rvin bse g S O ys th t r e a m E

THE EARTH OBSERVER May/June 2001 Vol. 13 No. 3

In this issue . . . EDITOR’S SCIENCE MEETINGS Joint Advanced Microwave Scanning Michael King Radiometer (AMSR) Science Team EOS Senior Project Scientist Meeting...... 3

Clouds and The Earth’s Radiant Energy I’m pleased to announce that NASA’s Earth Science Enterprise is sponsoring System (CERES) Science Team Meeting a problem for the Odyssey of the Mind competitions during the 2001-2002 ...... 9 school year. It is a technical problem involving an original performance about environmental preservation. The competitions will involve about The ACRIMSAT/ACRIM3 Experiment — 450,000 students from kindergarten through college worldwide, and will Extending the Precision, Long-Term Total culminate in a World Finals next May in Boulder, CO where the “best of the Solar Irradiance Climate Database . . . .14 best” will compete for top awards in the international competitions. It is estimated that we reach 1.5 to 2 million students, parents, teachers/ SCIENCE ARTICLES administrators, coaches, etc. internationally through participation in this Summary of the International Workshop on LAI Product Validation...... 18 program.

An Introduction to the Federation of Earth The Odyssey of the Mind Program fosters creative thinking and problem- Science Information Partners ...... 19 solving skills among participating students from kindergarten through col- lege. Students solve problems in a variety of areas from building mechanical Tools and Systems for EOS Data ...... 23 devices such as spring-driven vehicles to giving their own interpretation of literary classics. Through solving problems, students learn lifelong skills Status and Plans for HDF-EOS, NASA’s such as working with others as a team, evaluating ideas, making decisions, Format for EOS Standard Products . . . 25 and creating solutions while also developing self-confi dence from their experi- ences. For more information on NASA’s involvement with Odyssey of the PoDAG Meeting XVIII, National Snow and Mind, see www.odysseyofthemind.com/ and our Odyssey of the Mind web Ice Data Center ...... 28 site at earthobservatory.nasa.gov/odysseyofthemind/. EOSDIS’ EOS Clearing HOuse (ECHO) Architecture Designed to Assist Diverse After lengthy discussions, the U.S. Geological Survey has secured funding User Community ...... 29 to keep Landsat 5 operational through this September. The decommissioning of Landsat 5 had been scheduled to begin in early July due to the high cost Earth Science Education Program Update — of operating both Landsat 5 and 7 simultaneously. The continued operation Black Rock Forest Carbon Initiative . . . 33 of Landsat 5 is important as a complement to Landsat 7, providing 8-day repeat coverage during the North American growing season. Without this ANNOUNCEMENTS coverage, cloud free imagery used to determine crop vigor and projected NASA Earth Scientist Honored with AMS harvests could be scarce. Anderson Award ...... 13 On a related note, the decommissioning of Landsat 4 has been completed. A EOSDIS Training ...... 32 fi nal re-orbit thruster burn has been executed, which is expected to EOS Scientists in The News ...... 34 gradually send the spacecraft into the atmosphere within the next 8 to 10 years. Also, a draft Request For Proposals for the Landsat Data Continuity Science Calendars ...... 35 Mission was released in early August. This mission is intended to follow

The Earth Observer Information/Inquiries (Continued on page 2) ...... Back cover THE EARTH OBSERVER

Landsat 7, and extend the data record of the Landsat series of spacecraft.

A proposer conference for the Earth System Science Pathfi nder (ESSP) program was held on June 14, with approximately 75 attendees. The pro- cedure for the proposal was described along with the mechanism for funding. This will include Step 1 limited propos- als, followed by solicitation of six to eight Step 2 proposals and site reviews, then three to four missions carried into design reviews. The cost cap for the fi nal two projects will be $125 M each, not including launch services. The ESSP budget is $4.5 M for FY 2002, increasing incrementally for two years, then taper- ing off over the following few years.

Finally, I’m pleased to report that the Earth Science Enterprise Education CD-ROM was recently translated to Por- tuguese by the Instituto Nacional de Pesquisas Espaciais (INPE), Brazil. The ESE CD-ROM is a collection of satellite imagery and animations of the Earth, with an accompanying audio narration explaining the interactions between the ocean, atmosphere, ice and land surface. Mt. Etna, July 2001 It is intended as an introduction to Two volcanic plumes from Mt. Etna, each composed of different materials, are visible in new Earth system science for high school images from NASA’s Multi-angle Imaging SpectroRadiometer on the Terra satellite. A bright, and higher grade levels. Its translation plume drifting southeast over the Ionian Sea is made up primarily of volcanic ash—tiny frozen to Portuguese will greatly increase fragments of lava. A fainter plume seen near the summit contains very fi ne droplets of water and dilute sulfuric acid. The images, taken July 22, 2001, are available at http://www.jpl.nasa.gov/ the value and utility of this popular images/earth/volcano. educational resource in Brazil. The Eng- lish edition is available through the Etna is located near the eastern coast of Sicily, to the southwest of mainland Italy. Major erup- EOS Project Science Offi ce web site at tions have been issuing from both summit and fl ank vents. Fine ash falling onto the Province of eospos.gsfc.nasa.gov which is currently Catania closed the local airport, and a state of emergency was declared for the town of Nicolosi, which was threatened by lava fl ows from the southern fl anks of the volcano. averaging over 1,000 orders per day.

At the left of this image set are views from the instrument’s 70-degree forward-viewing camera, the vertical-viewing (nadir) camera and the 70-degree backward-viewing camera, concentrating on the area around the volcano. Each of these images covers an area of 143 by 88 kilometers (89 by 55 miles). The right-hand image is a full-swath view from the instrument’s 70-degree backward-viewing camera; the area imaged is approximately 400 kilometers wide (about 250 miles).

Although Etna is one of the world’s most studied volcanoes, it is diffi cult to classify, being a mixture of overlapping shield and strato volcanoes, partially destroyed by repeated caldera collapse and partially buried by younger volcanic structures. Eruptions are related to a complex tectonic situation, including subducting plates, numerous major faults intersecting at the volcano and perhaps also hot-spot volcanism.

Mt. Etna is Europe’s highest (3,315 meters, 10,876 feet) and most active volcano. In ancient Greek mythology, Etna was identifi ed with the forge of Volcan. Image courtesy of NASA/GSFC/ LaRC/JPL

2 May/June 2001 • VOL. 13 NO. 3

Tracking & Control Division and Satel- lite Communication Network Division. Joint Advanced Mr. Nasu mentioned briefl y the calibra- tion and validation plan and the new DAS (Data Analysis System) that will be Microwave Scanning implemented at the new EORC offi ce in Radiometer (AMSR) early August 2001. A. Shibata (JMA/MRI) had the diffi cult Science Team Meeting task of explaining a problem with the warm calibration target (hotload). The — E. Lobl ([email protected]), AMSR-E Science Team Coordinator, AMSR and AMSR-E warm targets are Earth Systems Science Center, University of Alabama in Huntsville manufactured from a material with a ADEOS-II AMSR homepages: thermal conductivity of 0.13 W/m/K. http://yyy.tksc.nasda.go.jp/Home/Projects/EOS-PMI/index_e.html (SSM/I’s target had an epoxy covering http://sharaku.eorc.nasda.go.jp/AMSR/index_e.html an aluminum core with a thermal con- Aqua AMSR-E homepage: http://wwwghcc.msfc.nasa.gov/AMSR.html. ductivity of 1.37 W/m/K) The plan is to move 2 PRTs (Platinum Resistance Ther- mometer) from inside the pyramids to the outside surface of the warm target, and to develop a method for calibration A Joint AMSR Science Team meeting repairs fi nished by May 23, 2001. The of the data that has two independent was held in Boulder, Colorado, March most important milestone is the thermal variables: temperature of the instrument 22-23, 2001. Topics covered were Aqua vacuum test scheduled for June 29 to and channel frequency. and ADEOS-II program status, science August 9, 2001. The plan is to ship the updates and validation plans. Two air- spacecraft to Vandenburg Western Test Dawn Conway (AMSR-E lead software borne AMSR simulators were also dis- Range on December 3, working toward engineer, UAH/MSFC) reviewed a cussed: the existing, operational PSR a launch on December 21, 2001. number of items: the current software (Polarimetric Scanning Radiometer) and status, production rules, browse images, the design concept of AESMIR (Air- S. Nasu (NASDA – Earth Observation Q/A statistics, fi le versioning borne Earth Science Microwave Imaging Research Center) presented the approaches, metadata, reprocessing, Radiometer). The day and a half meet- ADEOS-II hardware and software production histories, delivered algo- ing ended with a tour of the NSIDC status. The launch target month for rithm packages, science software update (National Snow and Ice Data Center) ADEOS-II is February 2002, provided schedule, and MOSS testing. All algo- facilities. that the H-IIA rocket is successfully rithms are delivered and used in pro- fi red prior to February. All system tests duction testing at the AMSR-E SIPS. Roy Spencer (AMSR-E Science Team have been completed in early December Several algorithms will be updated fol- Leader, MSFC) opened the meeting, 2000. A system PQR (Post Qualifi cation lowing MOSS 3 testing in May. Browse and introduced Ramesh Kakar (NASA Review) is being held March 21-23, 2001, images will be available on the TLSCF Headquarters). Kakar discussed briefl y with an initial fl ight operations team web site. For the swath products, the the proposal process for all Aqua instru- to be organized in early April 2001. browse images will be daily composites. ment teams. The MODIS and CERES The seven standard algorithms (water Other Q/A statistic summaries will also teams will have to write proposals this vapor, cloud liquid water, precipitation, be available on the web site. The TLSCF year. The decision was not made yet sea surface wind speed, sea surface software team will write and implement whether the AIRS and AMSR-E teams temperature, sea ice concentration and the metadata routines for the science will write proposals this year, (before snow equivalence) and one research algorithms. Production histories will be launch) or 18 months after launch. (soil moisture) algorithm have been suc- written operationally and will include cessfully tested at EOC (Earth Obser- such items as I/O fi le names, DAP name P. Hwang (Aqua Project Offi ce, GSFC) vation Center). In April 2001 there and version, PGE version, fi le version, presented the Aqua project status. The will be an end-to-end test involving science software error and informational three major problems in the spacecraft NASDA/Earth Observation Planning messages, and some Q/A statistics. Sci- I&T were FMU/SSR Single Bit Errors, Department (EOPD), NASDA/Earth ence software updates delivered to the TIE lockup and transponder reset. The Observation System Engineering TLSCF by COB July 15, 2001 (launch transponder reset should be completed Department (EOSD), ADEOS-II Hard- minus 6 months) will be integrated by May 4, 2001 and the FMU/SSR ware Project Team, NASDA/EOC, in the launch version operational SIPS

3 THE EARTH OBSERVER

software. The TLSCF is currently writ- face temperature (SST) and ocean wind Tb values were placed over oceans). ing a Q/A plan and science product speed. The SST retrievals show a pro- The decision was made to leave the README fi les. The AMSR-E objectives nounced diurnal warming in the mid- algorithm unchanged, and instead for MOSS 3 are: afternoon when the wind speed is below use this problem as a QC tool for 4 m/s. The wind retrievals agree very the satellite navigation. • all science algorithms integrated into well (within 0.7 m/s) with buoys and SIPS; scatterometer winds. 2) Artifi cial spike in the rain rate his- • automated processing of all standard togram. Frank Marks (Hurricane products; K. Aonashi (JMA/MRI) focused his talk Research Center) identifi ed a prob- • delivery of products with cloned on the AMSR precipitation validation lem with the AMSR-E algorithm metadata; and project in Wakasa Bay in 2001. This (as applied to the TRMM-TMI). The • delivery of sample Delivered Algo- project studies precipitation at high lati- problem was an artifi cial spike in rithm Packages (DAPs). tudes, especially the microwave bright the rain rate histograms for hur- band signature. Measurements were ricanes, at about 12 mm/hr. This Following Conway, Melinda Marquis taken with a Doppler radar (9410 MHz), problem was corrected by modify- (NSIDC) presented general information a ground-based microwave radiometer ing the a priori databases. about the plans for managing the (23.8 and 31.4 GHz) and a microwave AMSR-E data by NSIDC. The main radar (10.3 GHz). The case study pre- 3) Ongoing research related to climate points of Ms. Marquis presentation sented was during the TRMM overpass signals. It was shown that differ- were: basic information for ordering 18522 (1054 UTC February 14, 2001). ences between the freezing level data, NOSE (the spatial search/location Aonashi found signifi cant differences and the bright band, as identifi ed mechanism for AMSR-E data) imple- between the MWR Tb peaks and the sur- by the TRMM PR, could be respon- mentation, and data distribution ser- face snow intensity. There are no plans sible for some of the different vices offered by NSIDC (including for further observations before launch of signals in the climatologies of El restricted distribution of L0 data, and ADEOS-II. Niño/La Niña transition between other topics, e.g., subsetting in the the radiometer and the radar. future). The documents that NSIDC is Roy Spencer (AMSR-E Science Team responsible for are: the DIFs (metadata Leader, MSFC) discussed the discrep- 4) 20% bias between TMI and TRMM submitted to the Global Change Master ancy between passive microwave esti- PR. It was shown that the DSD Directory), guide documents (detailed mates of rainfall variations during El could be obtained from the TRMM user guides), and post-launch com- Niño, wherein the satellite estimates are PR itself, and that the retrieved munications, (documented in the OA at least twice what theory predicts for DSD was smaller than the assumed between NSIDC and AMSR-E SIPS). monthly tropical oceanic anomalies. He one. Correcting the DSD brought showed new cloud resolving model sim- the two estimates to with 10% of The rest of the day, all scientists pre- ulations that reveal a similar, very large one another. sented a short update of their research. increase in atmospheric integrated rain- water content during warmer conditions Ralph Ferraro (NOAA/NESDIS) pre- Peter Ashcroft (Remote Sensing System) that was several times the increase in sented an update on the status of presented an overview of the Level 2A model surface rainfall. Since the satellite the rain over land algorithm and the format, and the associated Level 2A passive microwave radiometers mea- Eureka, CA, NEXRAD/rain gauge vali- Quality Assessment fl ags. The Level 2A sure the integrated rainwater content, dation site effort. Recent improvements fi le is now structured as three HDF-EOS and not surface rainfall, this might be to the GPROF - land component (by J. swaths containing the low-frequency a source of bias in rainfall estimates of McCollum and R. Ferraro) include the observations, the A scan high-frequency interannual rainfall variability. development of a stratiform and con- observations, and the B scan high-fre- vective set of profi les. This work was quency scans. Quality assessment fl ags Chris Kummerow’s (Colorado State done by utilizing co-incident match-up include four bytes for every scan in University) presentation covered four TMI and PR data. Then, the actual rain every swath, as well as two bytes for aspects: rate retrieved is a weighted average of every scan and channel (for both the the stratiform and convective profi les, original Level 1A channels, and the resa- 1) Analysis of the AMSR-E production where the fractional coverage of con- mpled Level 2A channels) in whatever run. During MOSS1, the rainfall vective rain in the FOV is determined swaths are appropriate. algorithm required approximately 7 through a statistical relationship based hrs. to complete one orbit. After upon four existing convective-stratiform Frank Wentz (Remote Sensing System) thoroughly checking the algorithm separation scheme. Application to global gave a presentation on recent advances the problem was found to be a func- TMI data results in about a 20% reduc- in the microwave retrievals of sea-sur- tion of a navigation error (land-type tion in monthly rainfall (primarily in

4 May/June 2001 • VOL. 13 NO. 3

convective zones) and brings the retriev- phase regions (Wilheit et al., 1977). radar measurements are now consid- als in much closer agreement with However, to estimate hydrometeor pro- ered as the only source of this infor- other data sets, such as the GPCP fi les more accurately in the tropical pre- mation within the framework of the merged product. The TRMM ground cipitation regions, understanding of the Aqua AMSR-E mission (as listed in the validation team (D. Wolff and D. Silber- brightness temperature (Tb) variations AMSR-E Science Data Validation Plan stein) has been processing the Eureka, in the mixed-phase regions is essential. ftp://eospso.gsfc.nasa.gov/ATBD/ CA NEXRAD data, in combination with Further, establishing physical assump- REVIEW/AMSR/validation-amsr.pdf.) gauge measurements, to develop pre- tions for the microwave transfer in This presentation substantiated the liminary ground validation products the mixed-phase regions is necessary potential and the method of determi- similar to those generated for TRMM. to minimize uncertainties in rainfall nation of the HHD parameters directly A preliminary web site has been retrievals. Consequently, the objective of from spectral/polarization microwave developed for this data, trmm- the study was to quantify uncertainties radiometer measurements based on the fc.gsfc.nasa.gov/Eureka/index.html. and to achieve a solid basis for improve- recently developed Beamfi lling Algo- ment of the current rainfall retrieval, rithm [Petrenko, B.Z. “The Beamfi lling The next two presentations were given which is based on a Radiative Transfer Algorithm for retrieval of hydrometeor by two of Tom Wilheit’s students, Dong Model (RTM). To accomplish this, they profi le parameters from passive micro- Lee and Kyoung-Wook Jin of Texas examined data taken by the Convair-580 wave measurements,” IEEE Trans. on A&M. Dong Lee examined the log-nor- aircraft during the KWAJEX (Kwajalein Geosci. Remote Sensing, 39, pp. 117-124, mal assumption on the distribution of Experiment). In order to calculate radia- 2001]. HHD parameters selected for the the rain rates for estimation of monthly tive transfer, they combined radiosonde retrieval are shown to provide an effec- rain totals proposed in Wilheit et al data and aircraft microphysics data with tive description of the HHD, and at 1991. Since the log-normal assumption AMMR (Airborne Multi-channel Micro- the same time their variations are detect- was originally used for the SSM/I, it wave Radiometer) data. Analyses were able and distinguishable from radiome- is necessary to re-evaluate the assump- performed for the stratiform and con- ter measurements. The method has been tion and to determine a more effective vective rainfall regions, respectively. In tested for available hydrometeor profi le method for estimates from TMI, which stratiform rainfall regions, tests were simulation and demonstrated with the has a 10 GHz channel that the SSM/I conducted to examine the validity of example of TMI measurements process- lacks. The minimum chi-square esti- physical assumptions made to describe ing. Unlike radar measurements, the mation was used for the log-normal the abrupt change in Tb observed just proposed method makes it possible to method. To check the credibility of the below the freezing level that causes the acquire the HHD information on the estimation routines, log-normally dis- characteristic bright band on the radar. global base along the satellite orbit with- tributed synthetic data were used. Using Results indicate that the bright band out any additional instrumentation. real data from TMI, Gaussian smoothing absorption would need to be twice on the rain rates was performed to as strong to explain the observed dra- Thorsten Markus (NASA/GSFC-UMBC get all three channels, 10, 19 and 37 matic change in brightness temperature. JCET) presented sea ice activities con- GHz, to have a common resolution so In convective precipitation regions, the centrating on the validation of sea that the rain histograms were merged effective additional rain layer corre- ice concentration retrievals during the into a single histogram. The log-normal sponding to super cooled water droplets Arctic summer season. The summer method estimate averaged about 5% above the freezing layer was examined. season is the most diffi cult time for more rainfall than the direct sum It was discovered that, due to strong passive microwave sea ice retrievals method, but the log-normal assumption updrafts, approximately a 1/2 km because of rapidly changing surface may have made a false estimation. Also, supercooled layer thickness is suffi cient conditions. Landsat 7 data have been the random error involved in the TMI to describe the additional hydrometeor acquired coincident with the Melt- measurement was estimate. The result layer that is observed. pond2000 aircraft experiments. The high showed that the log-normal assumption spatial resolution of Landsat 7 (15 and contributed more error than it removed, Boris Petrenko (USRA, supporting R. 30 m) and its various spectral bands especially when the number of rain sam- Spencer at MSFC) discussed the enable the distinction between different ples was small. retrieval of the horizontal hydrometeor surface types, such as wet snow and distribution (HHD) parameters within melt ponds. Because the Meltpond2000 Kyoung-Wook Jin discussed the micro- the fi eld of view of a satellite microwave PSR data currently still have some prob- wave radiative transfer in the mixed- radiometer. The information about vari- lems with the 89 GHz vertical polariza- phase regions of tropical rainfall. Cur- ations in a small-scale horizontal hydro- tion channel, comparisons of Landsat 7 rent physically-based Radiative Transfer meteor distribution within the micro- with passive microwave data were lim- Model (RTM) algorithms for estimating wave radiometer’s fi eld of view is nec- ited to SSM/I data as the 85/89 GHz oceanic rain use a very simplifi ed hydro- essary to control “beamfi lling” errors channels are critical for NASA Team 2 meteor profi le that ignores the mixed- in rainfall retrieval. The ground–based (NT2) sea ice concentration retrievals.

5 THE EARTH OBSERVER

Adequate comparisons between these tion analysis also show the intuitively the UAV, and the types of modeling to two sensors were ensured by con- expected negative relationship between be performed. volving the Landsat 7 retrievals with the anomalies in ice concentration and the SSM/I antenna pattern. The results those in surface temperature. However, Recently Al Chang (GSFC) showed show that the overall agreement of ice the correlation values were unexpect- that SSM/I passive microwave tends to concentrations is very good (correlation edly very strong at around 0.9 in the underestimate snow cover at a global coeffi cient of 0.98). However, in areas seasonal regions in both the Arctic and scale. Chang et al. (1987), Foster et al. with a high fraction of melt ponds, the the Antarctic. These results imply that (1997) produce snow estimates that are NT2 algorithm underestimates ice con- changes in the ice concentrations are not fi nely tuned to local conditions. centrations by up to 11% for areas with coherent with changes in the surface Good examples of SSM/I snow cover a high fraction of melt ponds. This will temperatures on a pixel by pixel basis. underestimation is demonstrated by Tait be investigated in more detail once the This means that dynamic forcings on the et al. (2000) and Hall et al. (2001). The PSR data calibration is completed. The ice cover, such as those caused by wind, solution is to develop a spatially and PSR data have a spatial resolution of tides, and waves, are refl ected in both temporally dynamic snow water equiv- between 0.5 and 1 km (depending on the ice concentration and surface tem- alent (SWE) calibration scheme to pro- aircraft altitude) so that distinct surface perature maps. Since these are indepen- vide improved local estimates of SWE features can be identifi ed and analyzed. dent measurements, in a sense, one data . Using the NASDA compiled data set A signifi cant change in producing daily set validates the other, and the results that consists of 100 stations over the average sea ice concentration maps, provide strong confi dence in the derived Northern Hemisphere each with snow compared to the SSM/I processing, products. This also means that both data depth, surface temperatures and SSM/I occurs because NT2 ice concentrations sets are very useful for polynya studies data from 1992-1995 (inclusive). An are now calculated from single swath and ice-ocean-atmosphere studies in the empirical calibration scheme is gener- data which are then averaged to provide polar regions, especially in divergent ated that is temporally and spatially daily maps (to date, SSM/I daily ice areas. In the perennial ice regions of the dynamic and that can be updated with concentrations have been calculated Arctic, the correlation values are not so new observations that become available. using daily averaged brightness temper- good, even positive at times, because of [Chang, A.T.C., J.L. Foster and D.K. atures). The reason for this is that for the impact in the rise in emissivity of Hall, 1987: Nimbus-7 derived global non-linear algorithms the use of daily sea ice during onset of melt. Such results snow cover parameters, Annals of Glaci- averaged brightness temperatures does may be useful in identifying variations ology, 9, 39-44. Foster, J.L, A.T.C. Chang not necessarily produce correct daily of melt areas from one year to the other and D.K. Hall, 1997: Comparison of average ice concentrations. This is par- over the perennial ice cover. snow mass estimates from a prototype ticularly critical for the NT2 algorithm passive microwave snow algorithm, a with its explicit weather correction since Jim Maslanik (University of Colorado) revised algorithm and snow depth cli- the atmospheric signal in the brightness presented the proposed plans to vali- matology, Remote Sensing of Environment, temperatures is diluted through the date AMSR sea ice and near-ice sea 62:132-142. Hall, D.K., R.E.J. Kelly, G.A. averaging process. surface temperature products. These Riggs, A.T.C. Chang and J.L. Foster, plans include collection and analysis of (submitted) 2001, Assessment of the Joey Comiso (GSFC) presented his in-situ surface and aircraft data in the relative accuracy of hemispheric-scale study of the spatial and temporal vari- Barrow, Alaska area in conjunction with snow-cover maps. Submitted to Annals ability of sea ice concentrations derived AMSR-E Science Team aircraft fl ights. of Glaciology. Tait, A.B., D.K. Hall, J.L. from currently available passive These data will be combined with radia- Foster and R.L. Armstrong, 2000, Uti- microwave satellite data and the co-reg- tive transfer modeling to determine the lizing multiple datasets for snow-cover istered high resolution infrared satellite contribution of specifi c physical condi- mapping, Remote sensing of Environment, data derived from the Advanced Very tions to algorithm errors. Rationale for 72:111-126.] High Resolution Radiometer (AVHRR). this additional data collection and mod- Cloud-free infrared data provide tem- eling effort includes the need for coor- Don Cline (National Operational perature at the surface and may be used dinated in-situ measurements as well Hydrologic Remote Sensing Center to validate spatial variability in ice tem- as the value of acquiring atmospheric National Weather Service, NOAA) pre- peratures that will be retrieved from column data and detailed surface map- sented the plan for the Cold Land Pro- AMSR-E brightness temperature data. ping coincident with NASA aircraft cesses Field Experiment (CLPEX). This Yearly anomalies in ice concentrations and satellite overpasses. The latter is campaign will address a number of and surface temperatures were gener- to be accomplished using unpiloted science objectives pertaining to snow ated for each year from 1981 to 1999 aerial vehicles currently being tested in and frozen ground, including the vali- and comparative analysis showed very Barrow. The presentation described the dation of AMSR snow water equivalent similar patterns. Results from correla- proposed fi eld plan, the capabilities of retrieval algorithms. A major theme of

6 May/June 2001 • VOL. 13 NO. 3

the experiment is the effect of scale and ture more complex, all algorithms tend Tom Jackson discussed his plan to relating processes, measurements, and to show improved agreement with the involve both ground-based observations models between scales. CLPEX will be validation data. When applied at the and model/data assimilation results in conducted in north central Colorado, in hemispheric scale, regional algorithms the validation of the soil moisture prod- a 5-level set of nested study sites. Data developed to detect wet snow using a ucts from AMSR. Ground-based obser- to be collected include: 1) active and polarization difference approach appear vations will be provided by dedicated passive microwave remote sensing data to be accurate for only limited land sur- validation sites, available soil moisture sets from ground, aircraft, and satellite face types. Results also indicate a gen- networks and intensive fi eld campaigns. sensors; 2) intensive ground measure- eral tendency for the algorithms tested The NASA and ADEOS-II AMSR sci- ments of snow and soil characteristics, thus far to underestimate SWE. Unlike ence teams have agreed to cooperate including snow depth, water equivalent, snow extent, differences between the in data collection and analysis at surface wetness, stratigraphy, and tem- SWE validation data and the microwave an ADEOS-II dedicated validation site perature and soil moisture, temperature algorithms appear to be generally con- in Mongolia. This effort (AMPEX: and frost content; and 3) intensive sistent throughout the winter season. ADEOS-II Mongolian Plateau Experi- gamma radiation observations from The algorithm signifi cantly underesti- ment) will be led by Prof. Kaihotsu low-fl ying aircraft to measure snow mates SWE as fractional forest cover and will involve continuous recording water equivalent over 25-km x 25-km increases. Errors begin to increase as weather and soil stations with data log- grid cells for AMSR-E validation. Land fractional cover increases beyond 30 per- gers. In addition, several intensive fi eld surface modeling will be an important cent but the algorithm can reasonably sampling campaigns will be conducted component of the experiment, and will accommodate fractional forest cover so throughout the fi rst two years of AMSR be used diagnostically to provide insight long as the cover is 50 percent or less. observations. The U.S. AMSR-E science into AMSR retrieval algorithm perfor- Future work will continue the com- team will join this effort and provide a mance. prehensive multi-year comparison of at Soil Climate Analysis Network (SCAN) least four of the most promising algo- station. This will be installed by an Richard Armstrong reported on the rithms as well as the development of a USDA team (Garry Schaefer, Ron Paet- study that evaluates several different methodology that will combine visible zold, Chien-Lu Ping) in June 2001 in passive microwave snow algorithms and passive microwave data to provide Mongolia. It is expected that this effort which include both mid- and high-fre- an optimal satellite-derived snow prod- will complement the ADEOS-II effort quency channels, vertical and horizontal uct. and provide a means of cross-referenc- polarizations and polarization difference ing to other SCAN stations in the world. approaches. All frequencies and polar- Eni Njoku presented an update on The USDA in cooperation with others izations being evaluated will be avail- the soil moisture algorithm develop- has a network of 140 soil climate sta- able on the AMSR-E instrument. Eval- ment and validation activities. Results tions worldwide. Most of these are uation of snow extent is undertaken from the PSR-C airborne radiometer on data loggers and are retrieved through comparison with the NOAA (~6.9 GHz) fl own during the 1999 South- annually, while 40 of these stations Northern Hemisphere snow charts, ern Great Plains Experiment (SGP99) in are available in real time on their web- while evaluation of SWE is based on Oklahoma show good sensitivity to soil site www.wcc.nrcs.usda.gov/scan/. The snow course data from the former Soviet moisture at the 1-km resolution scale SCAN network will be a key component Union and North America. Results from for the pasture and low biomass crops of AMSR validation. Plans are progress- the current phase of the study indicate in the region. This is consistent with ing for the development of a set of dedi- that horizontal polarization-based algo- achieving the expected soil moisture cated validation sites in the U.S. built rithms, while apparently underestimat- accuracy for those vegetation conditions around several of these stations. An air- ing snow extent during shallow snow when scaled to the AMSR-E footprint. craft fi eld campaign is being planned conditions of early winter, appear to An extended experiment in Oklahoma for the summer of 2002 (SMEX02) pend- provide the best overall estimates of and Iowa (SMEX02) covering higher bio- ing the successful launch of Aqua and snow extent. Vertical polarization-based mass conditions is planned for summer ADEOS-II. algorithms provide similar results but 2002. Time-series comparisons of point with a tendency to overestimate snow in-situ soil moisture data (USDA SCAN On the second day, most of the discus- extent in the presence of desert soils site) and 25-km microwave radar data sions involved the validation plans and and/or frozen ground. Algorithms (QuikScat) were shown indicating the the airborne radiometers involved in the which include the 85 GHz channel are high temporal correlation observed campaigns. E. Lobl presented a sum- capable of detecting shallow snow but in between footprint and point data at a mary of the planned validation cam- their current form may frequently over- given site. These activities and compari- paigns. SMEX02 and the Antarctica Sea estimate snow extent. As the snow cover sons will be a key element of the post- ice experiment will take place in the becomes deeper and the layered struc- launch validation. summer of 2002. In early 2003 there

7 THE EARTH OBSERVER

are three campaigns planned: precipi- research and development, and support and Proteus). A real-time operating tation in Wakasa Bay in collaboration of missions including Aqua, GPM, system and new instrument control with the Japanese AMSR Science Team, NPOESS, etc. and science goals includ- and data acquisition software was a snow campaign in collaboration with ing soil moisture, snow and cold sur- developed in 1998 in preparation for Don Cline of NOHRSC (see description faces, precipitation, and ocean surface CAMEX-3. Infrared radiometers and of campaign above), and a sea ice exper- conditions. color video cameras are standard within iment in the Arctic. In 2004 and 2005, the each PSR scan-head, and software for plan is to repeat all campaigns as neces- Marian Klein (ETL/NOAA) described calibration of the PSR is improved sary (SMEX04 and sea ice in the Antarc- the PSR and plans to improve it. The with each mission. Synchronized opera- tic in 2004, and precipitation, snow and Polarimetric Scanning Radiometer was tion of more than one PSR scan-head sea ice in the Arctic in 2005). originally developed at the Georgia on either a NASA or Navy P-3 has Institute of Technology starting with been proposed for AMSR-E and Wind- At the March 2001 AQUA/AMSR-E a concept proposal in 1995, and fi rst Sat Coriolis underfl ight missions during Science Team meeting, Peter Hildeb- operated on the NASA P-3B aircraft in 2002. Integration of at least one PSR into rand (GSFC) reported on Goddard’s 1997 for the Labrador Sea Ocean Winds the NASA WB-57F aircraft is also ongo- project to construct an AMSR-E airborne Imaging (OWI) experiment. Since this ing in support of SSMI/S calibration and fl ight simulator. The Airborne Earth Sci- initial deployment it has been upgraded validation underfl ights. (PSR home page ences Microwave Imaging Radiometer and successfully operated by the NOAA at www1.etl.noaa.gov/radiom/psr/) (AESMIR) is a seven channel, polar- Environmental Technology Laboratory ization radiometer designed for a mul- (ETL) during several airborne cam- Roy Spencer ended the meeting with titude of science measurements. The paigns, for example the Third Con- a discussion on a few miscellaneous AESMIR system includes 6.9, 10.65, vection and Moisture Experiment topics: ideas for fi rst images, work- 18.7, 23.8, 36.5, 50.3-52.8, and 89 GHz (CAMEX-3) in August and September around data issues, and Science Team radiometer channels, all of which are 1998, Southern Great Plains Experiment proposals. Then M. Marquis conducted dual polarization and three of which (SGP99) during July 1999, and Meltpond a tour of NSIDC for those interested. will include measurement of all 4 2000 during June 2000. As a result of Stokes parameters. The AESMIR scan- these campaigns the fi rst airborne pas- Acronym list: ning system is a dual axis—azimuth sive microwave imagery of ocean sur- below elevation—scanner that provides face wind vector fi elds [Piepmeier, J.R., AMR Airborne Microwave Radiometer easily programmable, complete scan and A.J. Gasiewski, “High-Resolution AMSR fl exibility—conical, cross-track, fi xed Passive Polarimetric Microwave Map- Advanced Microwave Scanning pointing, etc.—within the full 360° azi- ping of Ocean Surface Wind Vector Radiometer for ADEOS-II muth and up to 65° incidence angle. Fields,“ IEEE Trans. Geosci. Remote Sens- AMSR-E Considerable attention is being given ing, vol.39, pp.606-622, March 2001], the Advanced Microwave Scanning Radiometer for EOS to AESMIR radiometer calibration, with fi rst multi-band conical-scanned imag- AMSS the goal of less than 1 K absolute error. ery of hurricanes and intense convec- Advanced Multi-Spectral Scanner Hot and cold loads will be mounted tion, the fi rst high-resolution C-band AVHRR close to the radiometer scan head, on imagery of soil moisture, and the fi rst Advanced Very High Resolution the azimuth turntable, but above the 65° high-resolution conical-scanned imag- Radiometer AVIRIS incidence angle limit. The radiometer ery of sea ice have been produced. Sev- Airborne Visible and Infrared Imaging can be programmed to observe these tar- eral deployments in support of AMSR-E Spectrometer gets as desired. Visible and IR channels post-launch calibration and validation DAP will be included. The system has been activities are planned for 2002-2003. Delivered Algorithm Package designed to be compatible with a wide Since its initial deployment, the PSR DIF Data Interchange Format variety of aircraft including P-3, DC-8, system has been continuously improved DSD ER-2, Proteus, etc. On the P-3 aircraft, to make the instrument more reliable, Drop Size Distribution mounting options will include both the accurate, versatile, and to provide EOSD belly and a new tail mounting option. shorter turn-around times for calibrated Earth Observation System Development AESMIR is being built in-house at God- data. There are currently two oper- GAME GEWEX Asian Monsoon Experiment dard and will be ready for fi eld opera- ational PSR scan-heads (PSR/A and GEWEX tion in the spring of 2002. Once com- PSR/C), two high-emissivity calibration Global Energy and Water Cycle pleted, AESMIR will be an in-house targets, three new positioners and two Experiment Goddard instrument in support of new scan-heads (PSR/S and PSR/L) GRASP NASA Earth Science Enterprise goals, under development, and two new air- Greenland Arctic Ocean Shelf Project GTS Goddard science and engineering craft integrations under design (WB-57F Global Telecommunications System

8 May/June 2001 • VOL. 13 NO. 3

IOP Intense Operating Period MICRORAD Microwave Radiometry and Remote Sensing of the Environment Clouds and The Earth’s MOSS Mission Operations and Science System Radiant Energy System MRI Meteorological Research Institute MWR (CERES) Science Team Microwave Radiometer NASDA National Aeronautics and Space Meeting Administration of Japan NOSE — Shashi K. Gupta ([email protected]), and Gary G. Gibson ([email protected]), NASA Langley Nominal Orbital Spatial Extent Research Center NT2 NASA Team (algorithm) 2 PSR Polarimetric Scanning Radiometer SIPS Science Investigator-led Processing System SMEX The 24th Clouds and the Earth’s Radiant be as good or better than CERES, but Soil Moisture EXperiment Energy System (CERES) Science Team NPOESS plans to leverage EOS develop- SSM/I meeting was held in Newport News, ments and is not expecting major new Special Sensor Microwave/Imager SST VA, on May 1-3, 2001. The meeting capability. A gap in radiation data still Sea Surface Temperature focused on Science Team results and the exists from the end of the Aqua mission Tb status of new data products in devel- in 2006 until the fi rst NPOESS mission Brightness Temperature opment and validation. This meeting in 2009. Wielicki led a discussion of the TIE was a major milestone as we saw value of science teams in peer reviewing Transponder Interface Electronics TLSCF the fi rst CERES Tropical Rainfall Mea- algorithms and data products. Science Team Lead Science Computing Facility suring Mission (TRMM) angular dis- teams have played critical review and TMI tribution models (ADMs), early valida- oversight roles on the Earth Radiation TRMM Microwave Imager tion results, and archival of the fi rst Budget Experiment (ERBE) and CERES TRMM year of Terra CERES ERBE-like top- projects which have signifi cantly Tropical Rainfall Measuring Mission (U.S.-Japan) of-atmosphere (TOA) fl ux data. The improved the content, quality, and use- UAV Data Assimilation Offi ce (DAO) God- fulness of the data products. The next Unmanned Airborne Vehicle dard Earth Observing System (GEOS CERES Science Team Meeting is sched- 3.3.1) meteorological fi elds are provid- uled for September 18-20, 2001 in Brus- ing improved results and, after further sels, Belgium. skin temperature tests by the cloud group, may replace the European Center Instrument Status for Medium-Range Weather Forecasts (ECMWF) data. There were interesting Larry Brumfi eld (LaRC) presented the results concerning aerosol radiative forc- instrument hardware status report. ing from TRMM, the use of long-term Terra instruments continue to operate radiative fl uxes in models, and data to without problems, and Aqua instru- evaluate climate clear-sky and all-sky ments are being readied for launch sensitivity. on schedule (currently no earlier than December 2001). The Terra Deep Space Bruce Wielicki (LaRC), CERES Co- Calibration for scan-dependent offset Principal Investigator, gave an Earth determination and verifi cation should Observing System (EOS)/CERES status take place at the end of the summer or report. The Request for Proposals for in early fall. CERES-like instruments on the National Polar Orbiting Environmental Satellite Kory Priestley (LaRC) analyzed the fi rst System (NPOESS) is expected in the year of CERES/Terra on-board calibra- fall of 2001. The new instruments will tion and consistency checks. CERES/

9 THE EARTH OBSERVER

Terra instruments are still performing fi rst year of archived Terra CERES ning a major reprocessing of all data very well, and anomalies are small. Evi- ERBE-like global and zonal mean TOA starting in mid-May. Minnis verifi ed dence of possible calibration drifts of SW and longwave (LW) fl uxes. Wong that in CERES processing we need only 0.25 to 0.5% in fi rst year for the total showed that the fi rst year of Terra global use every other MODIS pixel and scan channels is under investigation. Short- average LW fl ux was 6 Wm-2 higher line to keep cloud/radiation consistency wave (SW) and window (WN) channels than the fi rst year of combined Earth in CERES fi elds of view within 1%. show no detectable drift. The instrument Radiation Budget Satellite (ERBS)/ team will evaluate the necessity of fi xing NOAA-9 ERBE data (1985/86). About Angular Modeling an apparent 0.1% SW leak in the Flight 1.5 Wm-2 of this difference is due to Model 1 (FM-1) WN channel (an issue a NOAA-9 underestimate of daytime Norman Loeb (Hampton University, primarily for daytime deep convective LW fl uxes. The remaining 4.5 Wm-2 is HU) reported that the fi rst draft of the clouds). Spatial uniformity in the FM-2 under investigation. Zonal differences CERES TRMM angular models are com- WN channel detector’s response was vary from 1.5 to 6.5 Wm-2 with the larg- plete – a major milestone for the project. determined by conducting special lunar est differences near the poles. The SW He showed some of the early validation scanning. Mirror Attenuator Mosaic global mean Terra fl uxes cannot be eval- results from these models. The new (MAM) coatings on Terra instruments uated more accurately than about 3 to ADMs are the foundation for all of the appear to be degrading, so lamps will 5 Wm-2 until the 3-hourly geostationary advanced CERES data products. The be the only onboard source useful for data are merged with the Terra obser- fi rst draft TRMM angular models are monitoring SW channel stability. Mar- vations, and the new Terra ADMs are a major improvement on ERBE capabil- tial Haeffelin (Virginia Tech) presented available. ity. Further improvements are needed results of special scanning operations to to reduce cloud fraction dependence on obtain enhanced coverage of the upcom- Cloud and Aerosol Properties viewing angle, verify the cost/benefi t ing Chesapeake Lighthouse and Aircraft of spatially averaging CERES data to a Measurements for Satellites (CLAMS) Patrick Minnis (LaRC) summarized the more constant fi eld of view with view- experiment. progress and problems related to deter- ing angle, and complete validation of mining cloud properties. For TRMM, regional angular model biases. Data Production Status and comparisons with the Belgian Geosta- Issues tionary Earth Radiation Budget (GERB) TRMM draft ADMs were judged suffi - optical depth determinations showed ciently advanced to warrant producing a Jim Kibler (LaRC) discussed several sizeable differences in optical depth “beta” version of TRMM SSF Edition 2 Data Management System items. He with the CERES/VIRS (Visible InfraRed for all 10 months for science use. Fully reported dramatic improvement since Scanner) values for some viewing condi- validated TRMM SSF Edition 2 is sched- the beginning of the year in the delivery tions. This was attributed to approxima- uled to be available in September 2001. of CERES Level 0 data from the EOS tions in the cloud property parameter- This will be the fi rst validated matched Data and Operations System (EDOS) izations that account for surface anisot- cloud-aerosol-broadband radiation data as well as ephemeris and attitude data ropy and varying gas absorption as set. from the EOSDIS (Earth Observing a function of cloud optical depth and System Data and Information System) viewing condition. Minnis is testing an Surface/Atmosphere Fluxes Core System (ECS). He summarized improved parameterization to reduce the recent CERES code deliveries to this uncertainty by about a factor of 2. Thomas Charlock (LaRC) noted that the the Atmospheric Sciences Data Center. The cloud properties on VIRS/TRMM new beta TRMM ADMs represent Nearly half a million lines of executable were also found to be a function the fi rst chance for the Surface and code have been delivered along with of viewing zenith angle with a 10% Atmospheric Radiation Budget (SARB) an additional 140,000 lines of scripts increase from nadir to 48-degree view- team to begin serious validation of sur- and confi guration fi les. He reviewed the ing zenith. This change was correlated face fl uxes using the improved TOA CERES data products that are available with the viewing angle systematic fl ux constraint. As instantaneous ADM to both the science team and the public dependencies in the fi rst draft ADMs errors are reduced with improved and showed public CERES data product from the TOA fl ux group. The viewing ADMs, SARB consistency tuning should distribution statistics totaling 1.4 tera- zenith dependence is likely due to two become more linear and well behaved. bytes in the last 6 months. Finally, he different factors: 3-D cloud geometry, SARB tests of GEOS 3.3 water vapor requested feedback on the usability of and the ability to detect thin clouds fi elds looked comparable in accuracy to the data product documentation. easier from slant views where cloud past results using ECMWF data. optical path length increases. Terra ERBE-like TOA Fluxes The SSF estimates of surface fl uxes are The Moderate Resolution Imaging Spec- meant to tie most closely to the TOA Takmeng Wong (LaRC) summarized troradiometer (MODIS) team is plan- fl uxes, and to be least dependent on the

10 May/June 2001 • VOL. 13 NO. 3

radiative transfer theory or 4-D assimi- Plains (SGP) site showed larger differ- electronics system on November 1, 2000. lation data input. David Kratz (LaRC) ences for direct and diffuse fl uxes sepa- and Shashi Gupta (Analytical Services rately and smaller differences for global Potential algorithm improvements for and Materials, AS&M) have updated fl uxes. Edition 2 cloud processing include a 2 the Langley parameterized SW algo- CO slicing algorithm, new polar night- rithm, and results from this as well as Ellsworth Dutton (NOAA Climate time retrievals, multi-layer cloud deter- the Li/Leighton method were shown Monitoring. & Diagnostics Laboratory, mination, and improved retrievals of for clear-sky SW fl uxes. Results for Gup- CMDL) apprised the group of the status mixed phase cloud properties. Bing Lin ta’s LW surface fl ux algorithm looked of several surface-based radiation mea- (LaRC) will revisit TRMM liquid water remarkably good against Baseline Sur- surement programs. Data from these path retrievals over ocean as a possible face Radiation Network (BSRN) data. programs, e.g., the CMDL network and modifi cation for Edition 2. Tests of the Inamdar/Ramanathan clear- BSRN, are used widely by modelers for sky model and the new Zhou and Cess validation of their results. The BSRN ADM and TOA Flux Working LW cloudy sky algorithm are underway. archive has been out of service for Group about a year awaiting hardware and CERES Validation Experiment software upgrades, but is expected to be Norman Loeb led the ADM and TOA up and running within a few months. Flux Working Group meeting. He began Bill Smith, Jr. (LaRC) updated plans Fred Denn (AS&M) presented results of with a general overview of critical for the CLAMS mission. CLAMS will multi-fi lter rotating shadowband radi- ADM/inversion research issues. Yongx- focus on validating Terra data products, ometer calibrations at the Mauna Loa iang Hu (LaRC) then presented results improving SARB calculations for Observatory and at the COVE site. from CERES monthly deep convective CERES, and validating MODIS and albedo distributions for 9 months of Multi-angle Imaging Spectro-Radiome- Shi-Keng Yang (CPC/NOAA) reported TRMM and showed that the CERES ter (MISR) retrieved aerosol properties. on the status of the Stratosphere instrument has been very stable. Nitchie The intensive measurement campaign is Monitoring Ozone Blended Analysis Manalo-Smith (AS&M) showed results scheduled from July 6 to August 2, 2001. (SMOBA), which is the primary source from regional comparisons between LW of ozone data for the CERES Meteorol- and WN fl uxes from TRMM ADMs and Surface and Atmospheric ogy, Ozone and Aerosol (MOA) prod- fl uxes based on direct integration of the Radiation Budget Working Group uct. SMOBA processing was recently radiances. Meeting switched to Solar Backscatter UltraVio- let (SBUV-2) data from the NOAA-16 Konstantin Loukachine (Science Appli- Thomas Charlock (LaRC) chaired the satellite. NOAA-16 ozone values were cations International Corporation, SAIC) meeting and led the discussions. Zhong- slightly higher than both the NOAA-14 compared LW fl uxes based on pre- hai Jin (AS&M) presented results of a and Total Ozone Monitoring System Edition 2 ADMs with SSF Edition 1 1-year simulation of upward and down- (TOMS) values. fl uxes. The results show a 1-2% depen- ward SW fl uxes with a coupled ocean- dence on wind speed and precipitable atmosphere radiation model and com- Cloud Working Group Meeting water. Seiji Kato (HU) presented an pared results with measurements made improved method for computing clear- at the CERES Ocean Validation Exper- The CERES Cloud Working Group, led sky upward SW radiances over ocean iment (COVE) site. Site measurements by Patrick Minnis, established a time- surfaces with the discrete ordinate radi- varied with season; this could be related line for algorithm changes for the Edi- ative transfer (DISORT) model. Lin to the changes in sea state. tion 2 Beta and Edition 2 cloud products. Chambers (LaRC) presented the results of a study exploring the effect of angular Fred Rose (AS&M) compared CERES/ Minnis reported that the calibration bin size on direct integration. Erika SARB TOA LW fl uxes and correspond- of VIRS and MODIS is relatively Geier (LaRC) discussed the temporal ing computations from the Fu-Liou radi- well understood with some outstanding differences between the TRMM SSF and ative transfer model. The model con- issues related to the VIRS 1.6 and 3.7 TRMM ES8 cloud comparisons in the sistently overestimated LW parameters. micron channels as well as some MODIS Warm Pool and in the Tropics. Largest differences were observed for thermal channels. Larry Stowe (NOAA scenes with high clouds of moderate NESDIS) pointed out a thermal leak in Temporal Interpolation and Spa- optical depths. David Rutan (AS&M) the MODIS 1.6 micron channel that is tial Averaging (TISA) Working compared SARB-derived downward SW about one third the magnitude of the Group fl uxes with corresponding ground mea- VIRS 1.6 micron leak. Shaima Nasiri surements. Comparisons for clear-sky (University of Wisconsin) showed the David Young (LaRC) led discussions of conditions at the Atmospheric Radiation potential impact of MODIS calibration TISA issues. Studies are in progress to Measurement (ARM) Southern Great changes due to a switch to the B-side determine the sensitivity of Geostation-

11 THE EARTH OBSERVER

ary Operational Environmental Satellite the assumed aerosol model. Aerosol and the occurrence of fog and precipita- (GOES) calibration uncertainty for the optical depth (AOD) was derived using tion. The rawinsonde method also over- diurnal cycle correction and to verify broadband TOA radiances which, in estimated cloud top heights. that the constraint against CERES broad- turn, were used to compute the direct band radiances eliminates the sensitivity effect. Coakley recommended the use of Anand Inamdar (Scripps Institution of to GOES calibration. The TISA team is broadband retrieval of AOD over single- Oceanography) presented results of a currently focusing on TRMM Edition 1 channel retrievals. study of the atmospheric greenhouse development and testing. effect for the broadband LW, the 11mm Steven Dewitte (Royal Meteorological window, and the non-window regions. Invited Presentation Institute, Belgium, RMIB) presented an Comparisons of clear-sky weighted sea overview of an ongoing project on surface temperature (SST) with true Gary Rottman (University of Colorado) smooth blending of CERES and GERB SST over tropical oceans showed good briefed the team on the Solar Radiation TOA radiative fl uxes. RMIB is respon- agreement for the ERBE and TRMM and Climate Experiment (SORCE) which sible for developing the GERB inversion periods. For the Terra period, however, is a part of the NASA solar irradiance algorithms. The GERB team will use sev- the true SST was found to be about monitoring program and is scheduled eral CERES-developed algorithms and 0.5K higher than the clear-sky weighted for launch in July 2002 with an expected databases, including ADMs. value. mission lifetime of 5 years. The primary objective of SORCE is to monitor Xiquan Dong (University of Utah) com- A. J. Miller (NOAA Climate Prediction the variability of total solar irradiance pared boundary layer stratus cloud Center, CPC) presented the status of and its spectral components. This will microphysical and radiative properties long-term NOAA ozone and strato- enhance our understanding of solar pro- derived from surface observations, in- spheric temperature data sets. The drift cesses and the effects of solar variability situ airborne measurements, and sat- of the equator crossing time for NOAA on atmospheric and climate processes. ellite (GOES/MODIS) data during the satellites was addressed by adjusting all March 2000 Cloud IOP at the ARM SGP satellite data sets relative to NOAA-9. Investigator Presentation site. Good agreement with surface and The long-term ozone record is based Highlights aircraft measurements suggests that sat- on SBUV/SBUV-2 data from several ellite retrievals of cloud properties are NOAA satellites. This data set agrees Robert Cess (State University of New reliable. well with the results of 2-D photochemi- York at Stony Brook) analyzed atmo- cal models from the University of Illi- spheric LW feedbacks for clear and Leo Donner (NOAA Geophysical Fluid nois and NASA/GSFC for total column cloudy conditions and discussed what Dynamics Laboratory) presented results ozone. we could learn from a long-term of a General Circulation Model (GCM) record of ERBE-like LW observations. study of the role of the precipitation Shaima Nasiri (University of Wiscon- A 120-year run of the National Center from mesoscale convective clouds in sin) presented results of a study of for Atmospheric Research (NCAR) Cli- radiative forcing of the atmospheric cir- cloud overlap detection using MODIS mate System Model (CSM) found that culation. Cumulus cells were found to Airborne Simulator (MAS) data over the both clear-sky and cloud LW feedbacks be important heat sources, and they ARM SGP site. The algorithm, which is are slightly positive and exhibit con- became stronger in the absence of meso- based on the use of 8.5 and 11 mm tem- siderable temporal variability. Cess con- scale circulation. Donner concluded that perature differences, was able to detect cluded that a 15-year record of ERBE- mesoscale stratiform clouds were an overlap conditions. Nasiri also showed like LW observations would allow us to important heat source, a moisture sink, comparisons with ground-based mea- assess the realism of the temporal vari- and alter the mass fl ux profi les. surements from the SGP site. abilities seen in the CSM results and provide data to understand the feedback Qingyuan Han (University of Alabama David Randall (Colorado State Univer- variabilities. - Huntsville, UAH) analyzed methods sity, CSU) presented results of a study for determining cloud base heights and in which ECMWF analysis fi elds were James Coakley (Oregon State Univer- their uncertainties. The analysis was ingested into the CSU GCM. Forcing sity) presented estimates of aerosol based on rawinsonde and radar/lidar by the analysis data was used to simu- direct (radiative) effect derived using data taken at the ARM SGP site. The late weather events. In the current runs, CERES TOA radiances from Edition-1 rawinsonde technique often resulted in these were used to diagnose pressure SSF and Aerosol Robotic Network false detection of clouds, and the uncer- and moisture tendencies which, in turn, (AERONET) observations. Studies based tainties in the base height were highly can be used to force the cloud system on Indian Ocean Experiment (INDOEX) dependent on the threshold used, the model. A 2-D cloud system model has data had indicated that narrowband- non-uniformity of cloud base, the pres- now been incorporated into the global broadband relations were insensitive to ence of supercooled water in the clouds, GCM. Results from the new GCM runs

12 May/June 2001 • VOL. 13 NO. 3

showed reasonable agreement with sat- mask derived from MODIS data. ellite cloud and radiation fi elds. Bruce Wielicki (LaRC) reported on the Larry Stowe and A. Ignatov (NOAA results of a study using CERES TRMM National Environmental Satellite, Data data to examine the recent Lindzen et NASA Earth & Information Service) reviewed the al. (Bulletin of the American Meteorological development of the third generation Society, Vol. 82, No. 3, pp. 417–432) iris aerosol retrieval algorithm. Simultane- hypothesis for controlling the outgoing Scientist Honored ous retrieval of AOD in both VIRS chan- LW radiation in response to changes in nels (0.63 µm and 1.61 µm) and the surface temperature. Bing Lin and Lin With AMS Angstrom exponent was tried using Chambers found a small positive feed- a bimodal size distribution. A basic back of anvil clouds instead of a strong Anderson Award aerosol model developed by T. Nak- negative feedback. ajima provided the best results, but multi-channel retrievals underestimated Shi-Keng Yang (NOAA/CPC) pre- AOD when compared with single-chan- sented results from National Centers NASA Goddard Space Flight nel results. for Environmental Prediction (NCEP) Center research scientist, Dr. Joanne Global Data Assimilation System Taneil Uttal (NOAA Environmental (GDAS) forecasts of relative humidity Simpson, has been awarded the Technology Laboratory) reported on a (RH). The model exhibits drying in the Charles F. Anderson Award by the database of cloud properties derived Tropics and moistening at high latitudes from ground-based observations at the at all levels. Comparison of the RH American Meteorological ARM NSA site. This database is profi le averaged for the Tropics with Society (AMS). The Anderson valuable for validating satellite cloud the corresponding McClatchey profi le retrieval algorithms and GCM simula- showed the model profi le to be wetter. Award is given to an individual or tions because, in the Arctic, GCMs show Yang suggested that major convection organization in recognition of the largest disagreements, greenhouse centers in the Tropics advect moisture to effects are greatly amplifi ed, cloud- high latitudes only at the lower levels. outstanding and/or extraordinary radiation feedback and snow/ice-albedo feedback are strong, and the Arctic Education and Outreach contributions to the promotion of atmosphere is generally cloudy. In addi- educational outreach, educational tion, low contrast between clouds and The CERES Students’ Cloud Obser- service, and diversity in the AMS snow/ice surfaces, and the presence of vations On-Line (S’COOL) educational multiple layers and mixed phase clouds outreach now has over 800 schools in 55 and broader communities. Simpson confound the satellite retrieval algo- countries participating in the program. is Goddard’s Chief Scientist for rithms. A S’COOL summer teacher workshop will be held at LaRC July 16-20, 2001. Meteorology and the former Project Michel Viollier (Laboratoire de Météorologie Dynamique, LMD, France) Scientist for the Tropical Rainfall presented results of consistency checks Measuring Mission. She was the between CERES ERBE-like and Scanner for Radiation Budget (ScaRaB) results. world’s fi rst woman to obtain Comparison of CERES/Terra data with a Ph.D. in Meteorology in 1949. LMD GCM results showed good agree- ment for all months, while ScaRaB com- For more information, see: parisons showed CERES outgoing LW www.gsfc.nasa.gov/pub/ to be higher and refl ected SW to be lower than ScaRaB. PAO/Releases/2001/01-44.htm

Ron Welch (UAH) reported on the status of a database of surface-based measurements of cloud cover assembled from a variety of instruments at the ARM SGP and North Slope of Alaska (NSA) sites. This database is currently being used for validation of the cloud

13 THE EARTH OBSERVER

began in late 1978 with observations by the Earth Radiation Budget (ERB) experiment on the NOAA Nimbus 7 sat- The ACRIMSAT/ACRIM III ellite. This was followed in early 1980 by the fi rst experiment designed for Experiment — Extending and dedicated to TSI monitoring, the Active Cavity Radiometer Irradiance the Precision, Long-Term Monitor (ACRIM I) on NASA’s Solar Maximum Mission. The ACRIM mea- Total Solar Irradiance surement approach invoked a new mode of electrical and degradation cal- Climate Database ibration, providing a precision level that enabled unambiguous detection of intrinsic solar variability on timescales — Richard C. Willson ([email protected]), Center for Climate Systems Research Columbia from minutes to the sunspot cycle. University Four other satellite TSI experiments have since made contributions to the long-term database including the ERBS, UARS/ACRIM II, SOHO/VIRGO and ACRIMSAT/ACRIM III.

Introduction anthropogenic processes. To provide Past and present TSI monitoring results this understanding, a precise and/or are shown in Figure 1. The spread of Total solar irradiance (TSI) provides accurate TSI database must be estab- values refl ects the bounds of absolute the energy that, through interactions lished and maintained on centuries-long uncertainty for TSI observations by dif- with the Earth’s oceans, land masses time scales. This will require the ferent experiments, which have varied and atmosphere, determines the Earth’s deployment of many TSI monitoring from about +/- 0.3 % to +/- 0.1 % climate. A National Research Council experiments with precise overlapping between 1978 and the present. The most study recently concluded that gradual observations, periodic highly accurate recent values, in units of watts per variations in solar luminosity of as little ones or a combination of the two. The square meter, normalized to one Astro- as 0.25 % may have been the principal ACRIMSAT/ACRIM III experiment is nomical Unit, are considered the most forcing for the ‘little ice age’ that per- the most recent component of this mea- accurate due to continuing improve- sisted in varying degrees from the late surement strategy. ments in sensor calibrations. 14th to the mid-19th centuries. Paleocli- mate 14C proxy records implicate peri- TSI Monitoring and Strategy odic TSI variations as the forcings of past climate change on time scales of Monitoring TSI centuries to millennia. The sensitivity of with suffi cient climate to solar irradiance fl uctuations is precision and per- demonstrated by the subtle changes in sistence for a effective TSI caused by variations of the climate database Earth’s orbital and rotational parame- became possible ters. These so-called Milankovich cycles when a new gen- are known to be responsible for large cli- eration of mate swings, including numerous exten- electrically sive glaciations including the one that self-calibrating ended about 10 k-years ago, correspond- sensors, so-called ing to the combinations of the 100, 41 active cavity radi- and 22 k-year basic frequencies. ometers (ACR’s), and opportunities Monitoring TSI variability is clearly for extended an important component of climate space fl ight research, particularly in the context became available of understanding the relative contribu- in the 1970’s. The tions to climate change of natural and modern record Figure 1.

14 May/June 2001 • VOL. 13 NO. 3

The long-range objective of TSI monitor- 1999 into a 700 ing is to provide a database for compari- km. orbit with son with the climate record that will be a 98° inclination. capable of resolving systematic variabil- Solar pointing ity of +/- 0.1 % on time scales of a cen- by the spin-sta- tury. This is about equal to the absolute bilized ACRIM- uncertainty of the satellite experiments SAT satellite, responsible for the current database, while less accu- which operate at ambient temperature rate than pro- (~300K) and process results in the time vided by SMM domain. Clearly an ‘overlap strategy’ and UARS, is that relies on transferring measurement adequate for its precision, which is orders of magnitude mission pur- smaller than absolute uncertainty, is poses and well required to produce a useful climate within design database from these devices. limits of +/- 0.25 deg. ACRIM III The ACRIMSAT/ACRIM III can obtain more Experiment than 60 minutes Figure 2. of solar observa- as shown in Figure 2. The second part NASA accepted the ACRIM experiment tions per orbit, 10 percent more than of the second objective will be overlap- for three EOS fi ve-year mission seg- the SMM/ACRIM I and nearly twice ping comparisons between ACRIM III, ments following the initial selection that possible with the UARS/ACRIM II VIRGO and the SOURCE/TSIM exper- process in the late 1980’s. The 2nd experiment. The additional data makes iment (to be launched in mid-2002). mission segment has since been recom- ACRIM III daily mean results, its stan- Although the UARS/ACRIM II experi- peted with selection of the SOURCE/ dard data product, the most precise of ment failed abruptly in May 2001 after TSIM experiment. The 3rd has yet to all the ACRIM experiments to date. more than 9 ½ years of operation, the be defi ned. The fi rst mission segment year of overlapping comparisons with (ACRIM III) was scheduled for launch The primary objectives of the ACRIM III, coupled with reprocessing as part of the EOS/CHEM platform pay- ACRIMSAT/ACRIM III experiment are: ACRIM II results using more advanced load. EOS restructuring removed it from ACRIM III algorithms, have facilitated the CHEM satellite in the mid 1990’s and 1. The continuation of the precision a highly precise knowledge of the rela- placed ACRIM III in a ‘fl ight of oppor- TSI database during solar cycle 23 tivity of their observations. By the time tunity’ status. Since few opportunities for an EOS 5-year minimum mis- ACRIM III completes its 5-year mini- were available at that time the imple- sion. mum mission it will have several years mentation approach shifted to a ‘faster, of overlapping comparisons with the better, cheaper’ mode of a small ded- 2. Determination of its precise rela- SOURCE/TSIM observations, providing icated satellite and launch as a sec- tionship to previous and successive a high precision connection to previous ondary payload. The instrument was experiments. results and continuity of the TSI data- redesigned, updated and miniaturized base. to conform to the latest techniques and 3. Analysis of TSI variability on all technology, and maximize its adaptabil- time scales with respect to their cli- The results of the ACRIM III experiment ity to a small satellite environment. The matological and solar physics sig- demonstrate the effi cacy of its design. ACRIM III instrument embodies an opti- nifi cance. Detailed registration of virtually all mized combination of the best features signifi cant solar-induced variations can of the ACRIM I, ACRIM II and SpaceLab Satisfaction of the fi rst objective is be seen in the co-plotted results of ACRIM instruments, together with new well under way. Although it required ACRIM II, VIRGO and ACRIM III. The electronics and package design. The 4 months to stabilize and fi ne-tune downward excursions are caused by the entire cost of the instrument, dedicated ACRIMSAT’s solar pointing, the science radiative defi cit effect of sunspot area in satellite, launch and science to date is mission fi nally began without compro- solar active regions caused by their sub- less than the projected cost of the origi- mise in early April 2000. stantially lower average temperatures nally proposed ACRIM instrument for relative to the undisturbed photosphere. the EOS CHEM platform. The fi rst part of the second objective Excursions above the average are caused was accomplished through comparison by the radiative excess effects of faculae, ACRIMSAT/ACRIM III was launched of overlapping results from the ACRIM- areas slightly hotter than the average from Vandenburg AFB on December 20, II, VIRGO and ACRIM III experiments

15 THE EARTH OBSERVER

photosphere in the same solar active Table 1. Components of ACRIM Composite TSI database regions. The peaks before and after large ‘dips’ are an artifact of the location of Experiment Operational Span Data Used active region faculae on the periphery of Nimbus 7/ERB 1978–1993 1978–80, 1989-1991 the sunspot area, where they are the fi rst SMM/ACRIM I 1980-1989 1980-1989 radiative feature seen when the active UARS/ACRIM II 1991-2001 1991-2000 region rotates onto our side of the sun ACRIMSAT/ACRIM III 2000 2000 and the last seen as they rotate off the other limb of the sun. The sunspot def- which dictated the use of the following acteristic ~ 0.1 % peak-to-peak variation icit dominates during passage of the TSI data sets: Nimbus 7/ERB, ACRIM I, in solar luminosity between solar maxi- active region across the visible side of ACRIM II and ACRIM III. Some of the mum and minimum periods. The higher the solar disk. highest quality observations are avail- variation of TSI during solar activity able from the VIRGO experiment but maxima is caused by the sunspot and The ACRIM composite TSI are not used in this model because of facular components of active regions database their redundancy with ACRIM II and as discussed above. The quieter solar data ‘gaps’. minima refl ect the solar luminosity Construction of the long-term composite ‘background’ level for TSI. database required for climate change The use of Nimbus 7/ERB data is studies requires the combined results important for several reasons. The fi rst TSI trend discovered in compos- of many individual experiments. Since is that its observations precede ACRIM I ite database each experiment reports observations on and extend the composite database back the performance of their sensors (their to late 1978. The second is that it has the The discovery of climate-signifi cant, sys- so-called ‘native scale’) they must be most observations during the 2-year gap tematic long-term trends in TSI is the related to each other by normalizing to between ACRIM I & II. The third is that primary objective of monitoring. Detec- the same ‘scale’ using comparisons of it’s the best-calibrated database avail- tion of such a trend would most likely overlapping results. able for establishing the precise relation- occur by comparing successive minima ship between the results of ACRIM I & II in solar magnetic activity since active Degradation of the sensors in the space- through overlapping comparisons. The regions and their sunspot and facular fl ight environment is a signifi cant factor ERBS database also overlaps ACRIM I components are nearly absent then, as in understanding the quality of each & II but was not used because it has sub- are the chaotic variations of TSI they data set. The experiments with the most stantially fewer observations and larger cause. effective self-calibration approaches will uncertainties than Nimbus 7/ERB. The provide the best results. The composite composite database is normalized to the The composite record now allows com- database shown in Figure 3 has been scale defi ned by ACRIM- III, the most parison of two solar cycle minima. TSI constructed using these principles accurately calibrated ACRIM instru- has been integrated over 6 month peri- ment. ods centered on the reckoned minima at 1986.75 and 1996.5 years. An upward The ACRIM data trend, visually obvious in Figure 3, products are was found to have a value of 0.037 available on the %/decade. The 1-sigma uncertainty ACRIM Science quoted is computational and does Team web site: not include sources of possible system- acrim.com. Links atic measurement uncertainty, such as to the other data uncorrected instrument degradation or are provided errors in electrical self-calibration. The there as well. trend is believed to be signifi cant rel- ative to systematic uncertainties in the Several features results of the ACRIM experiments, of the 22+ year which incorporate the most comprehen- TSI record stand sive degradation and electrical calibra- out clearly. On tion schemes. Knowledge of the Nimbus the timescale of 7/ERB data quality over this period a sunspot cycle is less well understood since it lacked (~ 10 - 11 years) its own degradation calibration capabil- Figure 3. there is the char- ity. However, since sensor degradation

16 May/June 2001 • VOL. 13 NO. 3

is primarily dependent on the amount tion and transformation of active region order of magnitude accuracy improve- of solar-exposure, the much lower duty faculae to the ‘bright’ network, provid- ment would be required to calibrate a cycle of the ERB experiment would ing a defi nitive physical link between centuries-long climate database and this place the bounds for its possible deg- the evolution of active region faculae is likely unattainable with ambient tem- radation well below the approximately and the observed solar cycle TSI varia- perature sensors. 400 ppm signal between these minima. tion. The TSI monitoring paradigm would Wavelet analysis of the TSI A strong annual component is seen in change if fl ight observations could be database the wavelet spectra during the maxi- calibrated with 0.01 % absolute uncer- mum of cycle 22 (1989 – 1993). This tainty. The only known approach capa- Wavelet spectra of the TSI record have results from the semiannual orbital ble of this requires sensors operating provided a powerful analytical tool for movement of the Earth through the near liquid Helium temperatures. Peri- detecting non-stochastic structures in solar equatorial plane. The two solar odic fl ight calibrations by such cryo- solar variability. The composite TSI time hemispheres are viewed more directly genic sensors need to be deployed to series and its wavelet transform, cali- at different times of the year and any calibrate ambient temperature TSI moni- brated on a decibel scale, are shown in north-south asymmetry of TSI-varying tors. This would relieve the TSI moni- Figure 4. Each level in the plot repre- solar activity will produce an annual toring strategy of its dependency on an sents a corresponding amount of oscilla- signal. ‘overlap strategy’, protect the database tory power for a given periodicity and from catastrophic failure of a critical ele- time. Conclusions ment and provide a well understood approach for experimentally evaluating Evolutionary changes in the distribution The precision TSI database for climate the accuracy of new instruments and of spectral power on time scales up to has logged 22 ½ years of useful results. techniques. about 180 days can be seen during the The experiments providing this data- most active periods of solar cycles 21 base use ambient temperature sensors The TSI trend between the solar cycle and 22. This corresponds to the upper and process data in the time domain. minima of 1986 and 1996 is potentially limit for lifetimes of identifi able solar They are fundamentally limited to an signifi cant. However, its climate impli- activity complexes that drive short term absolute uncertainty of about 0.1 %. cations await future observations to TSI variability. The intermediate-term The ACRIM III follow-on experiment, determine whether a persistent varia- inclined patterns of high oscillatory SOURCE/TSIM, will use ambient tem- tion is present. The ~ 0.04 % per decade power (marked by large spots) show a perature sensors but process data in the trend for solar cycles 21 – 23 would not cascade of spectral power downward frequency domain. Some improvement be expected to produce detectable cli- to greater time scales during the declin- in accuracy is expected from fi ltering mate change on time scales shorter than ing phase of solar cycles. These are measurement cycle errors outside the several decades. believed to be the signatures of dissipa- basic sampling frequency. However, an The ACRIM data products are available on the ACRIM Science Team web site: acrim.com. Links to the other TSI data- bases are provided there as well. ACRIM II and ACRIM III data products are also available at the NASA Langley Research Center’s Atmospheric Sciences Data Center, eosweb.larc.nasa.gov/.

Figure 4

17 THE EARTH OBSERVER

product validation (all activities were less than 5 years old). Instead, partici- Summary of the pants endorsed a series of guidelines for selecting a plot size and location (e.g., International Workshop on with fi ne-resolution imagery) and spa- tial sampling strategy (e.g., at a mini- mum accounting for all spatial scales LAI Product Validation up to the lag of the semivariogram sill, — Jeffrey L. Privette ([email protected]) as well as including several “renegade” — Jeffrey Morisette ([email protected]) sample locations). Error budgeting, par- — Frèdéric Baret, S. Tom Gower, and Ranga B. Myneni ticularly in scaling point data up to thousands of hectares, was highly rec- ommended, although a single scaling approach (e.g., data correlation with Landsat ETM+, SPOT or IKONOS) was not endorsed.

More than 20 scientists representing results from a POLDER derived product The group outlined research priorities, 10 countries met on June 7-8, 2001 in based on inversions of a bidirectional and will collaboratively evaluate the Frascati, Italy for the fi rst workshop refl ectance model. Plans for an accuracy of the reprocessed MODIS LAI dedicated to the validation of satel- ENVISAT/MERIS product, determined product (staggered delivery will begin lite-derived Leaf Area Index (LAI) prod- via vegetation indices, and a Global in mid-2001) at test sites for which fi eld ucts. The gathering was the fi rst topical Land Imager (NASDA) LAI product data were or will be collected during workshop of the CEOS Working Group were also discussed. MODIS operations (see Table 1). For on Calibration and Validation (WGCV) these sites, different methods of pro- Subgroup on Land Product Validation Results from ongoing validation efforts ducing fi ne resolution LAI maps will (LPV). The workshop was convened conducted at sites in Canada (CCRS), be compared, and results will be cor- alongside meetings of the WGCV Ple- Europe (VALERI), U.S. (BigFoot) and related against the moderate resolution nary and Terrestrial Carbon Observa- southern Africa (NASA/SAVE), sug- LAI products. Further, the group will tion (TCO) group to encourage cross- gested that the MODIS and POLDER develop a WWW-based document out- fertilization. The LPV Subgroup was LAI products were most accurate for lining protocols endorsed at the work- chartered in 2000 to establish standard cropland, broadleaf forests, and savan- shop; a summary of this document will guidelines and protocols and to foster nas and woodlands. Both products were be prepared for peer-reviewed litera- data and information exchange relevant problematic for needleleaf forests and ture. The group agreed to meet several to validation. The Subgroup effectively wetlands, and neither had been quan- months following fi nal delivery of the operates via the topical workshops and titatively assessed for high biomass reprocessed MODIS data to evaluate resulting activities. locations. Participants agreed that the the results. Interested researchers with number of sites and results were grossly fi eld data/results from other sites are Participants were charged with: insuffi cient to determine statistically sig- strongly encouraged to participate (con- nifi cant global uncertainties. tact information provided at WWW site 1. assessing current LAI products on a given below). biome-by-biome basis, and Frèdéric Baret (INRA) led a round-table session on fi eld protocols and practices, After participating in the two-day work- 2. developing a “standards docu- which included lively discussions on shop, representatives of WGCV, TCO ment” outlining the present “best limitations and preferred operation of and the Global Observations of Forest practice” protocols for experimental widely used approaches (i.e. direct Cover (GOFC) announced strong sup- design, fi eld data collection, analy- versus indirect) and optical instruments port for LPV’s continuing activities. The sis and LAI product evaluation. (e.g., LICOR Plant Canopy Analyzer, next topical workshops co-sponsored by 3rd Wave Engineering TRAC, hemi- the Subgroup will be dedicated to fi re/ Although the Moderate Resolution spherical photography, and Decagon burn scar product validation (July 6-8, Imaging Spectroradiometer (MODIS) ceptometer). The session evolved into 2001, Lisbon, Portugal; Point-of-contact: LAI Product (Principal Investigator: R. an equally lively discussion on sampling José Pereira), and land cover/land use Myneni, Boston U.), is the fi rst opera- strategy. Consensus on optimal, gener- tional (daily and 8-day) global product, ally applicable approaches was diffi cult Marc Leroy (CESBIO/CNES) presented to achieve given the early state of land (Continued on page 22)

18 May/June 2001 • VOL. 13 NO. 3

Each goal is elaborated on below.

Increasing the quality and usage of Earth An Introduction to the science data. Images and information from Earth orbiting satellites have dem- Federation of Earth Science onstrated in compelling ways that our planet is fi nite with only a thin shell to Information Partners support life as we know it. Today, man- — Alan Ward ([email protected]), EOS Project Science Offi ce, NASA Goddard kind is affecting this life–support system Space Flight Center, SSAI to an extent that might prove irrep- arable. We also have unprecedented amounts of data available, which, if easily accessible, can be used to study the Earth and diagnose its condition more accurately than ever before. History A new Constitution and Bylaws were agreed upon in January of 2000. The suc- The Federation of ESIPs brings together NASA established the Federation of cess of the Federation experiment had scientists and students, engineers and Earth Science Information Partners begun to attract the interest of other farmers, government agencies and busi- (ESIP) in 1998 as an experiment in institutions and the new governance nesses to achieve its goals. By engaging developing a federated system for data structure provided a means by which the full range of stakeholders, the Fed- management. The notion was that this the Federation could add new partners. eration hopes to derive maximum ben- should be a decentralized, heteroge- In July 2000, the fi rst two new partners efi ts from data just becoming available neous and distributed data distribution joined the Federation. As of Summer from the powerful new generation of system - something different from the 2001, a total of eight new partners have Earth observing satellites. By combining existing homogeneous and centralized been added and others are considering efforts, the pace of Earth science accel- Earth Observing System Data and Infor- joining. The Federation experiment has erates and society’s understanding of mation System (EOSDIS) that is based succeeded; the Federation continues to the Earth System is improved. By on the EOSDIS Core System. One of the grow and should soon become a legal including groups that reach beyond the key aspects of the experiment was to entity, which is seen as a key step in the science communities, it becomes easier work on governance and interoperabil- evolution of the organization. for the best science to fi nd effective use ity issues. The hope was that the “Work- throughout society. The idea here is that ing Prototype” Federation would grow Goals of The Federation “the whole is greater than the sum of the beyond the initial NASA-sponsored parts.” That is to say, members can do experiment and come to involve institu- The Federation, in January of 2000, far more to benefi t society by working tions sponsored by other agencies as adopted the following as the preamble together as Federated partners than they well. of their Constitution. It is a concise state- could ever do apart and working inde- ment of their mission. pendently. Partners pool their resources The initial Federation partners each and put them to work to solve real entered into individual Cooperative “We believe that society’s quality of life, eco- world problems. Agreements with NASA. The Coopera- nomic opportunities, and stewardship of the tive Agreements required the ESIPs to planet are enhanced by regular use of scien- Creating interoperability tools for Earth participate in forming the Federation tifi cally accurate Earth science information data resources. NASA’s Terra and other and to work together to develop provided in a timely matter by a Federation Earth observing satellites have sophisti- processes for governance and system- of groups collaborating to improve their col- cated sensors to record Earth radiation wide interoperability. The task was lective services” patterns, surface land and sea tem- challenging, but some initial Rules of peratures, vegetation’s vitality, biomass, Governance were developed which Based on this mission, the goals of the water availability and clarity, ocean allowed the Federation to function until ESIP Federation are twofold: winds and currents, plus air pollution. a more permanent system of governance The Federation helps make available could be developed. Shortly after its 1. to increase the quality and usage of these data for science and community inception, NASA’s eight Distributed Earth science data, and empowerment. But the Federation real- Active Archive Centers and the Global izes that all the science data in the world Hydrology Resource Center (Huntsville, 2. to create interoperability tools for is useless if it cannot be accessed in a AL) were added to the Federation. Earth data resources. timely fashion and with minimal effort.

19 THE EARTH OBSERVER

The quality and usage of Earth Science • Public health offi cials who can pre- that sponsor Federation activity. The dif- data will only increase if it is practical dict outbreaks of disease and pesti- ference between these two membership for end-users to access and process this lence. categories is that a Category IV ESIP information. may vote in the Federation assembly • Farmers and ranchers, who can while a Category V ESIP elects not The Federation has developed interop- improve the productivity, quality to be a voting member. Both types of erability tools. They have developed and consistency of their crops sponsoring agencies offer major fi nan- a System Wide Interoperability Layer, and/or rangeland. cial and/or in-kind support of Federa- which is a common interface that all tion activities. Presently, NASA is the partners participate in (see Website: Types of ESIPs sole sponsor of the Federation (a Cat- esipfed.net/committees/interop/ egory IV ESIP) but this should change as swil.html). Not only does this promote In accordance with its Bylaws, the Fed- the Federation evolves. easy exchange of information among eration has established a number of partners but by creating common data different partnership categories. These List of Current ESIPs structures and processing capabilities, are not meant to be rigid divisions, those with data can be more easily con- and picking the appropriate category for Category I ESIPs nected with those who can make use of some institutions is diffi cult. However, this data. These tools also make it easier having these categories allows for the Alaska SAR Facility (ASF DAAC) special- izing in synthetic aperture radar, Fairbanks, for end-users to make use of the data Federation to strive to have balanced AK. without having to navigate through a representation among its participants. tel. 907.474.6116 maze of data and cataloging issues. Currently, fi ve categories have been e-mail: [email protected] established and they are described www.asf.alaska.edu Benefi ts below. The EROS Data Center (EDC DAAC) spe- cializing in land processes, Sioux Falls, SD. The information coming from space- Category I ESIPs. These are institutions tel. 605.594.6116 based imagery and related Earth Science whose primary focus is disciplined e-mail: [email protected] data products and services are powerful adherence to operational schedules in edcdaac.usgs.gov tools for many communities of users. data processing, archiving and distribu- The Global Hydrology Resource Center These include: tion. Currently all of NASA’s Distrib- (GHRC) specializing in global hydrologic uted Active Archive Centers (DAACs) cycle and climate, Huntsville, AL. • Scientists, who can better under- are participating in the Federation as tel. 256.961.7932 stand basic biological, geophysical Category I ESIPs, as are the Global e-mail: [email protected] and chemical cycles of the Earth. Hydrology Resource Center (Huntsville, ghrc.msfc.nasa.gov AL), and NOAA’s National Climatic • Business people who create value The Goddard Space Flight Center (GSFC Data Center. DAAC) specializing in upper atmosphere, in imagery and support further mis- ocean color, and meteorology, Greenbelt, sions. Category II ESIPs. These are institutions MD. whose primary focus is on Earth Science tel. 301.614.5224 • Teachers who use data and infor- e-mail: [email protected] research. These include universities and mation to enhance education. daac.gsfc.nasa.gov/ groups from a number of NASA spon- • Professors who are educating future sored projects including several of the The JPL Physical Oceanography DAAC (JPL PO.DAAC) specializing in physical oceanog- scientists, policymakers, etc. Regional Earth Science Application Cen- ters (RESAC). raphy and air-sea interaction, Pasadena, CA. tel. 626.744.5508 • Civil servants who need to assess e-mail: [email protected] and plan for sustainable resources. Category III ESIPs. These are institutions podaac.jpl.nasa.gov whose primary focus is on developing • Environmental decision makers, applications for Earth Science data. Part- Langley Research Center (LaRC DAAC) spe- including regulators, legislators, ners here range from non-profi t educa- cializing in radiation budget and upper atmo- sphere, attorneys, and land use planners. tional organizations to new e-businesses Hampton, VA. offering information through the inter- tel. 757.864.8656 • Emergency management authori- net. Several NASA RESACs have joined e-mail: [email protected] ties, who can make critical deci- as Category III ESIPs. eosweb.larc.nasa.gov sions, e. g. regarding evacuations. National Snow and Ice Data Center (NSIDC Category IV and V ESIPs. These two • Ship captains who can safely plot DAAC) specializing in snow and ice, Categories are reserved for institutions University of Colorado, Boulder, CO. their course around storms.

20 May/June 2001 • VOL. 13 NO. 3

tel. 303.492.6199 eos-webster.sr.unh.edu/ tel. 785.864.7369 e-mail: [email protected] Annette Schloss tel. 603.862.0348 e-mail: [email protected] www-nsidc.colorado.edu e-mail: [email protected] www.kars.ukans.edu/resac/ resac.htm

Oak Ridge National Laboratory (ORNL ESP2Net: Earth Science Partners’ Private Net- Numerical Terradynamic Simulation Group, DAAC) specializing in biogeochemical work, contact Silvia Nittel, University of contact Steve Running, University of dynamics, Oak Ridge, TN. California, Los Angeles. Montana, Missoula, MT. tel. 865.241.3952 tel. 310.825.0607 tel. 406.243.6311 e-mail: [email protected] e-mail: [email protected] e-mail: [email protected] www.daac.ornl.gov dml.cs.ucla.edu/projects/dml_ esip www.forestry.umt.edu/ntsg/

Socioeconomic Data and Applications Evolution of Snow Pack in the Southwestern Category III ESIPs Center (SEDAC DAAC) specializing in United States: Spatial and Temporal Variabil- socioeconomic applications, Columbia Uni- ity from a Remotely Sensed and In Situ Data Institutionalizing MTPE Data for Land and versity, Palisades, NY. Set, contact James Simpson, Scripps Insti- Environmental Management, contact tel. 914.365.8920 tute of Oceanography, University of Califor- Thomas Burk, University of Minnesota, St. e-mail: [email protected] nia, San Diego. Paul. sedac.ciesin.org tel. 858.534.2789 tel. 612.624.6741 e-mail: [email protected] e-mail: [email protected] National Climatic Data Center (NCDC), landlub.ucsd.edu/projects/esip /esip.html terrasip.gis.umn.edu/ world’s largest active archive of weather data, Asheville, NC. Tropical Rain Forest Information Center, con- California Land Science Information Part- tel. 828.271.4384 tact Dave Skole, Michigan State University, nership, contact Gary Darling, California e-mail: [email protected] East Lansing. Resource Agency, Sacramento. www.ncdc.noaa.gov tel. 517.432.7774 tel. 916.653.4279 e-mail: [email protected] e-mail: [email protected] Category II ESIPs www.bsrsi.msu.edu/trfi c ceres.ca.gov/calsip

The Distributed Oceanographic Data System, The Passive Microwave ESIP, contact The Earth Data Analysis Center, contact Stan contact User Services, University of Rhode Michael Goodman, Global Hydrology & Morain, University of New Mexico, Island, Narragansett. Climate Center, Huntsville, AL. Albuquerque. tel. 401.874.6283 tel. 256.961.7890 tel. 505.277.3622 x228 e-mail: [email protected] e-mail:[email protected] e-mail: [email protected] www.unidata.ucar.edu/packages/dods/ pm-esip.msfc.nasa.gov edacesip.unm.edu

The Earth System Science Workbench, con- The Global Landcover Facility, contact John Integrating Environmental and Legal Infor- tact James Frew, University of California, Townshend, University of Maryland, mation Systems, contact Konstantinos Santa Barbara. College Park. Kalpakis, University Space Research tel. 805.893.7356 tel. 301.405.4558 Association (USRA), Greenbelt, MD. e-mail: [email protected] e-mail: [email protected] tel. 410.455.3143 essw.bren.ucsb.edu/ glcf.umiacs.umd.edu e-mail: [email protected] www.csee.umbc.edu/~elis Seasonal to Interannual Earth Science Infor- GPS Environmental and Earth Science Infor- mation Partner, George Mason University of mation System: GENESIS, contact Thomas A Public Access Resource Center (PARC) Fairfax, VA. Yunck, Jet Propulsion Laboratory, Pasadena, Empowering the General Public to Use Menas Kafatos tel. 703.993.1997 CA. EOSDIS Phase III Operations, contact George e-mail: [email protected] tel. 818.354.3369 Seielstad, Upper Midwest Aerospace www.siesip.gmu.edu e-mail: [email protected] Consortium, University of North Dakota, Hank Wolf tel. 703.993.3637 www-genesis.jpl.nasa.gov Grand Forks. e-mail: [email protected] tel. 701.777.4755 The Ocean ESIP, contact Victor Zlotnicki, Jet e-mail: [email protected] Progressive Mining of Remotely Sensed Data Propulsion Laboratory, Pasadena, CA. www.umac.org/EOSDIS/ for Environmental and Public Health Appli- tel. 818.354.5519 cations, contact Howard Burrows, IBM, e-mail: vz@pacifi c.jpl.nasa.gov WeatheRoute, contact Kevin Meagher, Read- Yorktown Heights, NY. oceanesip.jpl.nasa.gov ing Information Technology, Inc., Reading, tel. 202.421.1049 MA. e-mail: [email protected] California Water Resources Research and tel. 781.942.1655 x15 www.research.ibm.com/ Applications Center, contact Norman Miller, e-mail: [email protected] networked_data_systems/esip/ University of California, Berkeley. www.riti.com tel. 510.495.2374 A Web-Based System for Terrestrial Environ- e-mail: [email protected] Earth Data Discovery Consortium, contact mental Research, University of New Hamp- www-esd.lbl.gov/RCC Bruce Caron, Planet Earth Science, Inc., shire, Durham. Santa Barbara, CA. Berrien Moore III tel. 603.862.1766 Great Plains Regional Earth Science Applica- tel. 808.569.1343 e-mail: [email protected] tions Center, contact Theresa Crooks, Uni- e-mail: [email protected] versity of Kansas, Lawrence, KS. www.planearthsci.com

21 THE EARTH OBSERVER

Bay Area Shared Information Consortium MTPE-Derived Data Products for the Fisher- tel. 562.985.2358 (BASIC), contact Dave Etter, San Jose, CA. ies, contact Patrick Simpson, Scientifi c Fish- e-mail: [email protected] tel. 408.345.1573 ery Systems, Inc., Anchorage, AK. wildfi re.geog.csulb.edu/resac/main/ e-mail: [email protected] tel. 907.563.3474 netresacmain.htm www.basic.org e-mail: pat@scifi sh.com www.scifi sh.com TERC, contact Tamara Ledley, Museums Teaching Planet Earth, contact Cambridge, MA Patricia Reiff, Rice University, Houston, TX. Northeast Regional Earth Science Appli- tel. 617.547.0430 tel. 713.348.4634 cations Center, contact Chester Arnold, e-mail: [email protected] e-mail: [email protected] Haddam, CT. www.terc.edu earth.rice.edu tel. 4511.4511.4511 e-mail: [email protected] Category IV ESIP Terrain Intelligence Products from EOS resac.uconn.edu/ Sensor Data, contact Doug Kliman, Veridian, NASA Headquarters - Code YS, led by Tucson, AZ. Mid-Atlantic Regional Earth Science Applica- Martha Maiden, Offi ce of Earth Sciences, tel. 520.326.7005 tions Center, contact Stephen Prince, Univer- Washington, DC. e-mail: [email protected] sity of Maryland, College Park. tel. 202.358.1078 www.terraindata.com tel. 3408.3408.3408 e-mail: [email protected] e-mail: [email protected] Stormcenter.com, contact Dave Jones, Wash- www.geog.umd.edu/resac ington, DC. tel. 410.647.4299 Southern California Wildfi re Hazard Center, e-mail: [email protected] contact Christopher Lee, California St. Uni- www.stormcenter.com versity, Long Beach.

(Continued from page 18) Summary of the International Workshop on LAI Product Validation product validation (January, 2002, Ispra, For more information on the Land gov/MODIS/LAND/VAL/ Italy, cooperatively with a Global Land Product Validation Subgroup and the CEOS_WGCV/ Cover 2000 meeting; point of contact: LAI Workshop, see: modarch.gsfc.nasa. Alan Belward).

Site Country Major Biome Latitude Longitude Sponsor

Alpilles France Cropland 43° 46‘ 59‘’ N 4° 15‘ 0‘’ E VALERI Bondville IN, USA Cropland 40° 00‘ 23“ N 88° 17‘ 27“ W BigFoot BOREAS NSA Canada Conif. Forest 55° 52‘ 46“ N 98° 28‘ 50“ W BigFoot/CCRS BOREAS SSA Canada Conif. Forest 53° 39‘ 21“ N 105° 19‘ 23“ W CCRS Brasschaat (De Inslag) Belgium Mixed Forest 51° 18‘ 00“ N 4° 31‘ 00“ E VITO Forêt de Nezer France Pine forest 44° 33‘ 58‘’ N 1° 1‘ 59‘’ W VALERI Harvard Forest MA, USA Decid. Forest 42° 32‘ 17“ N 72° 10‘ 17“ W BigFoot/BU Jrvselja Estonia Boreal Forest 58° 16‘ 00“ N 27° 18‘ 00“ E VALERI Kejimkujik Park, Nova Scotia Canada Decid. Forest 44° 30‘ 00“ N 65° 30‘ 00“ W BU/CCRS Konza KS, USA Grassland 39° 04‘ 56“ N 96° 33‘ 36“ W BigFoot Krasnoyarsk Russia Boreal Forest 57° 17‘ 59“ N 91° 35‘ 59“ E NASA/Sukachev Inst. Mali Mali Shrubland 15° 20‘ 00“ N 1° 32‘ 00“ W VALERI Maun Botswana Woodland 19° 55‘ 22“ S 23° 35‘ 40“ E BU/SAVE Mongu Zambia Woodland 15° 26‘ 16“ S 23° 15‘ 10“ E SAVE Okwa River Crossing Botswana Shrubland 22° 24‘ 33“ S 21° 42‘ 47“ E BU/SAVE Pandamentanga Botswana Woodland 18° 39‘ 19“ S 25° 30‘ 01“ E BU/SAVE Park Falls WI, USA Decid. Forest 45° 56‘ 45“ N 90° 16‘ 20“ W BigFoot Romilly France Cropland 48° 26‘ 00“ N 3° 48‘ 00“ E VALERI Ruokolahti Finland Conif. Forest 61° 31‘ 58“ N 28° 42‘ 62“ E BU Skukuza S. Africa Savanna 25° 01‘ 12“ S 31° 29‘ 48“ E SAVE Tshane Botswana Savanna 24° 09‘ 51“ S 21° 53‘ 34“ E BU/SAVE Watson Lake, Yukon Canada Conif. Forest 60° 5‘ 24‘’ N 129° 23‘ 1‘’ W CCRS

22 May/June 2001 • VOL. 13 NO. 3

developers to discuss in an open forum how such tools fi t into the end-to-end data analysis cycle, user needs, and Tools and Systems for EOS issues. Discussion also included ideas on future plans developers had for their Data software. It was clear that while the — Jeanne Behnke ([email protected]), and developers understand the need for a — Richard Ullman ([email protected]), ESDIS 23Project diverse set of tools, the users fi nd this diversity of tools confusing. Based on discussions and follow-up by the AGU attendees, one opinion given was that there is an indication “that a lot more people are interested in tools specifi c to an instrument than are interested in a general-purpose tool.” The develop- At the recent Spring Meeting of the In particular, poster papers addressed ers all plan to continue discussion on American Geophysical Union, a special progress on and issues with analysis enhancing and consolidating tools for poster session, titled Tools and Systems tools, data population tools, specifi c the benefi t of users based on infor- for EOS Data, was convened by the EOSDIS data sets and metadata types, mation and impressions gained at the Earth Science Data Information Systems tools for metadata creation and man- AGU session. Such discussions will (ESDIS) Project. The poster session, held agement, tools for distribution, EOSDIS also be central to discussions at the May 29-30, covered the use of tools to data formats and distribution tech- upcoming HDF-EOS Workshop, Sep- search, order, visualize and manipulate niques. The session was well attended tember 19-21 in Champaign, IL. See data collected from the various Earth over the two-day period, not only by hdfeos.gsfc.nasa.gov for details. Observing System (EOS) missions. The Earth scientists but also by scientists in purpose of the session was to inform other related fi elds (e.g. solar physics) The table on Page 24 lists the posters EOS Science data producers, data users, who were interested in tools being that were presented at the AGU session. planners and managers of available data developed for EOS. The interaction Abstracts for the posters can be found systems and tools for managing EOS between the authors (tool developers) on the AGU web page at www.agu.org/ data. These systems and tools help users and the AGU attendees involved discus- meetings/sm01top.html. Please contact to process, archive and access data and sions of tools, formats, data access, basic Jeanne Behnke (jeanne.behnke@gsfc. information for research, applications, questions on the data, and future direc- nasa.gov) or Richard Ullman planning and management. tions. As noted by one author, the ses- (richard.ullman@ gsfc.nasa.gov) with sion gave him the time to “directly any questions or comments about this Thirty-two poster papers were pre- talk and interact with some of the real article. sented at this session. The papers cov- users of his software tool for their ered such categories as: research and routine applications.” Sev- eral attendees noted that they were • Systems for Finding and Getting saved a lot of “hunting for the right EOS Data tools” by virtue of their interactions at this session with several developers. • Tools Designed for EOS Data in General This session represented a good oppor- tunity for the developers to meet with • Tools Designed for Specifi c EOS scientists who are now in the position Instrument Data to really begin looking at Terra data. Fol- lowing the session, authors have noted • Standards Supporting EOS Data additional hits on tool web pages, phone both current and under develop- calls, and emails referring to the dis- ment cussions started at the AGU meeting. An unexpected outcome of the session PCs were available at several of the was that it also afforded the opportunity posters to provide hands on demon- for tool developers to see what is being strations of the tools and to support developed in the broad community. The impromptu queries and data analysis. unstructured time in the session allowed

23 THE EARTH OBSERVER

Category Title Authors

Introduction/Overview EOSDIS Science Data Systems for 2001 J Behnke, R Ullman

General Tools The Earth Observing System Data Gateway M S Nestler, R Pfi ster A New Guide System for the EOSDIS J Yang, P Agbu, J Tyler, A Warnock, R Ullman Data and Information Access Link (DIAL) R Suresh, L Di, K McDonald EDG Valids Processing (EVP): A web-based metadata ingest and test tool Z Yin, J Yang, R Ullman, F Corprew Innovative Technologies for Science Data Processing R Ramachandran, H T Conover, S J Graves, K Keiser, M R Smith Demonstration of the Application and Visualization of Environmental M P Soracco Satellite Remote Sensing Data in a Geographic Information System

Data Set Accessibility Making MISR Images Using Utilities Available From The ASDC J O Olson Web-based Hierarchical Ordering Mechanism (WHOM) tool for M S Sikder, P Eaton, G Leptoukh, MODIS data from Terra N McCrimmon, B Zhou Functionality of the Langley TRMM & Terra Information System (LATIS) S Sorlie, J O Olson EOS Snow and Ice Products at NSIDC DAAC A E Machado, B McLean, G Scharfen, M M Holm TERRA/MODIS Data Products and Data Management at the GES-DAAC A K Sharma, S Ahmad, P Eaton, J Koziana, G Leptoukh, D Ouzounov, A Savtchenko, G Serafi no, M Sikder, B Zhou MODIS Ocean Primary Productivity: Data Products and Services K R Turpie, W E Esaias

HDF-EOS Specializations HDF-EOS Tools and Information Center Web Site V Sagar, G W Schwenke, R E Ullman Interactive Visualizer and Image Classifi er for Satellites (IVICS) T A Berendes, R M Welch, U S Nair, D A Berendes HDF-EOS Dump Tools U Prasad, A Rahabi HDF-EOS in MATLAB(R) R P Comer, C Lawton, M M Yale Browse and Utility Tools for HDF-EOS Formatted Files F Milburn, L Klein Visualization of HDF/HDF-EOS Format Earth Observing System Data J M Torson Using the ISIS “cv” Program Vector Data Model: A New Model of HDF-EOS to Support GIS E Chi, R d Edmonds Jr. Applications in EOS NWGISS and GIS Translators: The Generic Tools for Making HDF-EOS L Di, W Yang, M Deng, K McDonald Data Accessible to GIS The Polar HDF-EOS Data Imaging and Subsetting Tool S S Khalsa, R L Weaver

MODIS Tools The MODIS Swath-to-Grid Toolbox T M Haran, S S Khalsa, K Knowles, L E Gumley Simple Mapping Tools from the Goddard DAAC Earth Sciences MODIS A K Savtchenko Data Support The MODIS Reprojection Tool J Dwyer, J Weiss, G Schmidt, T Logar, R Burrel, G Stubbendieck, J Rishea, B Misterek, S Jia, K Heuser Web-Based System for Visualization and Quality Assurance of R L Vogel, M Necsoiu, K R Turpie, MODIS Ocean Products W E Esaias Web-Based Tools for Checking EOS Data Integrity K Tewari, A K Sharma, L Fenichel, S Kreisler, G Leptoukh, C Lynnes, G Roth, R Strub Online Analysis of BUFR-formatted Quality-controlled NCEP C S Phelps, J Qin, M S Hegde, D R Final Observation Data at the Goddard Earth Sciences DISC/DAAC Frank

CERES Tools view_hdf: A tool to visualize CERES data products N R Silvers, L Hunt

MISR Tools MISR Data Visualization and Analysis Using the hdfscan Tool K A Crean, D J Diner, P K Banerjee Tools for Visualizing MISR data N A Ritchey MISR Data Access and Visualization Using misr_view J R Hall, C K Thompson

24 May/June 2001 • VOL. 13 NO. 3

about potential termination of support for HDF4. However, NASA will not terminate support for HDF4 as long Status and Plans for as the format is needed. And, NASA, ECS and NCSA are working together HDF-EOS, NASA’s to develop tools to help in data tran- sition. HDF5, the new standard, is dif- Format for EOS ferent from HDF4, the format for EOS standard products adopted for TRMM, Standard Products Terra, ACRIM, SAGE III and Aqua. The new standard is not backward compat- — Richard Ullman ([email protected]), ESDIS Information Architect, NASA Goddard ible either in code or in the underlying Space Flight Center, Greenbelt, MD. conceptual model with the old. This is not good news for long-term science endeavors. Changes in computing tech- Introduction to scientifi c data. Furthermore, EOS nologies pose a real and serious chal- teams have found HDF to be actively lenge to maintenance of long series data In the early 1990s, NASA’s Earth Science and effectively supported by NCSA, a collections. Data Information Systems (ESDIS) Proj- national leader in the advancement of ect began to address the technological applications computing. Why HDF5? challenges involved in producing, dis- tributing, analyzing and archiving the To further facilitate data sharing, certain As science computing systems evolved, Earth Observing System (EOS) Standard “idioms” with respect to geo-referenc- it became clear to NCSA’s HDF group Products. Since data interuse and inter- ing, data organization, and metadata that HDF4 would have diffi culty evolv- disciplinary science investigation are storage are encouraged. The EOS stan- ing to meet the demands of these sys- central to Earth systems science in gen- dard use of HDF for satellite swath data, tems. The future of Earth observing sys- eral, and particularly to the goals of gridded data, and point data are imple- tems is likely to include parallel pro- EOS, NASA found that a standard mented by HDF-EOS, developed by cessing environments, very large data data format would best facilitate data NASA under the EOSDIS Core System sets, data spanning multiple computing exchange and interoperability. Further- (ECS) contract. HDF-EOS is an API environments, new data models, and more, the existence of a common format and library of routines that invoke complex data analysis and visualization would encourage the development of HDF to create standard groups of HDF capabilities requiring industry standard tools for analysis that could be applied objects that form HDF-EOS idioms. interfaces. But, HDF4 supports only across the spectrum of data sets. Within HDF-EOS are ‘structural meta- datasets smaller than 2 gigabytes, with data’ that provide a common mecha- fewer than 20,000 datasets in any one When it was determined that a common nism for attaching geo-referencing infor- fi le, and is not capable of effi ciently per- format could provide important benefi ts mation to science data. In addition, forming I/O in parallel computing envi- to EOS, ESDIS engaged a number of a software library, called the ECS Sci- ronments. Size and complexity are an DAACs and science teams in examining ence Data Processing Toolkit is provided issue. The HDF4 library consists of over and testing their data products using a to implement a standard for attaching 300,000 lines of mature, heritage code variety of common scientifi c formats. In inventory metadata to HDF-based fi les. that represents a variety of disparate sci- 1993, after careful review of more than At the time NASA selected this standard entifi c data models. The lack of under- a dozen alternatives, NASA chose the for use by EOS, HDF was in version 4. lying commonality in the implementa- Hierarchical Data Format (HDF) as In this article, I refer to the version of tion of these models contributes to the the fi le format for EOS Standard Prod- HDF-EOS built on HDF4 as HE4. complexity of the code. This conceptual ucts. The HDF is a fi le format, applica- complexity, in turn, makes it diffi cult to tion programming interface (API) and In 1999, NCSA released HDF version adapt the library to modern high perfor- implementing library developed by the 5, a greatly improved, but structurally mance computing architectures. National Center for Supercomputing incompatible format. That is, the data Applications (NCSA). HDF is well model or internal storage implemented NCSA spent three years looking for suited as a standard for Earth Science by HDF version 5 is very different than ways to extend and adapt HDF4 to meet data. It is self describing, it is portable that implemented by HDF version 4.For these challenges, but in the end it was across many computing systems, and many, the new data format has been a clear that such an adaptation would it is designed explicitly for scientifi c reason for concern; either about costs only result in an extremely complex use with predefi ned structures common to transition from HDF4 to HDF5 or format and I/O library, which would

25 THE EARTH OBSERVER

not only be diffi cult to maintain, but NASA and NCSA understand this, and under the ECS contract and in the would not meet these new requirements are striving to assure that these chal- future. The HE4 API is stable at HDF- nearly as well as a completely new lenges will not be any more burdensome EOS 2.72 released March 1, 2001. The design would. Indeed, it was felt that, if for science data producers than neces- next release is planned for September the HDF libraries were not completely sary, especially for the science data end 2001. There are no major new capabili- overhauled, the data format and soft- users. We will do that by developing ties, only bug fi xes. NASA anticipates ware would gradually become unable compatibility and transition tools, by that HE4 will continue to have mainte- to support the modern computing needs working closely with teams that make nance releases that include error correc- of scientists. These pressures, and the the transitions, and by continuing to tions and porting to new hardware or lessons learned by NCSA in developing maintain the HDF4 code as long as operating systems at a rate of once to and supporting HDF4 over many years, required. twice a year for as long as support for led to the development of HDF5, a new HDF4 is required. data paradigm built from a solid foun- Despite the new HDF5, HDF4 and dation of computing science data prin- HE4 are not “dead” or even “heritage” Facilitating the Transition to ciples. formats. All standard products from HDF5 and HE5 Terra use these formats, as do the stan- The good news is that HDF5 is clearly dard products from other missions men- HDF-EOS helps superior to HDF4. The underlying con- tioned above. The standard products cepts are more robust and the workman- from the upcoming Aqua mission will The lack of direct backward compatibil- ship is cleaner, more compact, direct and also use HE4. NASA’s support for HDF4 ity of either the API or the fi le format simply more maintainable. HDF5 will and HE4 will continue for as long as sig- complicates the transition from HDF4 to be a powerful, fl exible and pragmatic nifi cant data holdings use this standard. HDF5. However, by minimizing the sci- data format for many years, or decades. The level of support for HDF4/HE4 ence impact of the changing format stan- There are no plans for a future transition includes correcting errors, porting to dard, HDF-EOS simplifi es things enor- to another, different “HDF6.”We believe new operating systems, adding func- mously. To the extent that EOS products that HDF5 “got it right;” that the capa- tionality as required and providing adhere to the HDF-EOS standard format bilities built into HDF5 will directly help-desk support for installation and and API, we will be able to provide tools benefi t the Earth Science Community use. that automatically convert to the new because they directly map into our representation or provide common read needs in the near and distant future. NCSA understands the requirements of access across implementations. That is, Just as NASA defi ned certain aggregates the EOS community and NASA’s Earth if applications use only the HDF-EOS of HDF4 structures to represent HDF- Science Enterprise. The overriding con- swath, grid, and point objects, and HDF- EOS point, swath and grid in HE4, cern is that we must assure that our data EOS metadata, the transition from HDF4 so HE5 is a standard usage of HDF5 remain usable over time. NCSA shares to HDF5 is greatly simplifi ed. Another to implement these same structures. In our goal of maintaining viability of sci- benefi t for product developers and users October, 2000, the EOS Aura Data Sys- entifi c data over an extended period. is that the HE5 API bears a very strong tems Working Group adopted HE5 as At NASA’s request, they have produced resemblance to the HE4 API. This is one the Aura platform standard. This is the detailed documentation of the HDF4 of the primary reasons for defi ning stan- fi rst EOS mission to use HDF5 and data format and software. This docu- dard EOS types using HDF. HE5 for all standard products. The Aura mentation is indispensable for long term instrument teams are together working preservation of the data because it will Transition tools to further standardize their products to retain the format design independent assure compatibility among the teams of implementing code. This is necessary Together with our software developers, by defi ning standard fi le metadata and insurance, but we do not intend to rely NASA and NCSA are investigating sev- other conventions. on it. NCSA will continue to maintain eral transition tool strategies. In evaluat- HDF4 for as long as NASA identifi es ing potential tools, we are considering Continued Support for HDF4 this as a requirement. The most recent the following kinds of questions: release of HDF4 (4.1r4) is dated Novem- HDF4 heritage and transition from ber 2000 and the next is expected this • How diffi cult is it to map HDF4 to HDF4 to HDF5 are important consider- coming fall. That release will include HDF5? ations. Many years of effort have gone some minor new functionality as well as into developing high quality data pro- bug fi xes and tools updates. • Is it possible to extend the HDF5 duction software based on the HDF4 library so that it can read and write and HE4 standards, and any change to The HE4 libraries will also continue to both formats? a new standard is a rightful concern. be supported under NASA’s direction

26 May/June 2001 • VOL. 13 NO. 3

• Can we provide stand-alone con- HE libraries. We are exploring ways designers. One tool facilitating complex verters that will effectively convert to publish this capability for HDF-EOS conversions is an HE fi le cracker tool, data from HDF4 to HDF5 or HE4 to objects as a separate API. Such a high called Java EOS Browser. This applica- HE5? level compatibility library would rely on tion can open and display contents of the HDF4 API but would incorporate either HE4 or HE5 granules. • How can we accommodate the tran- code to detect whether the fi le is HDF4 sition of EOS products that contain or HDF5 to call the corresponding HE Computer aided data engineering tool not only HE4 objects, but also con- library. tain other HDF4 objects such as We are studying the possibility of creat- images and annotations? Converting simple HDF4 fi les to HDF5 ing a computer aided data engineering tool. This tool would combine compo- • Should products be converted a Likewise, NCSA has developed a tool nents of several tools already developed suite at a time, a collection at a time called “h4toh5” that converts an entire to display the structure of an EOS prod- or one granule at a time? HDF4 fi le into a corresponding HDF5 uct in HDF4 and permit the user to fi le. “h4toh5” can be used as a stand- interactively transform selected objects • How long can we assume support alone utility, or it can be incorporated into equivalent HDF5 objects. Guided from third party applications? into other software. For example the by standard transformations, the user NCSA HDF5 viewer/editor, H5View, would effi ciently create a standard We think that the answers to these ques- uses “h4toh5” to read an HDF4 fi le, con- translation for a particular product type. tions, like so many technical details that vert it to HDF5, and view it. Hence, One result of this interaction would be relate to such a large and varied fi eld, it is already possible with H5View to the production of the framework of a will differ by application area, science view an HDF4 fi le as if it were an HDF5 C language program that could be used discipline and product. fi le. NCSA is now also developing an to transform other instances of the same “h4toh5” library, which consists of a col- EOS product. This code could be devel- NASA and NCSA are providing direct lection of function calls for converting oped into a stand-alone converter, an help in the form of “compatibility tools” individual HDF4 objects to HDF5. This import front end to a higher level tool, to many science product developers library provides more fi ne-grained con- or inserted into a later version of prod- and vendors of tools that work with versions than does the “h4toh5” tool, uct generation software. EOS data. With compatibility tools in and can, for example, be used to convert place, the science team can confi dently only selected HDF4 objects to HDF5, Wrapping the HDF4 and HDF5 time any contemplated conversion to the or to change the relative placement of libraries into one I/O library point of least disruption. It is likely that objects in moving them from HDF4 to such a change would be in conjunction HDF5. It has been suggested that a single with scientifi cally necessary reprocess- library be implemented that could ing. Such reprocessing generally makes Converting complex HE4 fi les to HE5 entirely support both the HDF4 and previous instances of the product obso- HDF5 APIs, making it possible for tools lete and so interuse between product The EOS standard does not require to read data either from HDF4 or from versions is not as great a concern. Still, products to contain HDF-EOS objects HDF5 without knowing the difference. data interoperability requires that a data exclusively. Typically, science data This approach has been studied exten- product in HE5 must be operable with product developers have placed both sively by NCSA, and unfortunately it older HE4 products. HE4 and HDF4 objects into a single was found to be nearly impossible to product. These products pose special create a wrapper for all HDF read and The most basic transition tool is a data challenges for automated conversion. write operations. This is because of fun- fi le converter. This is a utility or API Each HE4 object (a set of related HDF4 damental differences between the HDF4 whose input is a fi le in one format and objects) must be replaced with its cor- and HDF5 data models and APIs. While whose output is a fi le in another format. responding HE5 object and then each the HDF4 and HDF5 libraries are not remaining HDF4 object must be con- backward compatible in general, it may Converting simple HDF-EOS fi les from verted to an HDF5 object. If there were yet be possible to create a top layer that HE4 to HE5 implicit relationships between the HDF- wraps both libraries for specifi c HDF EOS objects and the other HDF objects, read and write operations (i.e. the HDF- The ECS contractor has developed a the conversion software would likely EOS operations). Such a narrow com- tool called “heconvert” that converts not know about these and so may not patibility may be suffi cient to provide HE4 Grid, Point and Swath objects to act appropriately. In these cases the data most necessary product capabilities to the same respective HE5 objects. The format conversion requires engineering general purpose application layers such “heconvert” program incorporates both intervention to retain the intent of the as Matlab, IDL or other data analysis or visualization environments.

27 THE EARTH OBSERVER

That said, it is almost certain that there will be products that use HDF4 capabili- ties in unique ways. High level HDF4 to HDF5 compatibility for these products PoDAG Meeting XVIII, may not be possible for a general library. National Snow and Ice Data Writing and accessing metadata Center, Boulder CO. An example of specifi c high level com- patibility is the writing and access of metadata in HE4 and HE5 fi les. — Ron Weaver ([email protected]), NSIDC DAAC Manager, National Snow HDF-EOS fi les contain inventory and and Ice Data Center archive metadata conformant to the EOSDIS data model and stored in Object Description Language (ODL). The inventory metadata is a copy of the metadata produced at the time the data product is created and stored in a data- base for purposes of locating data in The 18th meeting of the NSIDC DAAC • NSIDC will review the availability the archives. The ECS Science Data Pro- User Working Group (PoDAG) was held of data at other DAACs that could cessing Toolkit version 5.2.7.3 can access at NSIDC in Boulder April 25 and 26. be used to address some of the EOSDIS metadata in either HE4 or HE5 PoDAG members in attendance were: data gaps identifi ed by the NRC. If fi les. The API will recognize which ver- Dave Bromwich (chair), Cecilia Bitz, Jen- there is appropriate data available, sion of HDF-EOS is being processed and nifer Francis, Dorothy Hall, Gregory NSIDC will assess it usability for respond accordingly. Hunolt, Jeff Key, Thorsten Markus, polar applications. Chris Shuman, Koni Steffen, Mike Van Conclusion Woert, Anne Walker, Quinton Barker The above are the major recommen- (NASA ESDIS), and Waleed Abdalati dations. For a complete set please con- NASA recognizes that a burden has (NASA HQ). sult the minutes on the DAAC web- been placed on EOS data producers pages at: nsidc.org/NASA/PODAG/. and users by the introduction of HDF5. The purpose of this group is to provide These webpages also contain action In the case of EOS Terra, Aqua and user feedback on NSIDC DAAC datas- items and recommendations from all the other current HDF4 users, no require- ets, DAAC scientifi c data priorities, and past meetings, as well as the current ment will be levied to convert data to to generally represent the cryospheric membership and their email addresses. the newer format. HDF4 will be sup- user community’s data related interests Anyone wishing further contact with the ported by NASA as long as a require- in NSIDC DAAC programs. The PoDAG PoDAG should email either Dave Brom- ment exists. Data granules based on has met every six to nine months for the wich, the current PoDAG chairman or HDF4 will remain accessible though past ten years. Ron Weaver the NSIDC DAAC Man- EOSDIS and currently available services ager. will be supported indefi nitely. Given Discussion at the most recent meeting the superior quality of the newer HDF5 resulted in the following actions: The next meeting of the PoDAG will be format, we believe that science teams November 1-2 at Goddard Space Flight should, over the next few years, care- • The NSIDC DAAC will develop Center, Greenbelt, MD. For details of fully consider defi nition of new prod- information or tools to facilitate this meeting, please consult the PoDAG ucts using the HDF5 standard as the access and use of MODIS data sets. website noted above. Aura team has. For conversion of exist- ing products, change should likely be • PoDAG members will review and in the context of scientifi cally appropri- comment on NSIDC dataset Guide ate data processing or re-processing. To Documents with the aim to improve facilitate transition to the new standard, their readability. a variety of tools will be available. • NSIDC will review the gaps in The status of the HDF and HDF-EOS the topics spanned by the DAAC standards is the subject of an annual in response to the NRC report on NASA Polar Geophysical Data Sets. (Continued on page 32)

28 May/June 2001 • VOL. 13 NO. 3

Version-0 (V0) prototype was a proof-of- concept for interoperability among dis- tributed archives. The infrastructure EOSDIS’ EOS ClearingHOuse was designed to provide data search, browse (viewing a sample image), and (ECHO) Architecture order functions to scientists in all Earth Science sub-disciplines. Over a 4-year Designed to Assist Diverse effort, V0 was developed on top of the existing data and services at each DAAC. The V0 prototype succeeded in User Community providing single-point access, through a common interface, across distributed — Richard Ullman ([email protected]), ESDIS Information Architect data at the archives (Figure 2). During — Robin Pfi ster (pfi [email protected]), ESDIS IMS System Engineer the development of V0, information technology was revolutionized. As tech- nology evolved, the user community expected the user interface to be pre- sented in the latest technology on their desktop. In the beginning, character- based user interfaces run on VT100 terminals were the standard. After Introduction 1. The IMS components that provide a year into prototyping, X-Windows- search and order of data among the dis- based graphical user interfaces took Over the past decade the Earth Science tributed archives are based on a pro- over the market. At the operational Data and Information System (ESDIS) totype that was initiated in 1991. The release of the X-Windows-based Graphi- Project at Goddard has been developing, operating and improving components Figure 1. EOSDIS Functional Components. of an Information Management System (IMS) to provide users access to EOS Science Data Distribution, Data Capture, Data Processing, Access, and related data. There have been Flight Data Initial Processing, Transport Info. Mgmt, & Interoperability many changes since early prototyping Operations Acquisition Backup Archive to DAACs Distribution Reuse began a decade ago. User’s needs have changed and information technology has been revolutionized. These changes System have driven the evolution of the basic Extenders architecture and design of the IMS. Figure 1 illustrates the overall functional ASF EDC architecture of the EOSDIS in 2001. * GSFC JPL EOSDIS is an end-to-end ground data TDRSS * LaRC processing system that includes fl ight EOS Spacecrafe t ORNL Research NSIDC Users operations, data acquisition, data cap- EOSDIS EOS SEDAC o Backbone Internet ture, initial processing and backup Data Network DAACs o archival. Data are routed to the Distrib- and o Operations uted Active Archive Centers (DAACs) System * Receive Level 0 EOS Data Education or Science Investigator-led Processing Level 1-3 Products Users Systems (SIPSs) where data are further (EDOS) *Instrument processed into higher level products. PI facilities & SIPS The processed data are archived in the Data Value-Added DAACs and distributed to the end user Assimilation Providers Model community. Ground Stations EOS Operations Center (EOC) The Information Management Subsys- tem (IMS) is in the area of “Access and Interagency Int’l Partners Non- EOS Data Data & Data Interoperability” on the right in Figure Instrument Centers Centers Support Terminals

29 THE EARTH OBSERVER

cal User Interface (GUI), html interfaces search and order systems as Internet defi nes two tiers of metadata that emerged and quickly took over the navigation and paradigms of e-com- describe data holdings. The most gen- market. Just as users were getting used merce have become routine. Users have eral level is a catalog of datasets; a to the cumbersome click-and-wait inter- also become more familiar with the small core of metadata used to describe action of html, Java emerged as a more EDG interface to EOSDIS. Through this each collection. Key attributes are data- interactive option. Our development experience and user feedback, we have set name, topic area, coverage extent team could barely develop basic func- learned that different science disciplines and archive location. The more detailed tionality before it all needed to be sometimes desire very different ways tier is the inventory of granules (the redone with a new user interface tech- to access data. Using the “one-size-fi ts- smallest unit of data independently nology. all” approach with the V0/EDG, IMS managed in the inventory) for every cannot continue to satisfy our communi- dataset. In the heritage and present Operational since August of 1994, the ty’s expectations. EOSDIS, the granule is also the smallest prototype, now called the EOS Data unit of data that can be retrieved by the end-user. The EDG architecture rein- Figure 2. Version 0/EDG Interoperability forces the logical hierarchy of collection

Earth Science Global and inventory metadata by physical sep- Change Network Master aration of the classes and their hierar- View Directory chical application to the search process. Using catalog metadata, the EDG guides Information Management System (IMS) the user in construction of a search query. After it is constructed, the query ORNL LaRC EDC-V0 NSIDC-V0 ASF (U. GSFC JPL SEDAC/ Affiliated International is submitted by EDG in parallel to each (DOE) -V0 -V0 (USGS) (U. of Co./ of Alaska) CIESIN Data Center: Partners: NOAA) of the interoperable data centers that are U. of Alabama Australia identifi ed by the catalog as having gran- (Global Canada GSFC LaRC EDC - NSIDC Hydrology Germany ules that may result in a hit. The two-tier -ECS -ECS ECS - ECS Research Japan Center) Russia system is intended to reduce the occur- Israel rences of nonsensical queries. Even so, it is still possible to generate a query that results in too many or too few data Gateway (EDG), has grown in capabil- To address these changing expectations, items. ity, user base, and data provider base. NASA is re-architecting the IMS and In addition to the archives in the U.S., developing a component called ECHO Other shortcomings with the architec- EOSDIS has interagency and interna- (EOS ClearingHOuse) that will support ture have been identifi ed. The query is tional partners who supply data through varied paradigms needed by our users. performed in parallel among the distrib- interoperable links and collaborate on Through the use of the eXtensible uted archive systems. The fi nal result other EOSDIS activities. Today, EDG Markup Language (XML) and e-com- is not known until all systems have provides access to over 1200 datasets merce concepts, ECHO will provide responded. The performance and reli- at 18 heritage, international and ECS more fl exibility for end users and ability of the system as a whole and data archives. The URL for EDG is strengthen the ability of discipline spe- at any particular moment is bound to eos.nasa.gov/imswelcome. cifi c groups and data providers to serve the performance of each member. If net- particular sub-communities. The ECHO- work traffi c or systems problems affect New Architecture Supports Varied provided framework will enable better one member archive, all queries are Data Access Paradigms access to data because discipline specifi c affected. If a member archive experi- clients will show a view Changes in User Needs and Commu- of the data holdings that Figure 3. Present EDG Context Diagram. nity Characterization better meets the needs EDG of target user groups. Data Early in the EOSDIS specifi cation pro- Providers cess, scientists had limited experience The EDG Architecture Inventory Collection with large data search and retrieval Metadata WWW Metadata Catalog systems so it was diffi cult to know The V0/EDG Browsers Catalog Browse and communicate needs for data access. architecture (Figure 3) Messages: Search, Images

With the information technology rev- is designed to support User Interface Browse,& Order Data olution, information systems are now an iterative search and Holdings pervasive in everyday life. Our user retrieval paradigm for community has gained experience with data access. The system

30 May/June 2001 • VOL. 13 NO. 3

ences an outage, access to that respec- of navigation and discovery. The old communities will benefi t from ECHO tive inventory is unavailable. The tight EDG architecture mandated a multilevel because they will have data identifi ca- coupling of the EDG web server to search and discovery paradigm. ECHO tion and access tools that best suit their the collection catalog middleware is opens possibilities to new methods needs. another source of diffi culty. The addi- because bringing all the metadata into tions of new data collections to the a clearinghouse solves the problems of Data providers benefi t too. Providers catalog or new client/server function- “slowest member” IMS reliability and need not provide direct public access to ality are unacceptably intertwined and performance. System performance for data holdings. The resources required it is diffi cult to provide alternate inventory search is dependent only on for enabling public search and access views through the single server. Equally the ECHO system itself. Resources nec- can be off-loaded to ECHO. By publish- important is the emergence of new ways essary to maintain reliability can be ing data holdings through ECHO, pro- of searching. The EDG concept is a fi rst focused on the single system. Not only viders give data inventory access to all generation offi ce automation system. will clients have effi cient access to mul- clients. Novel clients will attract new The construction of queries based on tiple levels of metadata, but an added customers to fi nd these data. catalog metadata is a familiar and a nat- benefi t is that users can ural extension of traditional library cat- search and fi nd data Figure 4. ECHO Context Diagram. alog searches translated to a computer regardless of provider environment. Other data access para- down time. Only after Machine ECHO digms made possible with computer the user identifi es the And User Data & Service technology rely on much richer meta- granules of choice, is the Interfaces Collection Providers data than is available at the catalog link to the data provid- & Inventory level. Many of our users prefer the ers necessary to place Catalog Data paradigm of navigation and discovery. an order. Even then, EDG & Others Orders Holdings Rather than performing iterative que- if the provider or net- Sample Client APIs ries, these users want the system to work link to the pro- Browse Provider APIs WWW Browsers present the available data so they can vider is down when Images simply navigate the options (presented the user submits an by metadata) and discover desired data. order, ECHO will store Prototyping efforts have revealed that the order on behalf of the navigation and discovery paradigm the user and forward it cannot be supported with the present when the data provider system is acces- This new architecture offers other architecture because network separation sible. The user may not even know that opportunities to streamline user pro- of the collection-level metadata from the order was not immediately submit- cesses. Currently, we see several data the inventory metadata makes such an ted. services such as subsetting and on- application impractical. demand processing being developed The client message passing layer per- independently and remotely from the The ECHO Architecture mits varied clients to interact with the data providers. Although, these services metadata in the clearinghouse. In order are intended to operate with the data in The new ECHO architecture makes sev- to build a data access client, developers the provider archives, there is no pres- eral important changes from the heri- will code to the standard message pro- ent mechanism in the EDG system to tage EDG architecture. Two are most tocol. All clients will have the same connect data to these value-added ser- signifi cant. First, the metadata holdings level of access to the metadata repre- vices. The burden, therefore, is on the are no longer fragmented by the catalog senting data holdings from all providers user to fi nd the data, to separately fi nd and inventory hierarchy, all metadata to ECHO. The traditional iterative query the service, make the necessary connec- is brought into the clearinghouse. The interface of the EDG can be supported. tions to get the data transported to the second is that the strong coupling But new clients may become interfaces service location, and fi nally apply the between the metadata holdings and the of choice as they can be designed for service to the data. The ECHO design user interface is severed (see Figures 3 the specifi c needs of particular user includes a service brokering mechanism and 4 for comparison). In its place is an communities. These clients may provide to transparently interact between data abstraction layer consisting of an XML navigation- and discovery-based views and service providers so the user is based message protocol. of the combined data holdings. They never aware of the data transport and might provide tailored views of par- service application details. The integration of catalog and inventory ticular kinds of data products or short- Conclusions metadata with the inclusion of browse cut identifi cation techniques unique to imagery is the key to implementation certain classes of data. These end-user The clearinghouse architecture based on

31 THE EARTH OBSERVER

common web and e-business practices and open interfaces will enable better access to EOSDIS data. The infrastruc- ture will benefi t both users and provid- EOSDIS Training ers by permitting reliable access to sys- tem-wide metadata and fostering devel- — Robin Pfi ster (pfi [email protected]), NASA Goddard Space Flight Center, Green- opment of targeted client interfaces. The belt, MD heritage EDG client will continue to be supported, but niche community cli- ents will have new access to multiple provider sources. ECHO allows diverse NASA’s Earth Science Data Information Systems Project is sponsoring free train- communities to share tools, services and ing on the use of tools to search and order EOS and related data held in the metadata, while empowering commu- EOSDIS archives, and to visualize and manipulate HDF-EOS data. This includes nities of similar users to implement data collected from the Terra satellite. Instruments on Terra include ASTER unique views of the inventory and (Advanced Spaceborne Thermal Emission and Refl ection Radiometer), CERES methods of data exploration that best (Clouds and the Earth’s Radiant Energy System), MISR (Multi-Angle Imaging serve their needs. Greater system-wide Spectroradiometer), MODIS (Moderate Resolution Imaging Spectroradiometer), reliability and network response is more and MOPITT (Measurements of Pollution in the Troposphere). The training URL economically possible using ECHO. can be found at esdis-it.gsfc.nasa.gov/workshop Data users and data providers both will enjoy a higher level of service. This training is open to all interested scientists, data managers, data processing specialists, programmers, and others.

Two identical all-day sessions will be held at the GSFC Building 1 Training Center. Session 1 is Tuesday, October 16. Session 2 is Wednesday, October 17. Each session begins at 8:30 am and runs to 5:30 pm.

(Continued from page 28) Training will include: Status and Plans for • EOS Data Gateway (EDG) for search and order of data HDF-EOS, NASA’s Format • HDF-EOS Tools for organizing, dumping and some manipulation of data for EOS Standard Products • Data Visualization Tools for HDF-EOS Space is limited due to the hands-on nature of most of the sessions so workshop open to all. This year, the please sign up early. Questions about the training session should be directed Fifth Annual HDF/HDF-EOS Workshop to Robin Pfi ster, Code 423, NASA/GSFC, Greenbelt, MD 20771, e-mail: is to be held September 19-21 at the robin.pfi [email protected]; tel. (301) 614-5171). Hawthorne Suites Hotel in Champaign IL. Strategies and plans for the new Please complete the registration form below and e-mail to Robin Pfi ster by format standard will be a major focus of September 30. discussion. The workshop also features demonstrations and tutorials. See the Everyone must answer these: HDF-EOS web site: hdfeos.gsfc.nasa.gov Citizenship: If not U.S., do you have a visa? If so, what kind? for more details. Name: Affi liation: The following URLs contain information Address: about the HDF4 to HDF5 transition: Phone: http://hdfeos.gsfc.nasa.gov/HE5.html Fax: http://hdf.ncsa.uiuc.edu/h4-h5.html. E-mail: What would you like to get out of this class.

Our training website is at http://esdis-it.gsfc.nasa.gov/workshop If you are more interested in learning about HDF-EOS format (as opposed to data access tools) there will also be an HDF-EOS workshop in Champaign, IL September 19-21, that concentrates more on the content of the HDF-EOS fi les. The URL for the workshop is hdfeos.gsfc.nasa.gov/hdfeos/workshop.html

32 May/June 2001 • VOL. 13 NO. 3

New Earth Science Applications Site

NASA Earth Science Enterprise’s Appli- cations Division applies the results of the nation’s investment in NASA research to issues of national concern, Black Rock Forest Carbon such as environmental quality, resource management, community growth, and Initiative disaster management. Their new WWW site, gaia.hq.nasa.gov/eseapps/, pro- — Blanche Meeson ([email protected]), NASA Goddard Space Flight Center — Theresa Schwerin ([email protected]), IGES vides information and links to over 200 Earth science applications projects sponsored by NASA. Users can also search applications projects by state, affi liation, sponsoring program, and principle investigator.

Argonautica Project

NASA Goddard Institute for Space plants and soils, 3) analyze these data Agronautica is an initiative of the Studies (GISS) has begun piloting its sets to quantify carbon processes, 4) French Space Agency CNES’ (Center second research program for teachers apply the new knowledge gained to Nationale d’Etudes Spatiales) Education and students. The Black Rock Forest current science questions and environ- and Youth Department, developed to carbon study is being lead by two teach- mental issues of public concern in the promote the upcoming Jason-1 launch. ers who have previously participated New York metropolitan area such as It is intended to show both students in research through GISS’ Institute on urban development and sprawl from and teachers how space oceanography Climate and Planets (ICP) and their sci- New York City. may help to understand the Earth’s cli- ence advisor, Dorothy Peeteet. The Black mate and the socio-economic impacts of Rock Forest Director is also a science The ICP is a research, science education, global warming. advisor to the project. and minority education program. It involves pre-college and undergraduate One project sponsored by Agronautica Students and teachers throughout the students in current NASA climate and is the “Vendee Globe” buoys project. For NY metro area participate in research planetary investigations in collaboration this effort, skippers participating in the conducted by science teams at NASA with teachers and faculty from their “Vendee Globe” round-the-world yacht GISS. The majority of these projects take schools and colleges and GISS research race have released a total of 10 buoys. place at NASA GISS on the campus of scientists. Learning modules and tools They are tracked by satellite (Argos), Columbia University. This is the fi rst for teaching Earth system science are and the data are transmitted several summer that they will have an outdoor available on the ICP WWW site. These times per day. fi eld experience program for students include lessons, computer models, tuto- and teachers in Black Rock Forest. rials, and datasets and analysis tools. A total of 15 schools are participating, For more information on ICP and GISS, comprising 24 different classes ranging Black Rock Forest is a scenic area see icp.giss.nasa.gov/ or contact Caro- from nursery school to secondary comprised of several different ecosys- lyn Harris at [email protected]. school. They study a diverse range of tems and plant/animal habitats about 50 subjects including science, geography, miles north of New York City. Partici- JPL Releases New WWW Site computing and technology. The 550 pants will be a part of a science team youngsters involved are enthusiastically organized to improve our understand- On May 18, NASA’s Jet Propulsion Lab- following the progress of the boats and ing of human and natural contributions oratory released its new WWW site, buoys. Buoy positions are displayed on to the carbon cycle in this typical East which includes new Earth science pages, CNES’s educational website, www.cnes- Coast forest. By conducting a range at www.jpl.nasa.gov/earth/ edu.org, where participants from all of outdoor fi eld investigations over a earth_index.html. These pages focus on over France can exchange ideas, posi- three week period, the science team will: JPL research and missions by topic: Air tions, and even deductions. Each school 1) characterize and classify the forest & Ozone, Solid Earth, Ocean Motion, ecosystems, 2) collect measurements on Natural Hazards, Weather and Climate, carbon sources and carbon storage in Missions, and Images & Video. (Continued on page 34)

33 THE EARTH OBSERVER

(Continued from page 33) Earth Science Education Program Update

has been allocated a buoy and has received an educational kit with a map to plot the buoy’s position and compare its route with ocean current charts. Additional data on wind and wave — Rob Gutro ([email protected]), EOS Project Science Offi ce, heights, acquired by oceanographic sat- NASA Goddard Space Flight Center, SSAI, tel. (301) 286-4044, Fax: (301) 286-2322 ellites, can be made available for teach- ers who wish to do a more in depth study of this subject or for those who wish to have extra material to use in the classroom.

Major EOS news hotspots in this microbes are being transported in Digital Library for Earth System period include carbon sinks, clouds of dust blowing into the Ameri- Education (DLESE) Science cooling clouds, African dust, El cas from Africa. Policy Collections Group Niño effects and satellites. “Research Satellites to Ride in Tandem to The purpose of the DLESE Science “Measuring America’s Greenhouse Orbit,” (June 8) CNN.com. The TIMED Policy Collections Group is to create a Effect,” (June 21) MSNBC.com. New and Jason-1 satellites slated for launch national forum for organizing science research estimates that U.S. forests, this September will give scientists new policy documents which are relevant to crops and rivers absorb up to one-half information about the Earth’s upper Earth System Science education and the of all carbon dioxide emitted into the atmosphere and oceans. Lee-Lueng Fu sustainable development of our society air annually when Americans burn fossil (NASA JPL) explained that Jason-1 will at global to local levels. fuels, but that amount may decline replace the TOPEX/Poseidon satellite. according to Stephen W. Pacala (Princ- To achieve its purpose, the INSPIRE eton University). Jerry M. Melillo (Eco- “El Niño Phenomenon Boosts Snow- program (Integrating Science into systems Center of the Marine Biological fall,” (May 31) Associated Press by Ran- Policy: Interdisciplinary Research Edu- Lab in Woods Hole) commented on the dolph E. Schmid, and USAToday.com. cation) was designed for the National U.S. carbon sink. New research from James J. O’Brien and Science, Mathematics, Engineering, and Shawn Smith of Florida State University Technology Education Digital Library “Both Sides Now: New Way That indicate that the El Niño phenomenon (NSDL) ‘collection track’ through the Clouds May Cool,” The New York brings more snowfall to Stowe, Ver- National Science Foundation. Examples Times, by Andrew Revkin (June 19). mont, and Amarillo, Texas, and can of science policy documents that have While scientists largely agree that reduce snowfall in other areas of the been organized into searchable data- humans are warming the Earth by United States. bases are: Antarctic Treaty Searchable adding greenhouse gases to the atmo- Database: 1959-1999 webhost.nvi.net/ sphere, debate continues about the role “MOPITT Satellite Paints Smoggy Por- aspire, and United Nations Law of that clouds play as a cooling infl uence. trait of the Planet,” (May 30), CNN the Sea Convention Searchable Database O. Brian Toon (University of Colorado), Cable News, MSNBC.com, and CBS webhost.nvi.net/inspire. John H. Seinfeld (Caltech), and James Evening News. John C. Gille (NCAR) E. Hansen (NASA/GISS) appear in this and Daniel J. Jacob (Harvard University) If you are interested in participating in article. said the MOPITT instrument on NASA’s the Science Policy Collections Group, Terra satellite has tracked plumes of please contact: Paul Arthur Berkman, “African Dust Brings Bacteria to the carbon monoxide moving around the Byrd Polar Research Center, Americas,” (June 14) The New York world, indicating that pollution is not a [email protected] or berkman.1@osu. Times, Associated Press, CNN, ABC local problem. edu. For more information about DLESE News, MSNBC.com. Jay Herman see www.dlese.org/. (NASA/GSFC) was one of four researchers who discovered that

34 May/June 2001 • VOL. 13 NO. 3

Contact Christopher Velden, e-mail: [email protected]; EOS Science Calendar e-mail: [email protected], URL: www.geomatics2002.org. URL: fermi.jhuapl.edu/sat_met_ocean/ Oct. 26-28 September 18-20 December 5-7 3rd International Symposium on CERES Science Team Meeting, Brus- ‘The times they are a changing’: Sustainable Agro-environmental Sys- sels, Belgium. Contact: Jennifer Hubble, Climate change, phenological responses tems: New Technologies and Appli- NASA Langley, tel. (757) 864-8333, and their consequences for biodiversity, cations, Cairo, Egypt. Contact Prof. e-mail: [email protected] agriculture, forestry, and human health, Derya Maktav, e-mail: Wageningen, The Netherlands. Call [email protected] September 19-21 for Papers. Contact Arnold van Vliet, 5th Annual HDF-EOS Workshop, Uni- [email protected], November 8-15 Integrating Remote Sensing at the versity of Illinois at Urbana Champaign. tel. 31 317 485091/484812, URL: Global, Regional and Local Scale, Contact Richard Ullman, www.dow.wau.nl/msa/epn/conference/ The 15th William T. Pecora Memorial e-mail: [email protected]. Remote Sensing Symposium/Land Sat- December 10-14 ellite Information IV Conference and the 2001 AGU Fall Meeting, San Francisco, September 24-26 ISPRS Commission I (Platforms and CA. For more informatino, MODIS Science Team Meeting, loca- Sensors) Symposium, Denver, Colo- tel. 1 (800) 966-2481 or 1(202) tion: TBD (local to GSFC). Contact: rado; conference Website www.asprs.org/ 462-6900; Fax: 1(202) 328-0566; Barbara Conboy, NASA GSFC, Pecora-ISPRS-2002. e-mail [email protected]; e-mail: [email protected] URL: www.agu.org/

October 29 - November 1 U.S. TRMM Science Team Meeting, 2002 Fort Collins, CO. Contact: Robert Adler, e-mail: [email protected] January 21-23

Non-CO2 Greenhouse Gases (NCGG-3) scientifi c understanding, control options October 30-November 1 and policy aspects, Maastricht, The EOS IWG Meeting, San Antonio, TX Netherlands. Contact Dr. Joop van Contact: Mary Floyd, Ham. e-mail: [email protected]; tel. e-mail: mfl [email protected], 31-15-285-2558; Fax: 31-15-261-3186; For registration and location details see URL: www.et.ic.ac.uk/Dept/LocalNews/ URL: eospso.gsfc.nasa.gov/ greenhouse.htm.

April 7-12 Global Change Calendar 29th International Symposium on Remote Sensing of Environment Oct. 2-5 “Information for Sustainability and Conference on Regional Haze and Development,” Buenos Aires, Argentina. Global Radiation Balance - Aerosol Call for Papers. Contact Secretariat: Measurements and Models: Closure, email [email protected], Reconciliation and Evaluation, Bend, URL: www.symposia.org. Oregon. Contact Terry Mohr, tel. (412) 232-3444, ext. 3147, July 7-10 URL: www.awma.org 2nd Large Scale Biosphere-Atmosphere Experiment in Amazonia (LBA) Science Conference, Manaus, Brazil. Contact October 7-10 Flavio Luizao of the National Institute for 2001 International Conference on Image Space Rsearch (INPA), Manaus, Brazil, Processing, Thessaloniki, Greece. Call e-mail: [email protected] for Papers. Contact Diastasi, tel. +30 31 938 203, Fax +30 31 909 269, July 9-12 e-mail [email protected] 2002 Joint International Symposium on GeoSpatial Theory, Processing and October 15-18 Applications, Ottawa, Canada. Call for 11th Conference on Satellite Meteo- Papers. For details, rology & Oceanography, Madison, WI. tel. +1 613 224-9577;

35 Code 900 PRSRT STD National Aeronautics and Postage and Fees Paid Space Administration National Aeronautics and Space Administration Goddard Space Flight Center Permit G27 Greenbelt, Maryland 20771

Offi cial Business Penalty For Private Use, $300.00

The Earth Observer

The Earth Observer is published by the EOS project Science Offi ce, Code 900, NASA Goddard Space Flight Center, Greenbelt, Maryland 20771, telephone (301) 614-5559. Fax (301) 614-6530, and is available on the World Wide Web at eos.nasa.gov/ or by writing to the above address. Articles, contributions to the meeting calendar, and suggestions are welcomed. Contributions to the Global Change meeting calendar should contain location, person to contact, telephone number, and e-mail address. To subscribe to The Earth Observer, or to change your mailing address, please call Hannelore Parrish at (301) 867-2114, send message to [email protected], or write to the address above.

The Earth Observer Staff

Executive Editor: Charlotte Griner ([email protected]) Technical Editors: Bill Bandeen ([email protected]) Jim Closs ([email protected]) Alan Ward ([email protected]) Design and Production: Winnie Humberson ([email protected]) Distribution: Hannelore Parrish ([email protected])

Printed on Recycled Paper