateful to Paul for his vision andleadership of IGPO.

Dr. Schiffer is currently the Chief Scientist ofthe Goddard Earth Sciences and Technology Cen-ter (GEST) at the University of Maryland, BaltimoreCounty. He retired from the National Aeronauticsand Space Administration (NASA) in early 2002,and has over 39 years of experience in planningand managing national and international atmosphericand climate research programs, including 11 yearsat the Jet Propulsion Laboratory as a researchengineer/scientist involved in the NASA PlanetaryExploration Program, and 28 years at NASA Head-quarters, planning and managing Earth scienceresearch and applications programs. These includedinteragency and international coordination, and sci-entific oversight of experimental global observingsystems, numerical modeling and data assimilationsystems, physical process studies involving field andairborne campaigns, and data management systems.

While at NASA, Dr. Schiffer was awarded theDistinguished Service Medal and two Medals forOutstanding Leadership for his contributions to in-teragency and international climate research andglobal environmental observations. He earned hisMS and Ph.D. degrees (in atmospheric sciences)from the University of California at Los Angeles,and BAE and MAE degrees (in Aeronautical Engi-neering) from the Polytechnic Institute in New York.

ROBERT A. SCHIFFER TO SERVEAS DIRECTOR OF IGPO

Soroosh Sorooshian, ChairmanGEWEX Scientific Steering Group

Over the past 18 years, Dr. Schiffer was sec-onded part-time to the World MeteorologicalOrganization as Director of the World Climate Re-search Programme (WCRP) Radiation ProjectsOffice. Dr. Schiffer’s past assignments to the WCRPJoint Planning Staff and NASA’s Office of EarthScience Enterprise provide a nearly unmatched com-bination of international and interagency experienceto couple with his direct and long association withGEWEX and its remote sensing components to ensurecontinuing and strong leadership for the IGPO. Bob’sspace systems and radiative transfer backgroundblends especially well with my hydrology expertiseas we attempt to implement and address the criticalscience questions for Phase II in GEWEX.

IAA ELECTS SOROOSH SOROOSHIANAS MEMBER

Dr. Soroosh Sorooshian,chairman of the GEWEXScientific Steering Group, waselected as a CorrespondingMember of the InternationalAcademy of Astronautics (IAA)at the World Space Congress,held on October 13, 2002 inHouston, Texas. Election to theAcademy is in recognition of anindividual's record of service and achievement andleadership in space-related activities.

3November 2002

Collaborative research in Brazil has now shownthat it is possible to use up to 6-month daily rainfallpredictions from an atmospheric general circulationmodel (AGCM), together with a model of riverbasin response, to yield useful estimates of futureriver flow several months in advance, when appliedto a portion (75,000 km2) of the Uruguay Riverdrainage basin lying in southern Brazil.

With over 90 percent of the energy consumedby Brazil's population coming from hydropower,efficient exploitation of Brazil's water resources iscritically important and planning future energy pro-duction would be more efficient if the inflow toreservoirs used for hydropower generation could bepredicted. Where hydropower is generated, thebenefits of prior knowledge of reservoir inflows,even when knowledge is incomplete, are that(1) spillage is minimized; (2) reservoirs can operatewith greater head of water for longer periods; and(3) more energy can be generated at times whenenergy prices are higher. Since, in mixed generat-ing systems, the operational costs of hydropowerproduction are lower than for thermo-electric andother generating systems, there is a strong eco-nomic motive for maximizing the proportion ofenergy generated from hydropower. One way ofcontributing to this maximization is to make use ofhydrologic forecasts when decisions are to be madeconcerning power production, particularly where thereis a mixture of generating systems.

It is relatively easy in large rivers to predictflows several days or even several weeks ahead,but to be useful for planning purposes, predictionsof flow are needed several months in advance, andthis in turn needs predictions of future rainfall.While numerical models for weather prediction giveestimates of future rainfall for several days, andclimate prediction models yield rainfall sequencesextending up to several months, rainfall predictionremains one of the most difficult variables to fore-cast in quantitative terms, although important advancesin this difficult field have been reported recently.

Collaborative research between the Instituto dePesquisas Hidráulicas of the Federal University of

Rio Grande do Sul (IPH-UFRGS), the Institute forAtmospheric and Geophysical Research of the Uni-versity of São Paulo (IAG-USP), and the Centerfor Weather Forecasting and Climate Research ofthe Brazilian Agency for Space Research (CPTEC-INPE) is producing new capabilities. Output fromthe CPTEC's AGCM included sequences of futuredaily rainfall for periods extending up to 6 monthsahead, and these were used as input to a rainfall-runoff model for large basins, developed byIPH-UFRGS, to give estimates of future flow in theUruguay River. The period 1995–2001 was used tofit and test the modeling procedure. The usefulnessof predicted flows was assessed by calculating theroot-mean-square error (RMSE) between the pre-dicted flows, and the flows observed in the UruguayRiver. This RMSE was compared with the RMSEobtained when the mean flows recorded over theperiod of past record were used as predictions offlow in future periods. The extent to which the firstRMSE was smaller than the second is a measureof how well the modeled flows outperformed fore-casts based on the historic mean flows.

The highly nonlinear nature of meteorologicalprocesses causes uncertainty wherever hydrologicforecasts are derived from rainfall sequences derivedfrom predictive models of weather or climate. Becauseof the nonlinearities, predicted rainfall sequencesare strongly dependent on initial conditions. To evaluatethe uncertainty, predictions were repeated with initialconditions taken from five different days, resultingin an ensemble of predictions consisting of fiveindividual members. Each member of the ensemblewas used to generate a flow sequence, and vari-ability amongst the set of predicted flows thusgenerated gave a measure of their uncertainty. Beforeusing the forecasts of future rainfall given by theCPTEC AGCM, however, it was first necessary tocorrect them, because the CPTEC AGCM underes-timated rainfall over almost all the basin, particularlyin winter, although it reproduced inter-annual vari-ability in regional rainfall relatively well (Cavalcantiet al., 2002). A statistical procedure was used tocorrect for the underestimation of rainfall, whichconsisted of the following: At each grid-point of theCPTEC AGCM for which daily rainfall was pre-dicted, the probability distribution function (pdf) ofthe predicted daily rainfalls, in all members of theensemble, was calculated month by month. The pdfwas also calculated, at each of these grid-points,for the daily rainfalls interpolated from the 78 raingauges distributed over the drainage basin; eachdaily rainfall predicted by the CPTEC model wasthen simply transformed into the value correspond-ing to it in the pdf of interpolated rainfalls.

UP TO 6-MONTH DAILY GCMRAINFALL PREDICTIONS SHOW

SUCCESS IN STREAMFLOWFORECASTING IN BRAZIL

C.E.M. Tucci1, R.T. Clarke1, W. Collischonn1,P.L. da Silva Dias2, and G. Sampaio3

1IPH-UFRGS, Porto Alegre, 2USP, São Paulo,3CPTEC-INPE, São José dos Campos, Brazil

November 20024

Figure 2. The period of 1999–2001.

Figure 1. For the period 1995–98, the first of the fivecolumns shows the RMSE obtained when the historic meanflows are used as forecasts of future flow.

Figures 1 and 2 summarize the value of flowpredictions given by the combined AGCM and rain-fall-runoff modeling components. Both figures referto one flow-gauging station (Iraí) on the UruguayRiver, and refer to mean flows calculated over3-month periods.

In the case of Figure 1, which refers to theperiod 1995–98, the first of the five columns showsthe RMSE obtained when the historic mean flowsare used as forecasts of future flow; this error isclearly large, almost 1600 m3s-1. The second col-umn shows the RMSE when the rainfall sequencespredicted by the CPTEC model are used, withoutthe statistical correction referred to above, as inputto the rainfall-runoff model; here the RMSE iseven larger than in the case of the historic means.The third column shows the RMSE when the rain-fall sequences predicted by the CPTEC model arecorrected statistically, as described above. The RMSEis greatly reduced relative to the RMSEs in thefirst and second columns. The fourth column ofFigure 1 shows the RMSE that would have beenobtained if future rainfall had been known exactly;this column measures errors of interpolating rainfallbetween rain-gauge sites (needed to run the hydro-logic model) and imperfections in the modelrepresentation of physical hydrologic processes. Thefifth column of Figure 1 shows the RMSE obtainedwhen 3-month river flows are related, by means ofa multiple regression scheme, to the principal com-ponents of sea-surface temperatures (SSTs) in thePacific and Atlantic Oceans. For the period inquestion 1995–1998, this gave a RMSE smallerthan that given by using historic mean flows topredict future flows, but larger than the RMSEobtained by using statistically corrected rainfall pre-dictions as input to the IPH rainfall-runoff model.

Thus, column 4 of Figure 1 gives the RMSEassuming perfect knowledge of future rainfall. The

RMSE obtained by using predictions of future rain-fall, after statistical correction, is greater (column3), but is very much less than that obtained byusing the historic means to estimate future flows(column 1). Time-series models of ARMA type werealso explored, with results not shown in Figure 1,however the Uruguay River basin responds veryrapidly to rainfall, and ARMA forecasts only1 month ahead were very close to the historicmeans.

Figure 2 gives an even better demonstration ofthe value of 3-month-ahead streamflow predictionsfor the Uruguay River. The figure refers to theperiod 1999–2001, for which no rain-gauge observa-tions were available. As in Figure 1, the first columnof Figure 2 gives the RMSE obtained when pastmean flows are used to predict future flows, andthe second column gives the RMSE when the sta-tistically uncorrected rainfalls predicted by the CPTECmodel were used as input to the hydrologic model;here the RMSE is again larger than that given byusing the historic means. The third column of Fig-ure 2, however, shows the RMSE obtained whenstatistically corrected rainfall predictions were usedby the hydrologic model to predict streamflow; andthe important point here is that the probability den-sity functions used to make the corrections werederived from rainfall recorded prior to the period oftest 1999–2001. While the RMSE is still high, slightlymore than 600 m3s-1, it is substantially less than 900m3s-1 obtained by using past flow records to predictfuture flow. As in Figure 1, the fourth columnshows the RMSE obtained by using an empiricalregression relationship between mean 3-monthly flowSST principal components; the resulting RMSE issmaller than that where statistically uncorrected rainfallrecords were used (second column of the figure),but is larger than the RMSE obtained by usinghistoric mean flows to predict future mean flows.

5November 2002

It can, therefore, be concluded that the useof daily rainfall sequences predicted by theCPTEC AGCM, after they were statisticallycorrected for bias, as input to the rainfall-run-off model, gave better estimates of future flowat Iraí on the Uruguay River, than where his-toric mean flows were used as estimates offuture flow, even though the errors of esti-mate remain large. The error reduction in streamflowprediction is also related to the ability of the AGCMto predict seasonal rainfall with higher skill for someareas on Earth in comparison to others. This is truefor Southeastern South America, where the Uru-guay River basin is located, as shown by Cavalcantiet al. (2002), but cannot be generalized for alldrainage basins. That is, the ability of the statisticalcorrection of model-calculated rainfall bias for anyparticular basin will be limited if the AGCM haspoorer skill in predicting seasonal variability of rainfallfor that basin. (This research, was funded by theBrazilian Agency for Electrical Energy – ANEEL.)

ReferencesCavalcanti, I.F.A., J.A. Marengo, P. Satyamurty, C. Nobre, I.Trosnikov, J.P. Bonatti, A.O. Manzi, T. Tarasova, L.P. Pezzi,C.D. Almeida, G. Sampaio, C.C. Castro, M.B. Sanches, H.Camargo, 2002. Global climatological features in a simulationusing the CPTEC-COLA AGCM. Journal of Climate (in press).

TRMM AND GPCP INITIALCROSS-COMPARISON

Robert F. Adler1, George Huffman2,and David Bolvin2

1NASA/Goddard Space Flight Center2Science Systems and Applications, Inc.

The Tropical Rainfall Measuring Mission (TRMM)has been orbiting the Earth for almost 5 years. Thejoint U.S.-Japan mission was launched in late No-vember 1997 and produced its first full month ofdata in January 1998 (Kummerow et al., 2000).Now with nearly 5 years of data this compari-son of the TRMM estimates with those of thecommunity analysis produced routinely by theGEWEX Global Precipitation Climatology Project(GPCP) (which currently does not use TRMMdata) is becoming the baseline for both incor-porating TRMM data into the GPCP productand extending the understanding provided byTRMM back for 20 plus years.

TRMM standard products include estimates fromsingle instruments on TRMM (radar and passivemicrowave), from a combination of those two in-struments, and from a multisatellite product thatcombines information from TRMM and geosynchro-nous infrared (IR) satellites. In the current Version5 of the TRMM standard products these variousestimates produce a spread of mean rainfall in thetropics with a range of about 20 percent over theocean (Adler et al., 2003a). Although this range issomewhat smaller than a pre-TRMM analysis ofthe Special Sensor Microwave/Imager (SSM/I)-basedrainfall estimates, it should be reduced as differ-ences between the passive microwave and radaralgorithms are resolved. A major step in that di-rection will occur in the Spring of 2003 when theentire TRMM data set will be reprocessed with theimproved Version 6 of the TRMM algorithms.

For this comparison the TRMM Version 5multisatellite product was chosen because it is basedon the radar/passive microwave combination cali-brating or adjusting the geosynchronous IR data toincrease sampling. The mean value of the radar/passive microwave combination falls in the middleof the range of the TRMM estimates. This TRMMproduct (TRMM product identifier 3B-43) also uti-lizes rain gauge information over land, as does theGPCP product. The GPCP satellite analysis alsouses a multisatellite approach with passive micro-

Dr. Piers Sellers, whostepped down as chair ofthe International SatelliteLand-Surface ClimatologyProject in 1996 to begintraining as an astronautmission specialist in Hous-ton, was finally launchedinto space on October 6thaboard the space shuttleAtlantis (STS-112). Piers

was part of as i x - m e m b e rcrew whosemain task wasthe installationand hookup of a15-ton, 45-footlong truss con-taining the newexternal coolingsystem for the space station. During the 11-daymission, Sellers and his colleague made threespacewalks to install and activate the truss.

FORMER GEWEX SCIENTIST WALKSIN SPACE

November 20026

MOUSTAFA CHAHINE AWARDEDWILLIAM NORDBERG MEDAL

Dr. Moustafa Chahine,former chairman of theGEWEX Scientific SteeringGroup (1989–1999) and nowsenior research scientist atNASA’s Jet Propulsion Labo-ratory and Science TeamLeader for NASA’s Atmo-spheric Infrared Sounder(AIRS), was awarded the Wil-liam Nordberg Medal in space

science from the Committee on Space Research(COSPAR) of the International Council for Sci-ence. This medal commemorates the work of thelate William Nordberg and is awarded each year toa scientist who has made a distinguished contribu-tion to the application of space science in a fieldcovered by COSPAR.

wave information (from low-orbit SSM/I instruments)calibrating the IR information (Huffman et al; 1997;Adler et al; 2003b).

The mean daily rainfall from January 1, 1998 toJuly 31, 2002 is shown in the figure at the top ofpage 1 for TRMM and GPCP. Both analyses havevery similar patterns, as expected, with maxima inthe Intertropical Convergence Zone (ITCZ) of thePacific and Atlantic Oceans, in the eastern IndianOcean and over land areas in Brazil, Africa andIndonesia. The difference map in this figure (bot-tom panel) indicates a pattern of both positives andnegatives in the deep tropics over the ocean andgenerally positive values (GPCP larger) in midlatitudes.In the light rain areas in the western part of thesub-tropical oceans the TRMM estimates tend tobe larger, while in the very dry eastern parts of theoceans GPCP shows a slightly larger amount. Thetwo analyses produce nearly equal mean rainfallover the oceans between 20°N and 20°S, but withGPCP being higher in the midlatitudes, especially inthe Southern Hemisphere.

Even within the 20°N–20°S region over oceansthere are variations in relative rainfall amount be-tween the two estimates. In the heavier rain areasof the western Pacific Ocean, both above and be-low the Equator, the TRMM-based estimate is higher,whereas over the areas of somewhat less rain eastof the Philippine Islands and southeast of the SouthPacific Convergence Zone (SPCZ) maximum, thedifference is reversed. This relation between thetwo products may result from the GPCP estimatesaturating in very high rain areas. Other smaller,but interesting, areas where TRMM is larger arefound along coasts in the eastern Arabian Sea ad-jacent to India, in the eastern part of the Bay ofBengal along Indochina and in the Gulf of Panama.The difference here may be related to the ability ofthe TRMM radar to delineate areas of shallowconvection, especially in the monsoon regions of theIndian Ocean.

Comparisons between TRMM and GPCP willcontinue as algorithms and analysis procedures arerefined. The authors and others are workingon a method to incorporate the 5 years ofTRMM data into GPCP. This will likely be tiedto a parallel effort to produce GPCP global analy-ses at a 3-hr time resolution for recent years usingthe TRMM-based 3-hr multisatellite research analy-sis, which will be available in Version 6 in 2003. Areal-time version of the 3-hr multisatellite analysis(images and data) is currently available via the

TRMM web site (http://trmm.gsfc.nasa.gov). Thereal-time analysis uses TRMM to calibrate or ad-just both the polar-orbit microwave and geosynchronousIR rain estimates and merges the information usingas much microwave data as possible and filling inthe remaining gaps with the lower quality IR-basedestimates. Additional microwave data will be addedas available. It is expected that over the next fewyears GEWEX’s GPCP data products will be en-hanced through the use of TRMM data. Theseefforts in precipitation analysis fit very well withthe development of the Global Precipitation Mea-surement mission to be launched in about 2008.References

Adler, R.F., C. Kummerow, D. Bolvin, S. Curtis, C. Kidd, 2003a.Status of TRMM Monthly Estimates of Tropical Precipitation,Meteorolgical Monographs, Symposium on Cloud Systems, Hurri-canes and TRMM (in press).

Adler, R.F., G.J. Huffman, A. Chang, R. Ferraro, P. Xie, J.Janowiak, B. Rudolf, U. Schneider, S. Curtis, D. Bolvin, A. Gruber,J. Susskind, P. Arkin, 2003b. The Version 2 Global PrecipitationClimatology Project (GPCP) Monthly Precipitation Analysis (1979-Present). J. Hydromet. (submitted).

Huffman, G.J., R.F. Adler, P. Arkin, A. Chang, R. Ferraro, A.Gruber, J. Janowiak, A. McNab, B. Rudolf, U. Schneider, 1997:The Global Precipitation Climatology Project (GPCP) Version 1Data Set. Bull. Amer. Meteor. Soc., 78, 5-20.

Kummerow, C., J. Simpson, O. Thiele, W. Barnes, A.T.C. Chang,E. Stocker, R.F. Adler, A.T.C. Chang, E. Stocker, R.F. Adler, A.Hou, R. Kakar, F. Wentz, P. Ashcroft, T. Kozu, Y. Hong, K.Okamoto, T. Iguchi, H. Kuroiwa, E. Im, Z. Haddad, G. Huffman,T. Krishnamurti, B. Ferrier, W.S. Olson, E. Zipser, E.A. Smith,T.T. Wilheit, G. North and K. Nakamura, 2000. “The Status ofthe Tropical Rainfall Measuring Mission (TRMM) after 2 Yearsin Orbit”. J. Appl. Meteor., 39, 1965–1982.

7November 2002

NEW ISCCP GLOBAL RADIATIVE FLUXDATA PRODUCTS

Yuanchong Zhang1 and William B. Rossow2

Department of Applied Physics and AppliedMathematics, Columbia University1

NASA Goddard Institute for Space Studies (GISS)2

(NASA GISS climate model); and the spectral depen-dence of land surface albedo and emissivity byland-cover type (NASA GISS climate model). Theresults include the all-sky and clear-sky, upwelling anddownwelling, total shortwave (SW = 0.4 – 5 µmwavelength) and total longwave (LW = 5 – 200 µmwavelength) radiative fluxes at five levels: surface,680 mb, 440 mb, 100 mb and top-of-atmosphere. Allof these results are reported with a resolution of 3hours and 280 km (equal-area map equivalent to 2.5°latitude-longitude at the equator) in four data products:(1) TOA RadFlux, (2) Surface Radiative Fluxes (SRFRadFlux), (3) Radiative flux profiles including TOAand SRF fluxes (RadFlux Profiles), and (4) the com-plete input data collection (RadFlux Inputs). The firstthree products include a summary of the most rel-evant input physical parameters, whereas the fourthproduct contains the exact inputs used to calculate allthe fluxes.

Several papers are being prepared to describe thefeatures of the radiative transfer model and the inputdata sets and to provide results of comparisons withother more direct determinations of the surface andtop-of-atmosphere radiative fluxes (Zhang et al., 2003).More information can be found at

http://isccp.giss.nasa.gov/projects/flux.htmlor by contacting the authors (yzhang@giss.nasa.govor wrossow@giss.nasa.gov). These data products areavailable from the authors in a preliminary format.Note: In December 2002, the GEWEX Surface RadiationBudget (SRB) Project will release a 1x1 degree, 3-hourly,SW and LW surface radiative flux data set covering theperiod 1983–1995. These fluxes are also based on theISCCP cloud products, but use different sources ofinformation about the atmosphere and surface. Thesedata will be available through the Atmospheric SciencesData Center at NASA Langley Research Center at http://eosweb.larc.nasa.gov/PRODOCS/srb/table_srb.html.

References

Han, Q.-Y., W.B. Rossow, and A.A. Lacis, 1994. Near-global surveyof effective cloud droplet radii in liquid water clouds using ISCCP data.J. Climate, 7, 465-497.

Han, Q., W.B. Rossow, J. Chou, K-S. Kuo, and R.M. Welch, 1999.The effects of aspect ratio and surface roughness on satellite retrievalsof ice-cloud properties. J. Quant. Spectrosc. Radiat. Trans., 63, 559-583.

Wang, J., W.B. Rossow, and Y-C Zhang, 2000. Cloudset verticalstructure and its variations from a 20-year global rawindsonde data set.J. Climate, 13, 3041-3056.

Zhang, Y-C, W.B. Rossow, A.A. Lacis, V. Oinas, and M.I. Mishchenko,2003. Calculation of radiative flux profiles from the surface to top-of-atmosphere based on ISCCP and other global data sets: Refinementsof the radiative transfer model and the input data. J. Geophys. Res.(in preparation).

The International Satellite Cloud Climatology Project(ISCCP) has produced a new 18-year (1983–2000)global radiative flux data product called ISCCP FD.The figures on the cover and back page illustrate aunique aspect of this product, which provides physi-cally consistent surface and top-of-atmosphere (TOA)radiative fluxes by showing the global monthly meannet shortwave (SW) and net longwave (LW) anoma-lies at the surface, in the atmosphere and at the TOAover the whole time period. Notable features are:(1) a decrease of the net SW at the surface andTOA, as well as in the atmosphere produced by theMt. Pinatubo volcanic aerosols in 1991–92; (2) anoverall increase of the net SW at TOA and thesurface, but not in the atmosphere, from the 1980s to1990s associated with a decrease in low-latitude cloudcover; (3) three (possibly four) decreases in net LWat the surface and increases in the atmosphere, butnot at TOA; and (4) a small decrease of net LW atTOA and in the atmosphere and a larger increase ofnet LW at the surface occurring in the late 1990s.Another unique feature of the flux profile product isthat it provides, for the first time, a comprehensivedetermination of the synoptic scale variations of thevertical profiles of radiative diabatic heating, albeitwith crude vertical resolution, but sufficient to repre-sent radiative heating in the lower, middle or uppertroposphere and the stratosphere.

The product was created by employing the NASAGISS climate Global Circulation Model (GCM) radia-tive transfer code and a collection of global data setsdescribing the properties of the clouds and the surfaceevery 3 hours (ISCCP); daily atmospheric profiles oftemperature and humidity (National Oceanic and At-mospheric Administration Television InfraRed ObservationSatellites (TIROS) Operational Vertical Sounder); dailyozone abundances (Total Ozone Mapping Spectrom-eter); a climatology of cloud vertical layer distributionsfrom rawinsonde humidity profiles (Wang et al., 2000);a climatology of cloud particle sizes (Han et al., 1994,1999); a climatology of stratospheric aerosol and watervapor (Stratospheric Aerosol and Gas Experiment-II);a climatology of the diurnal variations of near-surfaceair temperature (surface weather observations andNational Centers for Environmental Protection [NCEP-1] re-analysis); a climatology of tropospheric aerosols

November 20028

WORKSHOP/MEETING SUMMARIES

The 13th meeting of the GEWEX Radiation Panel(GRP) was hosted by the Institute for Atmospheric andClimate Science ETH. An important focus of this meet-ing was to review the status of the Baseline SurfaceRadiation Network (BSRN) and the Surface RadiationBudget (SRB) Project, as well as the general activities ofthe other GRP projects.

The BSRN and SRB reviews indicated that bothprojects are now well underway and did not raise anysubstantive issues. Both projects are rapidly eliminatingdata processing backlogs and releasing their data sets.There are over 35 active stations participating inBSRN and the archives (located at ETH) have allirradiances available for over 1,800 data-months,equivalent to an average of 4 years of data for eachsite. Some sites have more than a decade of dataavailable. All the related ancillary data sets have alsobeen collected and are being placed on the BSRN ar-chives web site for ftp access. The SRB shortwaveproducts are available for almost 10 years; longwaveproducts will follow more slowly, but should be availableearly next year. Current plans are to process surfaceradiative fluxes for the period July 1983 through Decem-ber 1995; however, the GRP recommended that SRBcontinue processing the remaining years (1996–2001) byswitching to another source of atmospheric data.

The review of the other radiation projects highlightedseveral important activities that are underway. Recentanalyses of the 20+ year record of top-of-atmo-sphere radiative fluxes from Nimbus-7, the EarthRadiation Budget Experiment (ERBE), the Scannerfor Radiation Budget (ScaRaB) and the Clouds andEarth's Radiant Energy System (CERES) show notonly features associated with ENSO events and theEl Chichon and Mt. Pinatubo eruptions, but alsointer-decadal changes that appear to be associatedwith changes in clouds found in the InternationalSatellite Cloud Climatology Project (ISCCP) dataset and upper atmosphere water vapor found in theanalysis of High resolution Infrared Radiation Sounder(HIRS) data by John Bates [National Oceanic andAtmospheric Administration (NOAA)]. The report ofthe Intercomparison of 3D Radiation Codes (I3RC) Project(http://climate.gsfc.nasa.gov/I3RC/index.html) contained sev-eral notable items: (1) the first two phases of the projectto compare 3-dimensional (3-D) radiative transfer codes

are now complete and papers are being submitted forpublication of the results and conclusions; (2) a modeltest kit is now on-line to allow other investigators to testtheir 3-D radiative transfer (RT) codes; and (3) effortsare now being made to link with cloud-large-eddy-simu-lation and land-surface-vegetation modelers to examinethe role of 3-D radiative effects in planetary boundarylayers and in land-atmosphere exchanges. The prelimi-nary results of a survey of recent changes to globalcirculation model RT codes showed that there has beenrapid progress lately to improve the physical detail ofthese codes. GRP plans to complete and publish thissurvey and to foster some renewed attention to testingthese codes more thoroughly. The following issues re-garding radiative transfer modeling were discussed: (1) thenotable lack of quantitative data about the properties ofcirrus cloud particles, especially those smaller than about50–100 µm [some better data may come from the recentCirrus Regional Study of Tropical Anvils and CirrusLayers–Florida Area Cirrus Experiment (CRYSTAL-FACE)]; (2) the lack of a general radiative transfertheory to handle scattering by such small, nonsphericalparticles (also relevant for aerosols); and (3) no agree-ment on a practical way to represent the wide variety ofshapes/sizes of particles encountered in cirrus.

The water projects, especially the Global Precipita-tion Climatology Project (GPCP) and Tropical RainfallMeasuring Mission (TRMM) were reviewed and recentactivities highlighted three points, leading to some recom-mendations: (1) the GRP endorsed GPCP plans toproduce Version 3 products, which will be anchoredon TRMM results, but noted that the passive mi-crowave analysis should be made consistent withmicrowave-based water vapor and cloud water re-sults; (2) actions to obtain better snowfall data arestill not adequate, so the GRP recommended ex-ploring a tighter collaboration with the Climate andCryosphere (CLiC) Study to improve this aspect ofglobal precipitation; and (3) a possible gap in thetropical precipitation record may occur betweenTRMM and the planned Global Precipitation Mis-sion (GPM). A special presentation by Toshio Iguchi(Communications Research Laboratory) and the subse-quent discussion highlighted the fact that validation ofsatellite precipitation measurements is not yet successfuland that more needs to be done to understand the radia-tive transfer physics of this remote sensing problem.Finally, at the request of the GEWEX Scientific SteeringGroup (SSG), the GRP discussed its possible contributionsto a GEWEX-wide precipitation initiative based on theglobal satellite data sets.

The review of ISCCP and the Global Aerosol Clima-tology Project (GACP) indicated that both of these projectshad recently completed production of their data products

GEWEX RADIATION PANEL MEETING31 July – 2 August 2002

Zurich, SwitzerlandWilliam B. Rossow

NASA Goddard Institute for Space Studies

9November 2002

through September 2001, at which time the referencepolar orbiter was changed from NOAA-14 to NOAA-16,necessitating a transfer of the calibration standard. Twoissues were discussed: (1) GACP (also ISCCP) shouldnow make connections with the International Global At-mospheric Chemistry (IGAC) Project, the Global AtmosphereWatch (GAW) Programme, the Stratospheric Processesand their Role in Climate (SPARC) Programme, and theBSRN/Aerosol Robotic Network (AERONET) to exploitthe different sources of information that these groupshave that is related to clouds and aerosols; and (2) bothprojects need to formulate plans for exploiting the newsatellite instruments making advanced measurements ofclouds and aerosols.

The final discussion encompassed a number of pos-sible actions being considered by the GRP to foster moreintegrative analyses of the global data sets. A significantaction to facilitate better connections among the GRPsatellite projects and with other data analysis activitieswithin GEWEX was the decision to organize all of theGRP data activities into a single Working Group on DataManagement and Analysis (WGDMA). The first tasksof WGDMA would be to undertake some commonstatistical analysis tasks, possibly including the cre-ation of a merged collection of data, for all of theGRP global satellite projects (ISCCP, GACP, GPCP,SRB, SeaFlux) and to make plans for the exploita-tion of new satellite observations. This new groupwould also liaise with the GEWEX Cloud System Study(GCSS)/Data Intergration for Model Evaluation (DIME),the Global Land Atmosphere System (GLASS)/Assis-tance for Land-Surface Modelling Activities (ALMA),the International Satellite Land-Surface Climatology Project(ISLSCP), as well as the GRP Data Management Work-ing Group. The first meeting of this group is planned forApril–May 2003 in Asheville, North Carolina, hosted byNOAA's National Climatic Data Center.

Also discussed was what could be done to fill in themissing global land surface data sets needed to completea description of the global energy and water cycles(referred to as LandFlux). The GRP chairman reportedthat the GEWEX SSG had given the lead of this activityback to GRP, and as a result, he would be attendingforthcoming meetings of the GEWEX HydrometeorologyPanel (including the Water and Energy Balance Studyand ISLSCP) to discuss how to proceed.

Two workshops are being organized by GRP tofoster development of integrative data analysis methods:(1) the Workshop on Climate Feedbacks (jointly withWGCM) to be held 18–20 November 2002 in Atlanta,Georgia, USA; and (2) the Workshop on Objective AnalysisTechniques being organized by GPCP to be held 11–13March 2003 at the European Centre for Medium-RangeWeather Forecasts, UK.

GEWEX HYDROMETEOROLOGYPANEL MEETING

10–12 September 2002Palisades, New York

Ronald E. StewartMcGill University, Canada

The Eighth Meeting of the GEWEX Hydro-meteorology Panel (GHP) was hosted by theInternational Research Institute (IRI) for ClimateResearch at the Lamont-Doherty Earth Observa-tory in Palisades, New York. Focused workshopson the Water Resources Application Project (WRAP)and the Water and Energy Balance Study (WEBS)were held the day before the GHP meeting, and aCoordinated Enhanced Observing Period (CEOP)meeting was held the day afterwards.

A review of contributions to GHP/GEWEX byvarious national and international projects andorganizations including those within and outsidethe GEWEX framework was given by represen-tatives of such activities as IRI, the InternationalAssociation of Hydrological Sciences (IAHS); thelimited international regional study, the AfricanMonsoon Multidisciplinary Analyses (AMMA) inWest Africa; the PLATIN study (La Plata RiverBasin) in South America, the Isotope HydrologySection of the International Atomic Energy Agency;and the Data Assimilation Office of NASA GoddardSpace Flight Center; and others. The work ofthe GHP Working Group on Data Managementwas also reviewed. Topics discussed at the GHPmeeting include the following.

Four-dimensional data assimilation (4DDA) ofatmospheric fields was conducted over monsoonalAsia, and Version 1.5 of the GEWEX AsianMonsoon Experiment (GAME) reanalysis has beenreleased and is available as a CD-ROM set (6-hour interval data, gridded at 2.5 degrees; seepage 14). GAME has obtained new scientificresults on the hydrometeorological processes inthe Asian monsoon region from the Tropics tothe Siberian Arctic region. The land-atmosphereinteraction processes in typical climate and veg-etation in monsoonal Asia appear in diurnal throughseasonal time scales. GAME is also showing thatthe water fed rice paddy fields, which are atypical land-surface condition in monsoonal Asia,play an important role in developing and modify-ing precipitation systems.

November 200210

The GEWEX Americas Prediction Project(GAPP) studies of monsoonal processes are be-ing studied in conjunction with the ClimateVariability and Predictability (CLIVAR)/Pan-Ameri-can Climate Study (PACS) in support of the NorthAmerican Monsoon Experiment (NAME). GAPPis currently supporting several studies to assessthe relative role of land surface conditions andsea-surface temperatures in affecting the inten-sities and centers of atmospheric anomalies.Initial studies of the Southwest monsoon indi-cate that the strength of the Low Level Jet inCalifornia is inversely correlated with thestrength of the Gulf of Mexico Low Level Jet.

Currently, there are 102 ongoing field projectsin the Large-scale Biosphere-atmosphere projectin Amazonia (LBA). During the rest of 2002, theLBA-WET to DRY field campaign will study theatmospheric forcing linked to the dry season andthe onset of the rainy season in the southernAmazon, including the role of aerosols from bio-mass burning, and the South American Low LevelJet field that will take place during January–February 2003. This field experiment will allowfor a better knowledge of the moisture transportbetween Amazonia and the La Plata River Basin.

The Mackenzie GEWEX Study (MAGS) re-search in this past year had been focused on thedevelopment of intermediate level coupled models(e.g., coupled atmosphere-surface and coupledsurface-hydrologic models), which have been com-pleted, and preliminary results from these models(e.g., Canadian GEWEX Enhanced Study wateryear simulations) are very encouraging. Thedevelopment of a fully coupled atmosphere-surface hydrologic model is actively underwayright now.

Based on updated data, the Global RunoffData Centre (GRDC) is currently revising esti-mates of mean annual freshwater surface waterfluxes into the world oceans. Based on a newgeographic information system based methodol-ogy involving a digital elevation model, it will bepossible to estimate freshwater fluxes from arbi-trary reaches of the coastline. Depending on dataavailability, GRDC aims to extend this analysis toestimates of individual years later on.

The Global Precipitation Climatology Centre(GPCC) reanalysis based upon the Full Data Prod-uct has been carried out for the period January1986 up to December 1995 (ca. 28,000 to 40,000

stations). The results have been calculated on a0.5°-grid (and 1°-grid) and have been provided toNASA/GSFC for publication on the InternationalSatellite Land-Surface Climatology Project (ISLSCP)Initiative II (see page 15) CD-ROM.

The International Association of HydrologicalSciences (IAHS) Decade on Prediction in UngaugedBasins (PUB) is an international research initia-tive to promote the development of science andtechnology to provide the hydrological data wherethe ground observations are needed but are miss-ing. The Kickoff Meeting of PUB will be held inBrazilia on 20–22 November 2002.

Using representative models, available obser-vations and various reanalyses, the GEWEXContinental-scale International Project (GCIP)WEBS has produced a CD-ROM, with a synthe-sis of the water and energy budget for GCIP.MAGS WEBS has started budget studies overthe Mackenzie Basin using numerical weatherprediction, as well as regional climate models.These show general agreement with availablemeasurements for some parameters but problemswith others (such as orographic precipitation). LBAWEBS is using a combination of station rainfalland the National Centers for Environmental Pre-diction (NCEP) reanalysis to characterize the annualcycle of critical water budget parameters andtheir variations for the northern and southernsections of the Amazon River Basin. GAME WEBSis developing a comprehensive data set through aspecial reanalysis effort by the Japan Meteoro-logical Agency (JMA) and the bringing togetherof many observational measurements taken overmany of the GAME regions of Asia. Model stud-ies are using a hierarchy from high-resolutionreanalyses from JMA and standard reanalysisproducts from NCEP. AMMA is using reanalysisto identify the sources of moisture for the Sahelianrainfall. It has also been shown that GCMs donot reproduce correctly the monsoon onset. Theinfluence of rain variability on the computation ofcatchment water budgets is also explored. TheGlobal Land Data Assimilation System Project isusing a combination of numerical weather predic-tion modeled and observation-derived data to forceand constrain (via data assimilation) multiple, so-phisticated land surface models at 1/4 degree spatialresolution, 60S – 90N, on a 15 minute timestep.

All the Continental Scale Experiments are fo-cusing on their contributions to CEOP.

11November 2002

The Center for Ocean-Land-Atmosphere Studies(COLA), with support from the National Aeronautics andSpace Administration (NASA) Program on TerrestrialHydrology hosted the kickoff workshop of the SecondGlobal Soil Wetness Project (GSWP-2). The project isthe principal element of the large-scale uncoupled landsurface modeling action in the Global Land-AtmosphereSystem Study (GLASS; Polcher et al., 2000) and a majorelement of the International Satellite Land-Surface Clima-tology Project (ISLSCP), both contributing projects ofGEWEX. The overarching goal of the GSWP is toproduce as a community effort the best model estimatesof the global land-surface water and energy cycles(Dirmeyer et al., 1999). This will entail an evaluation ofthe uncertainties linked to the land surface schemes (LSS),their parameters and the forcing variables that drive them.

GSWP-2 will take advantage of the 10-year (1986–1995) ISLSCP Initiative II data set (http://islscp2.sesda.com)and LSS simulations will be conducted at a spatial reso-lution of 1-degree, sans Antarctica. The project will alsoadhere to the Assistance for Land-Surface ModelingActivities (ALMA) data standards developed in GLASS.

The main goal of the kickoff workshop was to dis-cuss details of the project planning with members of thescientific community who would be participating in GSWP-2, address unresolved issues and complete the draft Scienceand Implementation Plan for the project. Specifically, theworkshop provided a forum to resolve uncertainties andrequirements for input data to the various LSSs; addressmore general modeling group issues; introduce the GSWP-2 Inter-Comparison Center (ICC), housed at the Universityof Tokyo, and the data submission process; solidify thenew remote sensing element of GSWP-2; outline thevalidation program; narrow options for model sensitivitystudies; and discuss other science that can leverage offof GSWP-2. The workshop also established participantsand leadership for elements of the project, and demon-strated the data server and software tools that shouldmake model participation and data access much easierthan for past experiments.

Execution of GSWP-2 will follow in three streams—data, modeling and evaluation—given in the figure onpage 16. COLA is processing the 3-hourly meteorologi-cal forcing and complete boundary condition data forGSWP-2, including an extension of the data back in time

GSWP-2 KICKOFF WORKSHOP30 September – 1 October 2002

Calverton, Maryland, USAPaul Dirmeyer1, Taikan Oki2, and Xiang Gao1

1Center for Ocean-Land-Atmosphere Studies,Calverton, Maryland, USA

2University of Tokyo, Kyoto, Japan

to 1982. LSS integrations will begin at July 1982, andloop through the first 12 months of forcing data until themodeler is satisfied that soil moisture has spun up andsufficiently equilibrated. A lesson from the GSWP pilotproject was that this spin-up process overly amplifies theimpact of climate anomalies from that year on the landsurface state variables. Therefore, the models will thenproceed with their integrations forward from June 1983 –December 1985 so as to converge to a realistic "landclimate" at the start of the evaluation period. The 10-year baseline integration, which will be evaluated withinthe group of GSWP participants and later released to thecommunity at large, covers the period from January 1986to December 1995. Daily global output will be reportedfrom all models during this period. In addition, for thelast year (1995) there will be an "intensive model outputperiod" where model results will be reported at a 3-hourinterval, but likely with a reduced set of variables. Thesedata will be especially useful for validation and remotesensing applications. We may also specify a subset ofpoints for full 10-year histories at 3-hour output. Thedata server system will make such limited re-integrationssimple for the models to perform.

A major product of GSWP-2 will be a multi-modelland surface analysis for the 1986–1995 period. This willbe a land surface analog to the atmospheric reanalyses.There will be a climatological annual cycle data set, anda larger data set for the entire series. Compiling theresults of multiple LSSs to produce a single analysis willprovide a model-independent result. Of particular value,uncertainty estimates can be put on all of the fields,based on inter-model spread. Additional uncertaintiesregarding forcing data can be quantified, based on theresults of the sensitivity studies. The act of constructingan ideal multi-model analysis is a research topic in itself,and much can be learned from the experience of multi-model ensembling in the atmospheric and oceanic modelingcommunities. There will be three main modes of in situvalidation of participating LSSs:

Field campaigns. The GSWP-2 period overlapsa number of relevant field campaigns, including olderGEWEX experiments, which can provide validation

Participants attending the GSWP-2 workshop.

November 200212

data. Comparison of measured meteorological vari-ables from these campaigns with the reanalysis-basedforcing data will also provide an evaluation of thoseproducts.

Observational networks and long-term monitor-ing. There are also long-term monitoring networks of soilmoisture, carbon, radiative and turbulent fluxes that canprovide local or regional validation for LSSs. These willbe predominantly available for the latter years of theperiod.

Streamflow. Runoff fluxes from all participatingLSSs will be routed with common river routing schemesto compare with streamflow measurements across a largeportion of the globe, as an assessment of the simulationof annual, seasonal, and interannual variations in surfacehydrology. Similarly, large basin comparison of modelwater storage change with observed atmospheric mois-ture flux convergence minus discharge may also uncoverproblems in the forcing data and models at basin scales.

It is recognized that discrepancies exist between theobserved meteorology and land surface conditions at thevalidation sites, and the 1-degree gridded data that drivethe models, and that those differences can contribute toerrors at least as much as the shortcomings in the variousmodels. Representativeness of gridded data at the plotscale can also be evaluated in these locations. PILPS,in its Phase 2, has conducted and continues to craft localland surface modeling experiments built around nearlycomplete sets of forcing and validation data at a singlelocation (Henderson-Sellers et al., 2002). It is not theintent of this in situ validation program to duplicate thateffort. Rather, using the global forcing data sets, localvalidation may be performed when and where such dataare available.

One of the new thrusts for GSWP-2 is a strongerconnection to applications in remote sensing. The prin-cipal goal of the effort in remote sensing applications isto expand validation beyond those few areas where insitu data on land surface state variables are readilyavailable. In addition to the classical attempts to validatethe typical land-surface state variables using satellite re-trievals, GSWP-2 also intends to expand the capabilitiesof current LSSs. This is to be done by the applicationof algorithms by which LSSs can directly report bright-ness temperatures, like those sensed by instruments inorbit. These may be applied as forward algorithms forinfrared/skin temperature and microwave/soil wetness (andvegetation index for LSSs that predict vegetation phenology).

Modeling sensitivity studies will involve re-integratingthe LSSs over part or all of the global 10-year domain totest the response of the models to changes in forcingdata and surface parameters. Each participant will beencouraged to take part in some or all of the proposed

studies. The sensitivity studies are still being defined, butwill likely include sensitivity to choices in meteorologicalforcing data, such as choice of near-surface reanalysisproduct (NCEP/DOE Reanalysis versus ECMWF ERA-40), impact of hybrid forcing data (combining observedand reanalysis products for precipitation and air tempera-ture), and an assessment of the impact of rain gaugeunder-catch. There are also multiple land surface param-eter data sets available in ISLSCP Initiative II, such asthree choices of global vegetation maps. GSWP-2 willalso investigate sensitivity to basic choices of surfacevegetation data, as well as the impact of inclusion/exclu-sion of sub-grid information (for LSSs that include surfacetile schemes).

Forcing data and model results will be made avail-able to participants as data sets accessible from threeDistributed Oceanographic Data System (DODS) servers(http://www.unidata.ucar.edu/packages/dods/; http://grads.iges.org/grads/gds/) in the United States, Europe,and Japan. Using the ALMA data exchange stan-dards (http://www.lmd.jussieu.fr/ALMA/ ), and DODSdata subsetting capabilities, individual LSSs will beable to run globally each time step, each grid pointfrom start to finish, or in any other sequence ofintegration, accessing the data directly over the Internetwithout the need to download or otherwise a prioriprocess the complete data set on their local system.Software tools and source code to aid in access andproduction of ALMA-standard data, DODS clientsoftware libraries, PILPS consistency checks of modelresults, and interpolation of 3-hourly forcing data toshorter time intervals will also be provided to thecommunity. Additionally, standard and customizablebrowse images will be made available to the publicvia the web.

Release of all forcing and boundary condition data tothe modeling groups is scheduled for February 2003, withbaseline simulations due to the GSWP Inter-ComparisonCenter (ICC) in August 2003 (see figure on page 16).Complete descriptions and current information concerningthis evolving project are available at: http://www.iges.org/gswp/ and anyone interested in participating in land sur-face modeling, validation or other scientific participationshould contact the project at gswp@cola.iges.org.

References

Dirmeyer, P.A., A.J. Dolman, and N. Sato, 1999. The Global SoilWetness Project: A pilot project for global land surface modeling andvalidation. Bull. Amer. Meteor. Soc., 80, 851-878.

Henderson Sellers, A., A.J. Pitman, P. Irannejad, and K. McGuffie,2002. Land surface simulations improve atmospheric modeling. EOS,Transactions, American Geophysical Union, 83, 145, 152.

Polcher, J., P. Cox, P. Dirmeyer, H. Dolman, H. Gupta, A. HendersonSellers, P. Houser, R. Koster, T. Oki, A. Pitman, and P. Viterbo, 2000.GLASS: Global Land Atmosphere System Study. GEWEX News,10(2), 3-5.

13November 2002

A Coordinated Enhanced Observing Period (CEOP)status session was held in conjunction with the EighthMeeting of the GEWEX Hydrometeorology Panel (GHP).Dr. Toshio Koike, lead scientist for CEOP, confirmedthat the build-up phase of CEOP is on schedule andthat CEOP is focusing on the development of aninitial enhanced observing period (EOP-1) data set,which covers the period July through September 2001.The implementation of two other enhanced observingperiods covering annual cycles will be undertakenusing data collected from October 2002 to December2004. In response to a key action item for the CEOPData Management Working Group, a File TransferProtocol (FTP) for delivery of data from the CEOPReference Sites to the CEOP Central Archive hasbeen established.

A number of reports were given at the meetingfrom representatives of the GEWEX Continental ScaleExperiments (CSE) that highlighted the existence ofmultinational commitments that have been coordinatedand maintained in handling interactions among theoperators of the CEOP reference sites. Channels ofcommunication between the reference site operatorsand the CEOP data managers have been establishedto ensure that data will be provided for the purpose ofimproving the collective contribution of the CSEs tothe global requirements of CEOP. As a result ofthese reports and the discussions at the meeting, alarge amount of information about the characteristicsof the CEOP reference sites has been provided bythe CSEs and placed in the CEOP Reference SiteTable at: http://www.joss.ucar.edu/ghp/ceopdm/rsite.html.The CEOP Central Archive at the University Corpo-ration for Atmospheric Research continues to receivedata for the EOP-1 from the CSEs.

As a result of an action to standardize the CEOPModel Output requirements, a document has beenproduced that provides guidance for CEOP modeloutput generation at numerical weather prediction centers,meteorological agencies and data assimilation centers.Commitments have been obtained for the provision ofCEOP model products from major national and multi-national centers including the Japan MeteorologicalAgency, National Oceanic and Atmospheric Adminis-tration, National Centers for Environmental Prediction,National Aeronautics and Space Administration/Goddard

CEOP STATUS SESSION AT GHP-8

Palisades, New York13 September 2002

Sam BenedictCEOP International Coordinator

Space Flight Center Data Assimilation Office, theEuropean Centre for Medium-Range Weather Fore-casting, the United Kingdom Met Office, the Centerfor Weather Forecasting and Climate Research of theBrazilian Agency for Space Research and the Austra-lian Bureau of Meteorology Research Center. It wasannounced at the meeting that the Max Planck Insti-tute for Meteorology (MPIM) at Hamburg, Germanywould contribute support to CEOP by assisting withthe centralized handling and retention of the CEOPmodel output data being generated by the variouscontributing centers. Work is underway to integrateCEOP data into a World Data Center on Climatedatabase scheme at MPIM. The most efficient input,storage and access structure is currently being de-fined. Mirror sites for some or all of the CEOPmodel output data products may be established inAsia and the USA.

Presentations by the CEOP Water and EnergySimulations and Prediction (WESP) Working Groupclarified the methodology that CEOP will use in ap-plying enhanced observations to better document andsimulate water and energy fluxes and reservoirs overland on diurnal to annual temporal scales and to betterpredict these on temporal scales up to seasonal forwater resource applications. The CEOP WESPWorking Group strategy, as accepted at the meeting,is to build on work by the GHP related to closingsimplified vertically integrated water and energy bud-gets with observations and analyses, and beginningefforts to simulate these budgets regionally. WESPplans to transfer this knowledge to global scales; in-clude more land, water and energy cycle processes,and begin to examine the vertical structure in theatmosphere.

The CEOP Monsoon Systems Working Group,held its first implementation planning workshop, inparallel with the GHP meeting from 10 to 11 Septem-ber 2002. Results of the workshop were reportedduring the CEOP status session and it was recon-firmed that this working group will address theimplementation of one of the main CEOP aims asso-ciated with the documenting of the seasonal march ofthe monsoon systems, assessing the monsoon systemsdriving mechanisms, and investigating the possible physicalconnections between such systems. It was recom-mended that the Working Group proceed with a CEOPInter-monsoon Model Study (CIMS) as developed duringthe Workshop. CIMS will be an international re-search project to validate and assess the capabilitiesof climate models in simulating physical processes inmonsoon regions around the world. For CIMS, a majoreffort will be devoted to defining the data require-ments, and modeling strategy for validating model physics.

November 200214

Validation data will be derived from CEOP referencesites, which include the GEWEX CSEs and plannedClimate Variability and Predictability (CLIVAR) fieldcampaign sites. Numerical experiments will be de-signed to target the simulation of fundamental physicalprocesses that are likely to uncover limitations in modelphysics. A draft report of the workshop findings withthe versions of the presentations made at the meetinghave since been put on the internet at: http://monsoon.t.u-tokyo.ac.jp/ceop/meeting/CEOP-MSS/index.html.

The CEOP Satellite Data Integration Working Groupreported that a data integration, storage and accessscheme under development by the National SpaceDevelopment Agency of Japan (NASDA) and theUniversity of Tokyo (UT) has been demonstrated asan integral part of the satellite integration process inCEOP. It was reconfirmed that this 500 tera-bytedata integration and archival system at UT will beavailable for the CEOP satellite data products work.The scheme that utilizes the NASDA/UT capabilityfor production and archiving of satellite data productsfor CEOP reference sites has been presented as athree-phased process. The new schedule shows thatthe first phase (June 2002 to November 2002) willfocus on data received from NASDA and the Univer-sity of Tokyo related to all of the CEOP ReferenceSites. Specifically this will be for the DMSP SpecialSensor Microwave/Imager and Tropical Rainfall Mea-suring Mission Microwave Imager and PrecipitationRadar data. It was announced that NASDA and theUT would host a CEOP Satellite Data IntegrationIssues Workshop from 9 to 10 October 2002 in To-kyo, Japan. The proposed agenda included a discussionof details associated with a NASDA proposal for aCEOP Committee on Earth Observation Satellites Work-ing Group on Information Systems and Services TestFacility (CEOP-WTF) that would be developed toassist with the derivation of CEOP special productsfrom each satellite sensor. The CEOP WTF pro-posal, which now includes a Satellite Data IntegrationCenter in Japan and, possibly, one in the USA, hasalready been accepted for further implementation withthe support of the Integrated Global Observing Strat-egy Partnership (IGOS-P), including Space Agencies.

The CEOP Science Steering Committee reportedthat a number of important issues related to the effi-cient organization and management of CEOP to achievethe main science objectives have been addressed bythe Committee. These actions have included finalizingthe CEOP Data Policy statement; setting minimumstandards for temporal sampling of CEOP ReferenceSite parameters, maximizing the science and technol-ogy benefits from CEOP, especially associated withsetting a goal for delivery of a CEOP seasonal data

product (EOP-1); and providing inputs on CEOP pub-lications including the CEOP Brochure (see the CEOPweb site at http://www.ceop.net). It was also con-firmed that the CEOP Advisory and OversightCommittee would be activated by the end of 2002under the co-chairmanship of Drs. A. Sumi (NASDA)and J. Kaye (NASA).

CEOP held its initial implementation planningmeeting at the Earth Observation Research Centerof NASDA in Tokyo, Japan, from 6–8 March 2002.More specifics about CEOP and the Kick-off Meetingcan be found at: http://monsoon.t.u-tokyo.ac.jp/ceop/.All of the main actions and recommendations inCEOP are being undertaken in reference to thegoals and objectives contained in the CEOP Imple-mentation Plan. The Plan, which was finalizedfollowing recommendations formulated at a CEOPImplementation Workshop held at the GSFC in March2001, was published in May 2001 and can be foundat: http://www.gewex.com/ceop/ceop_ip.pdf.

CEOP has gained the interest of other interna-tional organizations outside of the WCRP community,as evidenced by the proposal for an Integrated GlobalWater Cycle Observations (IGWCO) theme withinthe framework of the IGOS-P, which has reaffirmedCEOP as "the first element of the IGWCO." Thenext implementation planning meeting will be held inBerlin, Germany from 2–4 April, 2003. Presentationsassociated with preliminary results from the applica-tion of the available site data in the EOP-1 data setwill be part of the agenda.

GAME RESULTS AVAILABLE ON CD-ROMVersion 1.5 of the GEWEX Asian Monsoon Experiment(GAME) Intensive Observation Period (IOP) reanalysisfor the period of April – October 1998 has been releasedand is available as a CD-ROM set (6-hour interval data,gridded at 2.5 degrees).

GAME-IOP data for the 1998 summer is also availablein a CD-ROM set containing the following:

Vol. 1: JMA routine observationVol. 2: GAME-Tropics, GAME-Tibet, India (sonde)Vol. 3: GAME-HUBEX, GAME satelliteVol. 4: GAME AAN, GAME radiationVol. 5: India surface observations (for the use of the Asian scientific community)

To obtain copies, contact Mr. Kiyotoshi Takahashi atinfo-gain@mri-jma.go.jp. For more information aboutGAME data sets, see the GAME Archive and Informa-tion Network (GAIN) web site at http://gain-hub.mri-jma.go.jp/

15November 2002

NEW WEB LOCATION FORISLSCP INITIATIVE II DATA

http://islscp2.sesda.com/

Due to security problems, the original web sitefor International Satellite Land-Surface ClimatologyProject (ISLSCP) Initiative II has been closed. Datasets from the ISLSCP Initiative II data collectionare available now at islscp2.sesda.com.

18–20 November 2002—GRP/WGCM WORKSHOP ONCLIMATE FEEDBACKS, Atlanta, Georgia, USA.

18–22 November 2002—WGNE/GMPP MEETING, Météo-France, Toulouse, France.

21–23 November 2002—12TH LBA SCIENCE STEERINGMEETING, Manaus – AU, Brazil.

6–10 December 2002—AGU FALL MEETING, San Francisco,California, USA. Special theme session on research in climateand hydrology in the Southern Hemisphere.

15–17 January 2003—US-JAPAN WORKSHOP ONCLIMATE CHANGE, Irvine, California, USA.

20–24 January 2003—15TH SESSION OF THE GEWEXSCIENTIFIC STEERING GROUP, Bangkok, Thailand.

9–13 February 2003 — 83RD AMERICAN METEOR-OLOGICAL SOCIETY ANNUAL MEETING, Long Beach,California, USA.

11–13 March 2003—GEWEX WORKSHOP ON OBJECTIVEANALYSIS OF PRECIPITATION, Reading, UK

17–21 March 2003—WCRP JOINT SCIENTIFIC COMMIT-TEE MEETING, Reading, UK.

2–4 April 2003—CEOP SECOND FORMAL INTER-NATIONAL IMPLEMENTATION PLANNING MEETING,Berlin, Germany.

3–6 November 2003—14TH SESSION OF THE GEWEXRADIATION PANEL, Toronto, Canada.

GEWEX/WCRP MEETINGS CALENDARFor calendar updates, see the GEWEX Web site:

http://www.gewex.org

CEOP PLANS PRESENTED ATINTERNATIONAL FORUMS

Toshie Koike, lead scientist for the CoordinatedEnhanced Observing Period (CEOP) and Professorat the University of Tokyo, outlined the need forbetter observations of precipitation patterns causedby changes in the global water cycle at the EighthSession of the Conference of the Parties of theUnited Nations Framework Convention on Cli-mate Change held in New Delhi, India, on23 October – 1 November 2002. Prof. Koike describedhow CEOP will gather, assimilate, and archive datapertaining to the water cycle through the use ofsatellites to create coordinated observations of theglobal water cycle. He emphasized that CEOP isthe first step in meeting the challenge of providingcoordinated observations of the water cycle in aglobally integrated system.

At the World Summit on Sustainable Devel-opment (WSSD), held in Johannesburg, SouthAfrica, 16 August – 4 September 2002, the IGOSWater Cycle Theme and its first element, CEOP,

were included inthe NASDA exhibiton cooperative sat-ellite and in situobservations of theglobal and regionalwater cycles andtheir scientific con-tributions to waterresource manage-ment in theWaterdome Pavil-

ion. At a side event at the WSSD organized byNASDA and the IGOS Water Cycle Theme Team,Prof. Hartmut Grassl, chair of the GEWEX Conti-nental Scale Experiment BALTEX Science SteeringGroup, gave the keynote presentation on the "ob-servation of the intensified global water cycle" andDr. Koike presented "CEOP as the first element ofthe IGOS water cycle theme."

The Initiative II global data sets are mapped atconsistent spatial and temporal resolutions and arecompiled in four key areas: land cover, hydrometeo-rology, radiation, and soils. They span the 10-yearperiod, 1986–1995, and are mapped to consistent grids(0.5 x 0.5 degree for topography and land cover, 1 x1 degree for meteorological parameters). The tempo-ral resolution for most data sets is monthly; however,a few are at a finer resolution (e.g., 3-hourly). Cur-rently, 35 of the 47 data sets are available for distribution.The complete collection will be delivered to the GSFCDAAC in January 2003.

The final data set will consist of approximately230 parameters organized under the following broadcategories:

• Carbon (15)• Vegetation (29)• Hydrology, Topography and Soils (39)• Snow and Sea Ice (4)• Oceans (1)• Radiation and Clouds (45)

Prof. Koike at the WSSD CEOPexhibit.

November 200216

The figure illustrates the newISCCP FD data product pro-viding physically consistentglobal monthly mean net SWand net LW anomalies at thesurface, in the atmosphereand at the top-of-atmosphereover the whole time period.See article on page 7.

GEWEX NEWSPublished by the International GEWEX Project Office (IGPO)

Dr. Robert A. Schiffer, DirectorEditor: Dawn P. ErlichMail: International GEWEX Project Office1010 Wayne Avenue, Suite 450Silver Spring, MD 20910, USA

WWW Site: http://www.gewex.org

Layout: Erin McNamaraTel: (301) 565-8345Fax: (301) 565-8279E-mail: gewex@gewex.org

OVERALL INCREASE IN NET SHORTWAVE RADIATION AT TOA ANDAT THE SURFACE, BUT NOT IN THE ATMOSPHERE

(New 18-Year ISCCP Data Set)

See article on page 11.

FOLLOW-ON GLOBAL SOIL WETNESS PROJECT (GSWP-2) TIMELINE