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

THE EARTH OBSERVER A Bimonthly EOS Publication July/August 1999 Vol. 11 No. 4

In this issue EDITOR’S CORNER Michael King SCIENCE TEAM MEETINGS EOS Senior Project Scientist Minutes of The Fifteenth Earth Science Enterprise/Earth Observing System (ESE/EOS) Investigators Working In the past month, the 1999 EOS Reference Handbook was completed and is now Group (IWG) Meeting ...... 6 being printed. The purpose of this Reference Handbook is to provide a broad overview of the Earth Observing System (EOS) program to both the science SOlar Radiation and Climate Experiment community and others interested in NASA’s Earth Science Enterprise (ESE). (SORCE) Science Team Meeting..... 18 This edition includes a brief history of EOS from its inception, science CEOS Working Group on Calibration objectives, mission elements, planned launch schedules, descriptions of each and Validation Meeting on Digital instrument and interdisciplinary science investigation, background informa- Elevation Models and Terrain tion on team members and investigators, international and interagency Parameters ...... 19 cooperative efforts, and information on the EOS Data and Information System SCIENCE ARTICLES (EOSDIS). A number of figures and tables are included to enhance the Status of NASA’s Earth Science Enter- reader’s understanding of the EOS and ESE programs. It is available electroni- prise: A Presentation by Dr. Ghassem cally from http://eospso.gsfc.nasa.gov/ eos_homepage/misc_html/ Asrar, Associate Administrator for Earth refbook.html, and will be available in hard copy by September 30. Copies may Science, NASA Headquarters ...... 3 be obtained by sending e-mail to Lee McGrier at [email protected]. NASA EOS Validation via SAFARI nasa.gov. 2000: Preparing the Way ...... 21 The EOS Project Science Office has completed an Educational CD-ROM SHADOZ (Southern Hemisphere entitled Welcome to NASA’s Earth Science Enterprise, which provides educators, ADditional OZonesondes): A New Ozonesonde Data Set for the Earth students, and the general public with an overview of the latest research Science Community ...... 27 related to our global environment, while providing the necessary background material to comprehend these studies. The CD-ROM is divided into four main Pre-Launch Validation Activities sections: Land, Sea, Ice, and Air. In addition, there are overviews of EOS for the MODIS Snow and Sea Ice Algorithms ...... 31 spacecraft and instruments, remote sensing, a description of what EOS will measure (the EOS 24 measurements), and a description of why and how we ANNOUNCEMENTS will study the Earth. Accompanying this CD-ROM is an Activity Supplement Landsat 7 Team Receives The American (Teacher’s Guide) which, together with the CD-ROM, enables students and Institute Of Aeronautics Space System teachers to further their comprehension of seven topics highlighted on the Award ...... 2 CD-ROM, including ozone, clouds, volcanoes, El Niño, sea level, sea ice, and Earth Science Education Program deforestation. Copies of the CD-ROM and accompanying Activity Supple- Update ...... 36 ment can be obtained by sending e-mail to Lee McGrier at lmcgrier@pop900. gsfc.nasa.gov. EOS Scientists in the News ...... 37

Calendars ...... 39 The SeaWinds scatterometer began global observing mode from the QuikScat satellite on July 15. Initial calibration analyses have resulted in global back- Information/Comments ...... Back cover THE EARTH OBSERVER

scatter accuracies of ~0.3 dB and ocean community. These teams are expected to vector wind accuracies of <2 m s-1 in wind deliver their ATBDs to the Project Science Landsat 7 Team Receives speed and <20° in wind direction with Office by December 15. Following these The American Institute Of respect to operational global surface wind written evaluations, an oral portion of this Aeronautics Space analyses. SeaWinds maps winds over review will be conducted in February. ~93% of the ice-free oceans daily. Follow- These documents, developed for each data System Award ing intensive, detailed validation by the product, consist of a detailed physical and science team, it is expected that SeaWinds mathematical description of the algorithm, — Lynn Chandler (Lynn.Chandler.1@gsfc. nasa.gov), Goddard Space Flight global backscatter and ocean vector wind variance or uncertainty estimates, and Center measurements will be available to the practical considerations, such as calibra- operational and research communities in tion and validation, exception handling, The Landsat Project Government-Industry November 1999, with a fully reprocessed quality assessments and diagnostics. Team has been selected to receive the American Institute of Aeronautics Space data set available in January 2000 through Systems Award for 1999. The award will be PODAAC at http://podaac.jpl.nasa.gov. These documents will be posted on the presented on Tuesday, Sept. 28 at the AIAA NOAA has the responsibility for distribut- World Wide Web following revisions that Space Technology Conference and ing near real-time data to operational result from the written reviews as well as Exposition in Albuquerque, N.M. weather prediction centers. panel report recommendations. All of The award recognizes outstanding these missions, (with the exception of SIM, achievements in the architecture, analysis, The SeaWinds global backscatter measure- recently selected as part of the SORCE design and implementation of space ments allow all-weather, moderate mission, and Jason-1, a joint mission systems. Receiving the award will be team resolution monitoring of land and ice- between NASA and CNES), have previ- leaders from Goddard Space Flight Center (Greenbelt, Md.), Lockheed Martin Missiles covered regions as well. As one example, ously developed ATBDs and have had and Space, SBRS/Raytheon Systems SeaWinds data have already been used to them reviewed and posted on the EOS Company and the US Geological Survey. monitor iceberg B10A in the southern Project Science Office Web site at http:// Goddard managed development of the ocean near Antarctica. The iceberg is the eospso.gsfc.nasa.gov/atbd/pg1. html. Landsat 7 project. size of Rhode Island and broke off from The Landsat 7 team received the award for the Thwaites Ice Tongue in Antarctica in In the process of updating the EOS outstanding teamwork and accomplish- 1992. Iceberg B10A measures 38 by 77 km, Reference Handbook, the Earth science ments in developing the latest and most and has also been imaged by the En- mission profile was redesigned so that it advanced satellite in the Landsat series of hanced Thematic Mapper Plus (ETM+) on now has a new look (see http:// Earth remote sensing spacecraft. “Landsat 7 truly was an outstanding team effort and Landsat 7 (http://landsat7.usgs.gov/ eospso.gsfc. nasa.gov/eos_homepage/ we are very honored to receive this award,” browse/iceberg.html). On September 7 the missions.html). All of the Earth science said Phil Sabelhaus, Landsat 7 Project U.S. Geological Survey announced the missions have been reorganized by launch Manager at Goddard. availability of Landsat 7 ETM+ data to the date in chronological order, with the Landsat 7 is the latest in a series that began general public. Full scenes that have been actual date of launch for missions already in 1972. These data will provide scientists corrected for sensor effects and spacecraft in orbit. All missions contain descriptions with new information on deforestation, geometry (Level One processing) are of the spacecraft name, launch date, receding glaciers, and crop monitoring. The available from the EROS Data Center in altitude, orbital inclination, and equator data also will be available commercially for land-use planning and urban development Sioux Falls, South Dakota(http:// crossing time (for sun-synchronous issues. Landsat 7 is part of the Earth edcwww.cr.usgs.gov/landdaac/landsat7), orbits), as well as the instrument comple- Observing System, the major component of at a price not to exceed $600 each. ment onboard the spacecraft. Spacecraft NASA’s Earth Science Enterprise, a long- and/or instruments not provided by term program that is studying changes in The AIRS/AMSU-A/HSB, AMSR-E, NASA are clearly indicated on the chart. Earth’s global environment. The goal of the Earth Science Enterprise is to provide SAGE III, ACRIM III, SIM, Jason-1, and This is quite common, as our program is people a better understanding of natural Data Assimilation science teams are international in scope with many valuable changes. Enterprise data are essential to currently preparing Algorithm Theoretical international partners. people making informed decisions about Basis Documents (ATBDs) that will be their environment. reviewed by the international scientific

2 July/August 1999 • Vol. 11 No. 4

collection. But this is by no means a Status of NASA’s Earth Science foregone conclusion yet. September often brings cooler weather and “cooler” heads. Enterprise: A presentation by Dr. Ghassem Asrar, Associate Administrator for Earth Science, So, we’ve had some successes and some setbacks this year. But if it wasn’t hard, it NASA Heaquarters likely wouldn’t be worth doing, and the Earth Science Enterprise is absolutely worth doing.

Outside our community, every audience Mr. Goldin or I speak to is genuinely excited to hear how NASA is using space I am pleased to be able to address the plished in the next week or so. Langley’s technology to study our own planet, and Earth Science community at Goddard SAGE III instrument, slated for ride on a how the results of our research are during our most exciting and productive Russian spacecraft and launch vehicle, is addressing issues they care about. Inside year yet. I’d like to take this opportunity being delayed while Russia studies adding our Earth Science community, we are to lay out the big picture of the Earth another instrument. TOMS has fallen off making the transition from a data-poor Science Enterprise, and enlist your the Russian table, but Goddard is stepping research effort to a data-rich one. Our continued support to achieve our goals. up to define a QuikTOMS mission as a crude pictures of ocean winds and rainfall, Let me start by updating you on recent free-flyer, much as we did for the QuikScat of atmospheric aerosols and tropospheric events, which include some good news mission, and still plan to launch in 2000. chemistry, of ice-sheet topography, and of and some, well, other news. This will be another validation of land cover and biomass are about to Goddard’s resilience and ability to fulfill become detailed, dynamic panoramas. So far this year we are three for three in Enterprise commitments. So, where we successfully launching satellite missions. started 1999 with a plan for 10 launches, Let me now lay out for you a portrait of The February launch of SUNSAT and we will probably end with 6 or 7, with the the Earth Science Enterprise along content Oersted, Dutch/South African Earth others sliding to early 2000. But that’s the and time dimensions. Along the content remote-sensing satellites, adds two of a space business! dimension, the Enterprise comprises four five-star constellation of GPS receivers in major elements. The most fundamental, of orbit. Landsat 7 continues to perform You have probably heard that the Con- course, is Research. Our research and beautifully, and NASA and USGS hope to gress, in deliberating on the FY 2000 analysis activities, including model declare it operational in a matter of days. budget, is considering a $305-million development and mission science teams, The first images from QuikScat data are reduction for Earth Science at NASA. The are the intellectual capital of the Enter- now available, and they look very good. problem stems from the future budget prise. We are committed to peer-reviewed, caps that were agreed to as part of the cutting-edge research to identify and We continue to have difficulties with 1997 budget agreement. NASA is not answer key questions about the Earth getting rides to orbit. Our next mission, alone in this dilemma; all the agencies in system. Chief among these are: the Shuttle Radar Topography Mission, the discretionary spending “pot” face will be delayed about a month to fix a similar challenges. The tack the Congress • Ecosystems & Global Carbon Cycle – wiring problem on the Shuttle orbiter, and has taken is to preserve the missions in How do changes in terrestrial and we hope not to run afoul of other time- development and allocate the cuts against marine ecosystems affect primary critical missions in the Shuttle queue. future missions. In reality, such a cut will productivity and the global carbon Terra, our EOS flagship mission has, along hurt both the present and the future. It budget? with GOES, been delayed by problems cancels the ESSP program and cuts • Global Water & Energy Cycle – In with the RL-10 engine on the launch algorithm and EOSDIS funding to the what way are atmospheric and vehicle. We are hopeful of an October point where we will not be able to realize hydrological processes that produce launch if the RL-10 certification is accom- the full return on our investment in data severe weather, cause floods, and

3 THE EARTH OBSERVER

control water resources related to information systems. The New Millen- The third set are prototype operational climate changes? nium and Instrument Incubator programs missions that satisfy requirements of other • Climate Variability & Change – Are are perhaps the most visible parts of this agencies, and are co-sponsored by them. the changes we observe in the Earth effort today. Our advanced technology climate associated with mechanisms program leverages off several others, We have also just released an Announce- we can understand, model, and including the Space Science core-technol- ment of Opportunity for small University predict? ogy program. We plan a 10% level of Earth System Science (UNESS) missions to • Atmospheric Chemistry – Is strato- investment in technology, which will yield help train the next generation of Earth spheric ozone recovering as a result of as much as a 50% reduction in per-mission scientists and engineers. the Montreal Protocol? Can long- costs for future missions. distance transport cause significant For the long-term, beyond 2010, our intent global tropospheric pollution? Let me now take you along the time is that NPOESS and related operational • Solid Earth & Natural Hazards – Can dimension. In the near-term, from now systems will assume the responsibility for we understand the dynamics of the through 2002, we are launching the EOS long-term climate observations. Mean- Earth interior, and use this knowledge first series. Our priority is to safely and while, NASA will move on to conceive a to prepare for natural hazards such as successfully launch 22 instruments and new generation of remote-sensing tools volcanoes and earthquakes? spacecraft during that time period. Some which will enable the environmental of these are multi-instrument satellites like prediction systems of the future. We are working to increase the portion of Terra and PM-1; others are single instru- investment in Research to 25% of the ments provided for launch on foreign Much of the work of envisioning this Enterprise budget. platforms, like SAGE III and SeaWinds. future is being done right here at Our second priority is to implement a data Goddard. Beyond the first decade of the The second element is Applications, and information system that is capable of next century, we plan to take full advan- Commercialization, and Education. It is no operating these missions and receiving, tage of scientific discoveries and techno- longer possible to justify the Earth Science processing, archiving, and distributing logical innovations that will enable Enterprise on the basis of science alone. their data. As we do so, we will be national and international proactive We must demonstrate the value of our laboring to demonstrate the broad environmental prediction systems. Here research in the economy and in the scientific and applications value of the are a couple of examples: broader society. This requires us to work data from this first series of missions. with commercial firms, local governments, The current suite of operational weather and others to translate science data into For the mid-term, from 2003-2010, we satellites allows a 5-day forecast of practical solutions. We are moving to have begun to define a set of national regional weather. Over the next 20 years, achieve a 10% level of investment in this mission concepts of three types. One is a we would like to push that out to 14 days. area. set of systematic measurements that will The benefits to such industries as agricul- The third element is Mission Implementa- carry on those EOS measurements that are ture and construction, not to mention tion and Operations, in which we fund our required to build a data set useful for vacation planning, would be enormous. satellite and information systems develop- global-change studies. Many of these are What is required to achieve a 14-day ment. As we have developed the first planned for transition to the NPOESS, and forecast capability? series of EOS missions, this has consumed the missions now being considered will about 65% of the Enterprise budget. By bridge the EOS and NPOESS eras. First, we need improved observations. applying the Administrator’s better, faster, The most important new scientific cheaper principles, we plan to reduce this A second set are exploratory missions observation required is tropospheric to about 55%. similar to the Earth System Science winds measured with an accuracy in the Pathfinders that are already underway. range of 1-3 m/s, with a resolution of 1 The key to doing so is Advanced Technology, These missions will explore Earth system km vertically and 100 km horizontally. which is the fourth element. We have a processes that are not well understood That leads to a new technology require- three-pronged investment strategy for and, as such, employ innovative technical ment: tunable lidar systems. We are just technology—instruments, platforms, and approaches. approaching the beginning phases of work

4 July/August 1999 • Vol. 11 No. 4

with space-based lidars. This is part of a The key to turning all these new data into But in order to have these opportunities in broader push to expand from passive to useful information products is data the future, we need to deliver on our active remote-sensing capabilities. visualization; transforming data into promises in the present. This means running 3-D images representative of real successfully launching the missions we Another example is soil-moisture mea- processes such as storm development or El now have in development. It means surement. Accurate regional measure- Niño evolution or aerosol transport. realizing the full scientific return from our ments of soil moisture would be a great investment in those missions. And it boon for agriculture, and also aid in Currently, visualizations are limited to means working with people and organiza- assessing flood-damage potential. We plan sites with limited time on advanced tions outside the research community to to fly an aircraft instrument to demon- supercomputing capabilities. Data translate our scientific knowledge into strate one approach to this in 2001. volumes are such that visualization solutions to practical problems in the products cannot be easily shared electroni- economy and the broader society. One shortcoming present in all LEO– cally. But with advances in computing based observing systems is revisit time. power and research targeted on visualiza- And that is going to require all of us to Landsat, for example, has a 16-day revisit tion techniques, we can make progress work together toward these goals. While time. To address the problem of revisit toward an era where visualizations can be many nations are involved in Earth time for these and other measurements, World Wide Web-based and user driven. Science, the U.S. is the only nation, and we want to move toward constellations of One outcome of this will be that non- NASA is the only Agency, with the smart spacecraft, and to the use of orbits scientists will be able to gain an immedi- breadth of expertise to establish the vision beyond LEO and GEO, such as L1 and L2, ate appreciation of the power and scale of and framework for Earth System Science. to achieve continuous views of the Earth. Earth-system processes, because they will And Goddard Space Flight Center is the Such systems might be: be able to view them as they would an largest portion of this capability. animation. • Low and mid-level Earth orbit-based It is my hope, and my expectation, that the sensor webs and reconfigurable Summary: civil service and contractor science and satellite constellations. engineering community here will embrace • GEO-based large-aperture imagers As we move into the next decade, we will this vision as its own and work to see it and sounders, and event monitors be reaping the science and applications fulfilled. We must rise to the challenge of • L1 and L2 Earth and Space weather rewards of the last ten years of planning making good on our current promises to outposts the Earth Observing System. We are deliver quality data sets to the climate- already doing so for the pre-EOS missions change science community and demon- With these, we need on-orbit advanced launched since 1991. strate practical applications of these data. information systems with high-band- If we are successful, we will build confi- width communication to produce direct- But this is only the beginning. Our dence in our sponsors that the future we to-the-user data products. understanding of the Earth System is envision is possible. growing rapidly, as is our technological Second, we need fundamental research. capability to make new observations and The Earth Science Enterprise is here to I described our science themes and turn them into useful information prod- stay. With your help, we can ensure NASA questions earlier. ucts. will be there to help the nation and the world meet the challenges of proactive Third, we need modeling and data The next 20 years will be an exciting time environmental prediction for the new analysis capabilities. for Earth Science, and will witness the century. A 14-day prediction capability also leads evolution of a global environmental to new requirements for computing power monitoring capability that will enable for modeling and visualization. This regional- and local-scale decision-making includes capabilities for high-volume for a whole host of economic and societal distribution and storage as well as endeavors. processing speed.

5 THE EARTH OBSERVER

• Climate Variability and Change • Atmospheric Chemistry Minutes of The Fifteenth Earth Science • Solid Earth and Natural Hazards Enterprise/Earth Observing System Kaye pointed out that there are differences (ESE/EOS) Investigators Working Group between these ESE themes and those (IWG) Meeting adopted by the USGCRP.

— Renny Greenstone ([email protected]), Raytheon ITSS ESE also has six applications themes:

• Food and Fiber • Natural Resources The 15th meeting of the Investigators out that Michael Luther is now the Deputy • Disaster Management Working Group (IWG) of the Earth Science Associate Administrator of OES. • Environmental Quality Enterprise/Earth Observing System (ESE/ • Urban Infrastructure EOS) took place at Manor Vail Lodge in He presented a list of near-term chal- • Human Health and Safety Vail, Colorado, June 15-18, 1999. lenges. These included: Developing the first series of EOS and selected Earth Kaye said that there are two cross-cutting Tuesday, June 15 Probe missions, with ten launches planned themes: land use/land cover and weather Morning Plenary Session (Earth Science for CY 99 followed by 25 launches through and climate. Enterprise/EOS Status) CY 02; delivery of a functioning Data and Information System (DIS) to support these Regarding applications, initially there is Michael King, EOS Senior Project missions; implementation of scheduled an emphasis on the land surface and its Scientist, Chair. aircraft and field campaigns; collection ties to agricultural issues. The Stennis and analysis of data from operating Space Center focuses on the commercial In the absence of Ghassem Asrar, NASA’s missions such as TRMM; and establishing area. There is a concern that the science Associate Administrator for Earth Science, joint applications projects with other and commercial applications groups need Jack Kaye (Director of the Research Federal, State, and local agencies. Kaye to have improved interactions. Division, Office of Earth Science [OES], also listed planned EOS and “other” NASA Headquarters) gave an update on Earth-science-related missions for 1999, Commenting on the New Millennium “Earth Science Enterprise, Status and 2000, and 2001. Program (NMP) missions, Kaye reported Future.” Kaye, noted that the number of that EO-1 is “on track,” Sparcle is having EOS Interdisciplinary Science Investiga- ESE now has five science themes: cost and schedule difficulties, and EO-3 tions (IDS) has grown by a factor of three definition is progressing. Commenting on since the initiation of EOS. Kaye showed • Ecosystems and Global Carbon Cycle Earth Probes status, Kaye said that there an OES organization chart and pointed • Global Water and Energy Cycle has been a House bill that would kill Triana, but there is still hope for the Office of Earth Science Organization mission. OFFICE OF ASSOCIATE ADMINISTRATOR An ESE Technology Strategy has been Dr. Ghassem R. Asrar, AA approved and it is now in printing. Mr. Michael R. Luther, Deputy AA Included are Instrument Incubator, Advanced Information Systems Technol- ogy (AIST), and Core and Cross-cutting PROGRAM APPLICATIONS & BUSINESS RESEARCH Technologies. PLANNING & OUTREACH DIVISION DIVISION DEVELOPMENT DIVISION DIVISION The NPOESS Preparatory Project (NPP) Dr. Richard Beck Ms. Anngienetta Johnson Dr. Nancy Maynard Dr. Jack Kaye now includes an Advanced Technology

6 July/August 1999 • Vol. 11 No. 4

Microwave Sounder (ATMS). There is a ESE planning for the next decade includes Addressing a question from the audience, very important issue about how research completion of Version 2.0 of the ESE Kaye said that there is soon to be an EOS missions transition to operational mis- Science Implementation Plan. The Science IDS follow-on announcement that will be sions. Neal Lane, the President’s science Implementation Plan has 13 overarching very important. The current set of IDS advisor and head of the Office of Science questions covering the five thematic areas investigations will generally come to an and Technology Policy (OSTP), has of ESE. EOS scientists should be looking end in December 1999. The earliest due authorized drafting of a policy statement into these 13 questions. A Strategic date for new proposals is the second week on this issue. Operational missions must Enterprise Plan is in development and is of September. support research mission needs. This is an to be reviewed with ESSAAC in Septem- urgent policy requirement. Kaye showed a ber. Kaye said that the OMB is looking for In answer to a question from Ricky Rood, chart giving current schedule assump- a coherent exposition of ESE’s post-2002 Kaye said that it is essential to show that tions. It showed the dates for which NPP strategy in support of the FY 01 budget we have a Science plan that leads to a would provide coverage, leading into the request. This is a very strong OMB Mission Plan. launch of NPOESS C1 between July and concern. December 2008. Phil Sabelhaus (Goddard Space Flight Kaye concluded his formal presentation Center), Deputy Director for Develop- ESE is now supporting a New Investigator with a chart labeled “Tasks for the EOS- ment, gave the talk entitled “EOS Flight Program. There have been 17 selections IWG.” The key thought for the IWG was Program Status.” He said that there has out of 84 proposals, and there is an intent that “EOS IWG members should play a been considerable progress since the last to have solicitations every 18 months with leading role in promotion of EOS data use IWG, which was held last fall in Durham, 30 awards each time. outside the traditional EOS community New Hampshire. Notable was the highly and in recruiting/welcoming new successful launch of Landsat 7 on April 15. NASA’s outreach strategy is based on the members of the Earth science community The unfortunate failure of the RL-10 concept that it is necessary to communi- to bring new ideas and measurement engine on the Centaur upper-stage rocket cate to the public that “Earth science at concepts.” has led to a delay of at least one month to NASA is truly science in the public not earlier than August 27, 1999 for the interest.” Terra launch. TOMS is to be a free flyer and not part of the PM-1 mission. CHEM A National Research Council (NRC) report Tasks for the EOS-IWG integration has begun. ICESat is now co- has been issued assessing the ESE post- EOS 1st series plan and budget is manifested with CATSAT (The Co- 2002 mission scenario. The report says that firmly in place. operative Astrophysics and Technology ESE needs a fully integrated science plan Satellite). and embraces NASA’s call for a national • Priorities should shift toward effective implementation policy discussion on the transition from - Calibration/validation of new The FY 99 budget is OK but tight. Contin- research to operations. instruments as they come on gency has been restored to PM-1. The line Terra science-processing software has been Among ESE’s long-term “challenges” is to - Production of promised data delivered. Additional resources have been sets & scientific progress change the budget investment balance to - Organized, regular reporting of requested to support CHEM. Priorities are increase emphasis on research, advanced science results to assure spacecraft safety, provide technology, and commercialization/ resources for the launch of Terra, and to applications. ESE is challenged to adopt • EOS-IWG members should play a establish a stable baseline for ESDIS to leading role in: the Science=>Technology=>Mission - Promotion of EOS data use proceed. paradigm, with an aggressive instrument outside the traditional EOS technology development program. ESE community Landsat has now provided more than 6000 should increase science data purchases - Recruiting/welcoming new scenes. Appropriate spacing of Landsats 5 members of the Earth science from commercial suppliers and should community to bring new ideas and 7 on orbit allows an eight-day repeat form new partnerships and test pilot and measurement concepts cycle instead of the standard 16 as long as applications with regional users. they are both functioning. USGS is to take

7 THE EARTH OBSERVER

over management and funding as of work split between the two organizations mental prediction it is essential that we October 2000. Sabelhaus showed the first has been agreed upon. drive sensor costs down. NASA technology Landsat 7 image. is the key to achieving our national goals. Sabelhaus ended his presentation saying Terra is now in a stand-down mode that Landsat data operations are stable, We look to our Agency partners: NOAA awaiting the solution of the launch- and that the Terra control center is able to for weather prediction, EPA for air/water vehicle problem. In the meantime there support operations. quality, and DOI and USDA for land have been three successful ground system management. to TDRSS tests. Mark Schoeberl (Goddard Space Flight Center) addressed “Earth Science Visions Looking to the future of space observa- The PM observatory is now entering the for the Next Millennium.” He asked: tions, Schoeberl sees the need for NASA to integration and test phase. Instrument “What should NASA be doing for Earth go to smaller sensor systems at lower development activities are at or near Science in the 2020 time frame?” In the costs. He envisions a sensor web where completion. There is adequate schedule 2020 era Schoeberl feels that such factors multitudes of sensors are spaceborne at slack to December 22, 2000. ICESat will as possible changes in climate, population the same time. have a Delta II launch vehicle. Its control growth, biodiversity, and land use all will center will be supplied by Ball and the impact habitability of Earth and quality of One of Schoeberl’s charts had the mes- Laboratory for Atmospheric and Space life and have economic consequences. sage: “NASA must achieve the goal of Physics (LASP). Things that NASA can do include: enabling proactive environmental prediction.” CHEM instruments are making good • Enable accurate 14-day weather progress. The Algorithm Theoretical Basis predictions with 12-hour severe storm Looking to new technologies, he illus- Document (ATBD) reviews for MLS, forecasts. trated the value of dense networks of HIRDLS, and TES were successful. The • Enable accurate short-range climate/ space-based wind observations leading, SOlar Radiation and Climate Experiment environmental prediction (mostly for example, to improved storm locations. (SORCE) is to combine SOLSTICE and seasonal to interannual). He showed the value of improved SST TSIM. It now has a July 2002 launch date. • Enable long-range climate/environ- observations, and said that adding a soil- Sabelhaus said that significant progress mental prediction (decadal). moisture capability would also lead to has been made in the EOSDIS arena. • Enable prediction of global air and significant improvements. The successful Mission systems tests for Terra have water quality. forecasting of the 97/98 El Niño season shown that the flight operations system is • Enable prediction of natural hazards. was an example of what can be achieved getting better with successive trials. 85% • Enable efficient management of natural with the added spaceborne capability. of the segments have been verified as of resources. June 4. Connectivity tests have been Schoeberl has a vision of “nano spinners.” successful. Sabelhaus also listed areas Schoeberl said that Science and Discovery These might have 10-kg mass with no where improvements are still needed. are the stepping stones to discovery. The propulsion systems (using magnetic goal is to answer: “What can we do to torque to maintain their attitudes). He Landsat 7 is functioning routinely but make people’s lives better.” called attention to the ever-increasing there are “stability” problems—correctable need for computer power. If grid sizes for problems with the software and with the The basic requirements for prediction forecast models are to decrease from 10 operational people who are still learning include: robust scientific research activities km to half that size, then 8 times the their roles. Terra End-to-end Science focused on understanding the Earth computer power will be required. In 10-to- System Tests (TESS) have revealed some System; a stable, calibrated global observ- 15 years ECMWF will be able to use data problems with ASTER and CERES— ing system; and an integrated information at the AIRS pixel size. everything else is fine. and data-management system. (There must be the ability to present and commu- New approaches to software development As a result of a joint NASA-Integrated nicate scientific results to the public.) On call for open-model architectures and Program Office (IPO) feasibility study, a the way to achieving improved environ- automatic parallelization schemes.

8 July/August 1999 • Vol. 11 No. 4

In situ sensors will have to be improved. Potential real-time MODIS products The current status of NPP is this: There is They often serve as prototypes for space- include regional ocean color and chloro- an agreement between NASA Headquar- based sensors. phyll. MOSIX is a free software package ters and the Integrated Program Office that provides a way of parallel processing (IPO); results of a feasibility study were Rosenzweig cautioned that, despite the the DB data. The University of Wisconsin presented to Ghassem Asrar (NASA) and successes with recent El Niño forecasting, will provide packets of information on Bob Winokur (NOAA/NESDIS) on April serious errors were made affecting the how to work with the DB system. 8, 1999. It was noted that NPP meets the agricultural community. requirements that were established for Afternoon Plenary Session, Part 1 EOS-2 and EOS-3 at last year’s ESE/EOS Jim Dodge (NASA Headquarters) (Earth Science Enterprise/EOS meeting in Easton, Maryland. Competitive described “EOS Direct Broadcast (DB) Status, continued) phase B studies are now under way for Activities.” MODIS will be the first EOS CrIS and VIIRS. Selections are to be made First speaker of the afternoon was Bob source of DB broadcasts (NOAA’s AVHRR for CrIS by August 1999 and for VIIRS by Murphy (Goddard Space Flight Center) is the prototype). DB is now available for mid 2000. (Ed. Note: ITT, Fort Wayne, on the subject of “NPOESS Preparatory both Terra and the PM satellite, but only Indiana, was selected to provide CrIS.). A Project (NPP).” (NPP is a joint NASA/ MODIS will use it on Terra, whereas all phase B study for ATMS is to begin in the DoD/NOAA activity.) Murphy began his PM instruments will use it. The DB data fall of 1999 with completion due in ten talk with a listing of the NPP mission are expected to be most useful for regional months. Contractors for these instruments objectives, which are to provide NASA systems, considering the ~$0.5 M cost of are to provide both the sensors and the with continuous global-change parameters the receiving system. Data are down- algorithms. Operational Algorithm Teams after Terra and EOS PM, and to provide loaded using an X-band link. DB data will (OATs) will provide governmental NPOESS with pseudo-operational risk- serve as primary sources for flight oversight. reduction demonstration and validation operations devoted to calibration/ for three of the four critical NPOESS validation activities. They will assure that An Integrated Operational Requirements instruments and algorithms. The three the aircraft get to the right places at the Document (IORD) sets the requirements instruments in question are Visible right times. for the sensors based on an interagency Infrared Imaging Radiometer Suite study. (VIIRS), Cross-Track Infrared Sounder About 100 users for DB have been (CrIS), and Advanced Technology NASA requirements have been merged identified internationally. (He listed the Microwave Sounder (ATMS). into the IPO requirements. The OATs likely sites.) Any Landsat site could use provide science guidance to the IPO on DB by performing a simple conversion of Murphy listed the key NPP personnel, led specific measurement suites. VIIRS their equipment. A 4-m program track at Goddard by Raynor Taylor as Project imaging requirements were based on antenna system is required. Data can go Formulation Manager. The launch MODIS but have been updated. CrIS directly into the ground-computer hard readiness date is set at 2005, and a five- extends the measurement series of AIRS, drive. Level 1B data and product images year mission life is planned. The orbit will and ATMS extends the measurement would be available in just one hour. (This be at 833 km with a sun-synchronous series of AMSU-A and HSB. capability is good for locating storms, 10:30 a.m. descending node. Data will go volcano plumes, and fires.) directly to polar ground stations, bypass- VIIRS could be several instruments ing TDRSS. There will also be direct making up the “suite.” There is a require- Dodge showed the coverage to be broadcast (DB) of real-time data. ment (from the Air Force) that VIIRS have provided by the three U.S. stations and a constant ground resolution of about 400 noted that there is no station planned for NPP will be the primary source for 13 of meters. It should match MODIS for Sioux Falls, although it would be feasible the 24 EOS measurements in the period selected long-term measurements such as to have one. The Australian CSIRO will between EOS PM and NPOESS (planned SST and land-cover determinations. It also have regional coverage with three for 2009). NPP will contribute to NPOESS omits ocean-fluorescence determinations. sites. Dodge and Michael King are through sensor risk reduction, ground- VIIRS resolution for surface albedo is working to establish DB stations for system risk reduction, and early user enhanced from 4 km to 1 km, and accu- Africa. evaluation of the NPOESS algorithms. racy is increased from 0.05 to 0.03. Land-

9 THE EARTH OBSERVER

surface temperature resolution is also Reprocessing will go back to at least the there was high predictability of precipita- enhanced from 4 km to 1 km. earliest satellite observations, but this tion. Predictability was greatest in wet/ cannot be done with existing computer dry transition zones. It was not good in Bob Atlas (Goddard Space Flight Center, capabilities. very wet or very dry regions. acting head of the Data Assimilation Office [DAO]) gave an update on “Im- The SSM/I product is used for sea ice. Koster made runs with 16 different provements in Data Assimilation for Terra Ground temperatures are represented by ensembles and found that response of soil and PM-1.” He said that the mission of the the 2-meter product. Half-degree resolu- moisture to initial conditions was very DAO is to provide a Data Assimilation tion is not used for long-term climate strong in some regions but not in others. System (DAS ) that makes the best use of prediction. Stratospheric chemistry is His overall conclusions were that land- satellite data. He listed the differences being assimilated offline, and there are no surface moisture initialization can between the DAO’s operations and plans to do chemistry directly with the contribute to the predictability of precipi- Numerical Weather Prediction (NWP): GEOS model. tation. Predictability is best in transition DAO produces more diagnostics; DAO’s regions. Large evaporation is needed to model starts with the surface of land and Afternoon Plenary Session, Part 2, have an effect, and this is typical of sea and extends through the stratosphere Global Water Cycle summer-time conditions. Roger Barry to the mesosphere; the DAO’s time scale is commented that Koster’s work points to finer; the DAO has more data types; and This session was chaired by Eric Wood, the need for in situ measurements, and the final DAO result is an analysis, not a NASA Headquarters. Eric Wood said that the Soil Conservation forecast. Additional DAO products are Service is currently doing more of this. parallel analyses, offline analyses, and Randall Koster (Goddard Space Flight Wood added that the Europeans are R&D products. The GEOS-2 model now Center) gave some insights into “The considering making satellite-based soil- being developed is superior to GEOS-1 in Influence of the Land Surface on the moisture measurements. representations of zonal winds and Predictability of Precipitation at Seasonal temperatures plus sea-level pressure. It is Time Scales.” Koster’s work is in connec- Steve Running (University of Montana) also superior in its representation of tion with NASA’s Seasonal-to-Interannual addressed “Cold Season Processes and integrated eddy fluxes. Prediction Project (NSIPP). The question Links to Continental Land Surface to be resolved is: “How predictable is Processes.” Running led off with an Half-degree resolution has been shown to precipitation?” Related to that question is animation showing day-to-day variations bring out fronts sharply even in situations the question: “How predictable is soil in Net Primary Production (NPP) across where they do not appear in the NCEP or moisture?” Another question is: “If soil the continental United States that followed ECMWF analyses. moisture can be predicted, how useful is the west-to-east movements of the air that prediction for the prediction of masses, illustrating tight daily coupling of Atlas said that the Physical Space Statisti- precipitation?” the biosphere with the atmosphere. cal Analysis Scheme (PSAS) permits direct Another animation of the daily global implementation in physical space. He Koster has a model called MOSAIC. Using surface freeze/thaw cycle clearly illus- listed the innovative features of PSAS. The the model for simulation, a correlation trated both synoptic and seasonal dynam- NASA/NCAR model now being devel- was found between observed precipitation ics in the areal extent of frozen land oped has the goal of providing global data and SST. surface. This suggests the valuable assimilation of both physical and chemical contributions that a high-repeat-cycle SAR features of the atmosphere. The model is Koster chose to address the question of system could make to land-surface capable of bringing out fronts in the isobar how much the prediction of precipitation monitoring. At high latitudes the impact patterns and also wind fields. DAS will is affected by the prediction of soil of regular monitoring of freeze/thaw and produce 1o × 1o × 36-pressure-level moisture. In a run called ensemble 1 he snow cover is of major importance to assimilated products. used prescribed SST, and in ensemble 2 he carbon- and hydrologic-cycle processes as used prescribed SST plus land-surface well as weather forecasting. The model runs with 1-degree resolution moisture. He found that with both SST in near real time—just one day late. and land-surface moisture prescribed, Running went on to make a presentation

10 July/August 1999 • Vol. 11 No. 4

on a proposed mission dubbed “Cold that river flow is a major component of the First speaker of the morning was Bob Land Processes Research Mission (EX-7).” global water cycle and that stream flow is Bindschadler (Goddard Space Flight According to Running, high latitudes are important, but is not well observed Center) giving “New Understandings of where land-surface monitoring is most worldwide. Monitoring in the U.S. is Ice Sheet Dynamics.” Bindschadler said needed. The earliest signs of global adequate, but it is not so good globally. that he would give “a whirlwind tour of warming are expected to be found at high The number of reporting stations world- ice dynamics research in the 90s.” His latitudes, but this is where the ground wide peaked in 1970 and then dropped focus was on ice streams and principally stations are rarest due to sparse popula- rapidly. Another problem is the large lag those of West Antarctica. Satellite sensing tion. Freeze/thaw and snow-cover-change in getting the data into databases at the has revolutionized the study of ice sheets. observations are needed because both the appropriate data-processing facilities. ecosystem and land-surface energy Some countries are unwilling to dissemi- He showed a reconstruction of Antarctica exchange respond sharply to these nate their data. at the time of the last glacial maximum transitions. Since so much of the land is (20,000 years ago). In its current form, the subjected to cloud cover and seasonal Lettenmaier offered two scenarios for his West Antarctic sheet has lost about two- darkness, it is natural to think of using mission. Scenario A uses radar altimetry, thirds of its mass. The significance of such radar to detect surface changes. Running but will not get river widths. Overpass changes lies in their possible effect on showed an example in which NSCAT data times must be on the order of three days to global sea level. He showed an illustration picked up the freeze/thaw cycle. This was capture important surface water dynam- of the rise in sea level around the world if due to the change in dielectric constant of ics—e.g., TOPEX/Poseidon. the sheet were to melt completely and the surface that occurs as the water noted that one-third to one-half of the changes phase in going from freeze to Scenario B calls for an innovative low-cost world population lives within the “coastal thaw. radar altimeter with doppler processing. zone.” The cost to the U.S. of a one-meter Over land it is possible to take advantage rise is estimated to run from about $270 B For an EX-7 baseline mission there would of fixed land topography. A large antenna to about $475 B. be two SARs, one in L band for freeze/ will be needed to get the cross-track width Early milestones in ice-sheet research go thaw detection, and one in X band for of the rivers. River widths should be back to the 1970s when airborne radar snow-melt detection. Thus there would be estimated to 10-m accuracy. There should revealed the existence of cravassed dual frequencies. The radars could be on be 5-cm rms altimetry. Rivers would have margins of the ice streams. AVHRR was one platform or two smaller platforms to be >250-m wide. River crossings should the first satellite-sensor system to map ice- flown in formation. They would give good be repeatable to ±125 m cross-track. SAR stream locations. transition-time determinations. The could be used to determine areal extent freeze/thaw cycle could be followed on a changes and thus volume changes in Bindschadler discussed the ice-stream daily-to-seasonal basis. This suggests the lakes. characteristics: width 50 to 80 km, length valuable contributions that a SAR system 300 to 400 km, thickness 1000 m, speed 100 could make. At high latitudes the impact Eric Wood noted that there has been some to 700 m/yr. Fast motions are explained of freeze/thaw and snow cover is of major research in the Amazon for measuring by the fact that the streams are underlain importance to NPP. wetlands. Dave Emmitt pointed out that, by marine sediments and supported by with 10-m lidar footprints, clouds would pressurized water so that the ice sheets Dennis Lettenmaier (University of not be much of an obstacle to the measure- float. The streams are slowed by basal and Washington) spoke on “Global Monitoring ments. lateral resistance. of Large River and Surface Water Bodies.” Lettenmaier was proposing a new mission Wednesday, June 16 Mid-decade milestones included the use Morning Plenary Session (Polar he called HYDRA-SAT for HYDrological of Landsat to map ice-stream velocity Missions, Science Results, and Data Radar Altimetry SATellite. The goal of this Products) patterns and the demonstration of shear- mission would be to monitor the water margin development. bodies in real time and thus to study the Claire Parkinson, PM-1 Project Scientist, effects of variations in water storage and chaired this session. New milestones include Radarsat interfer- discharge. The rationale for the mission is ometry measurements of velocity inland

11 THE EARTH OBSERVER

and the discovery of tributary networks mapped glacier change, but there is no atmosphere and oceans there have been feeding the ice streams. The discovery of general digital database of glacier sizes changes in pack-ice thickness and extent. non-uniform flow of ice sheets was and extent. (At the conclusion of the talk Martin showed Manabe’s model results

“stunning.” Roger Barry pointed out that there is no for CO2 doubling, leading to 10º C world map of land ice at the current warming in “autumn.” The observed Future milestones include: ICESat resolution of climate models, and Michael warmest northern land-surface tempera- elevation measurements, which will King suggested that MODIS cloud tures only occurred in the spring! Warming highlight areas undergoing change; masking could help.) was not observed in the autumn. Radarsat measurements, which will carry out a second mapping of Antarctic ice- Recognizing the above shortcomings, In observing the Arctic Oscillation, Martin stream velocities; and Landsat 7 determi- Kieffer proposed a program of space- cited investigators who find strong nations of ice-stream motion. based applications with these objectives: warming over the Asian landmass. They also find a warming trend in the Alaskan In concluding remarks, Bob said that 1. Measure all glaciers within a 5-year permafrost temperatures for the period satellite remote sensing has revolutionized period. 1983 to 1995. ice-sheet research. ICESat will make the 2. Develop a comprehensive uniform Cavalieri and others have found decreases measurements to detect change in mass image record. in pack ice of about 3 percent per decade balance. Zwally added that ICESat will 3. Establish a GIS database with these over the last 20 years for polar regions. make 0.1 mm/yr sea-level equivalent objectives: measurements. Bob stated that we don’t a. provide lateral extent and sur- SCISEX is an oceanographic project that think of ice-sheet “collapse” any more. face velocities; has been using a nuclear submarine called b. provide surface topography USS Pargo since 1993. SCISEX data show Hugh Kieffer (U.S. Geological Survey where practical; about a 2º C warming in the upper ocean Astrogeology Branch) discussed “Viewing c. determine rates of change; and since 1993 as compared to an earlier U.S.- Glaciers from ASTER.” Kieffer said that d. provide periodic global assess- Russian “Atlas” data assembly. Near- we’re here (in Vail, Colorado) because ment. surface salinity has also increased, leading glaciers shaped the valleys allowing to decreased stability in the surface layers. Interstate 70 to bring us here from Denver. Kieffer would use three types of informa- tion: imaging, SAR, and altimetry. Drew Rothrock has analyzed changes in Kieffer said that Global Land Ice Measure- Comparing SAR with imaging, Kieffer ice thickness and found that it has ments from Space (GLIMS) is an interna- said that, to first order, each is best where decreased by 1.5-to-2 meters in the Eastern tional organization. The GLIMS coordina- the other is poorest. ASTER is almost ideal Arctic. tion center is located at USGS, Flagstaff, for a glacier survey. It has high spatial and the GLIMS World Data Archive is at resolution, provides stereo, SWIR bands Martin described changes observed over NSIDC. GLIMS is much concerned with distinguish ice from snow, pointing ability the last decade that may be related to distinguishing clouds from the surface in allows more-frequent coverage of surges. changes in the Arctic Oscillation, and he polar regions. stressed that the observed changes are big Kieffer said that steps still to be taken are: compared to climatology. Remaining Kieffer pointed out that glaciers are more algorithm development, database questions are these: Will this trend “natural integrators” of climate. They implementation, and acquiring formal continue? What are the feedbacks? What represent the balance between precipita- support. A final comment, from Jim are the effects on the ecosystem? tion and temperature/insolation. Many Hansen, was that the Earth is getting glaciers including Glacier National Park warmer and that the best confirmation of Judy Curry (University of Colorado) gave are disappearing rapidly. The Alps lose this is the retreat of the glaciers. “An Overview of the FIRE Arctic Clouds about nine glaciers per year. Worldwide, Experiment.” The experiment was glaciers are disappearing at the rate of 2 or Seelye Martin (University of Washington) motivated by uncertainties in satellite 3 per week. The glacier connection to reported on “Recent Changes in the (ISCCP) cloud retrievals in the Arctic and climate is not simple. Oerlemans (1994) Arctic.” Along with changes in the also by the large discrepancies in simula-

12 July/August 1999 • Vol. 11 No. 4

tion of arctic cloud and radiation fields by phases. Plots of water-, mixed-, and pure- Tucker found that snow depth has a large climate models. ice clouds vs. time were obtained as were spatial variability. It increases slowly and rawinsonde data. then decreases rapidly. The average depth Curry pointed out that satellite retrievals in spring was 34 cm. It took eight months of Arctic cloudiness and surface character- A meteorological tower was 20-m tall from to build up and just one month to melt. At istics are hampered by many factors, top to bottom and instrumented all the any time the amount of snow cover was including the frequent presence of thin way. A “miracle” of SHEBA was that it determined by independent storms. Melt multiple cloud layers and mixed-phase worked all the time and didn’t need ponds grew ~50 cm deeper and ~2 m clouds. She then described aircraft repositioning throughout the duration of wider during SHEBA. Lateral melting was experiments conducted April to July 1997 the project. The average surface tempera- significant. During ice melt the albedo that used ER-2 and in situ (low-level) tures during SHEBA under cloudy skies decreases and spatial variability increases. aircraft overflights of surface-based were -24º C and they were -17º C under The formation of melt ponds leads to a measurements at ice station SHEBA. The clear skies. SHEBA summers were almost sharp drop in albedo. Ice permeability biggest challenge to remote sensing of 100% cloudy. rose during the melt process. clouds lies in separating the cloud signal from the bright, heterogeneous underlying Four Portable Automated Metnet (PAM) In summary, the average albedo changed surface. stations supplied by NCAR operated from 0.9 to 0.4. There was a net mass loss throughout the year, providing coverage during the SHEBA year. The ice thinning Curry believes that the FIRE Arctic Clouds from October 1997 to September 1998. A was specific to the latitude of the study. Experiment will lead to improvement of NOAA ETL cloud-sensing radar at 35 GHz The warmth of the underlying ocean cloud retrievals and determination of picked up light precipitation, recorded water was probably the cause. radiation fluxes from EOS. Active sensors doppler winds, and classified cloud types. (radar, lidar) will provide considerable Also recorded were cloud boundaries and Afternoon Plenary Session (A improvement to the satellite retrieval of bases and tops of multilayer cloud Selection of Results from Recently Arctic clouds, shedding light on their systems. Cloud thickness determinations Launched Spacecraft and Analysis complex vertical structure. (Joe Waters Results) were very accurate. Determinations of ice/ commented that MLS can provide cirrus water contents based on radars and cloud coverage that is unaffected by This session was chaired by Daniel Jacob radiometers were used to establish surface characteristics.) (Harvard University). profiles of particle characteristics. Prob- Taneil Uttal (NOAA Environmental lems arose in certain situations when there Darrel Williams (Goddard Space Flight Technology Laboratory [ETL]) described were alternating layers of ice- and liquid- Center) gave a very enthusiastic review of “Atmospheric Measurements over the particle clouds. Uttal said that her group “Landsat 7 Early On-orbit Performance SHEBA Ice Camp.” (SHEBA is the will be doing surface validation for CERES and Science Goals.” The spectacular acronym for Surface Heat Budget of the at Barrow, Alaska. launch took place on April 15, and the first Arctic Ocean.) Taneil said that her talk image came on Sunday April 18—it would complement the FIRE ice talk given Terry Tucker (U.S. Army Cold Regions included EDC! The project is now carrying by Judy Curry. SHEBA constituted the Research and Engineering Laboratory) out its normal post-launch on-orbit largest ice camp in history. A Canadian described “SHEBA Snow and Ice Studies.” independent verification plan (OIVP). The ship frozen in the ice served as the power Tucker said that the object of his research satellite is currently flying lower and station for the project, which ran for a full is to better understand the processes faster than would be normal because of a annual cycle. governing surface-energy balance, planned underflight with Landsat 5. The particularly looking at ice-albedo feed- two satellites are actually 20 minutes apart Measurements were intended to cover an back. At SHEBA the surface changed along track. Having coverage from two entire atmospheric column and go down dramatically from month to month. satellites permits 8-day repeat coverage. to 300 meters in the ocean. Surface-based Although soot from the ship’s generator sensors included lidar from ETL to look at could be detected at up to 0.8 km from the Williams reported that Landsat 7 pixel clouds. The lidar gave heights vs. time up ship, it was not a problem. locations are good to ~ 100 meters. to 8 km and discriminated cloud particle

13 THE EARTH OBSERVER

There are some performance issues: ETM+ sponsored by ESIPs and the AMSR and Pat McCormick (Hampton University) has suffered loss of scan-mirror phase TRMM science teams. gave an “Overview of PICASSO-CENA.” lock, and there is a problem with onboard The acronym spelled out is Pathfinder calibrator stability—temperature variabil- Wentz pointed out that the TRMM Instruments for Cloud and Aerosol ity is excessive. Landsat 7 offers two new antenna aperture is about 70 cm in Spaceborne Observations-Climatologie noteworthy components: image assess- diameter. The atmosphere is nearly Etendue des Nuages et des Aerosols. This ment system and long-term data acquisi- transparent to clouds at the TRMM is an ESSP mission. LITE is regarded as a tion plan (LTAP). frequencies below 30 GHz. He can correct predecessor experiment. PI is Dave “out” water vapor effects to leave about Winker (Langley Research Center [LaRC]). In summary, Williams said that everything 0.3 K error and he can correct out cloud is performing well and the system will be optical thickness to leave about 0.06 K Four scientific instruments will be flying operational by July 15. error. Surface roughness remains the on the platform in close formation with the PM-1 platform. The instruments are: a Gary Rottman (University of Colorado) primary source of error. To correct for surface wind speed, Wentz uses both the 3-channel lidar, an oxygen A-band described “The Solar Radiation and H and the V polarization signals. Winds spectrometer (ABS), an imaging infrared Climate Experiment (SORCE).” He greater than 15 m/s are not computed. radiometer (IIR), and a wide-field camera discussed the evolution of the project (WFC). Launch is to take place in 2003. going back to the selection of SOLSTICE Wentz found that SST determined by the The mission will provide aerosol and through the original EOS Announcement TRMM TMI instrument compared well cloud vertical structures and have of Opportunity (AO) process. The with the Reynolds SST that is based on improved longwave radiative fluxes. National Academy of Sciences has AVHRR and buoys. Buoy temperatures recommended that there be both a total Mission partners are as follows: LaRC is correspond to about 1 meter, and micro- and a spectral irradiance project. the lead; the French will provide the wave temperatures correspond to about 1 PROTEUS spacecraft plus the IIR instru- The TIM instrument on SORCE is ex- mm. Wentz said that sun glitter can ment; Hampton University leads the pected to measure Total Solar Irradiance interfere with TRMM SSTs. Aerosols are algorithm implementation plus the (TSI) with an accuracy of 100 ppm. Early not a problem for microwave retrievals educational outreach; and Ball will data have shown than TSI varies by about but they are a big problem for infrared. develop the lidar, the ABS, and the WFC. 0.1%. There has been a spread in measure- TMI doesn’t give SST in rain, and AVHRR ments of TSI between previous instru- doesn’t give SST in the presence of clouds. Science objectives include developing a ments of about 0.5%. SIM on SORCE will global suite of measurements to give measure solar spectral irradiance. The A 3-day map of SST from TRMM brings observations of direct aerosol forcing. overall objective of SORCE is to determine out the circulation features of the ocean how our variable Sun affects Earth’s such as upwelling regions. In a time series PICASSO-CENA is to be co-manifested atmosphere and climate. of Hurricane Bonnie, areas of cooling with CloudSat for a Delta II launch. It is to indicated the hurricane’s strength. fly in formation with EOS PM at ± 60 The Laboratory for Astronomy and Solar seconds for cloud observations. The lidar Physics (LASP) (operating in the PI mode) TMI has a calibration problem, as indi- footprint will be have a 70.5-m diameter. will manage the entire SORCE program, cated by a 10-degree offset from cold Estimates of direct aerosol forcing will including the building of the four instru- space. Apparently, the problem occurred make use of MODIS, CERES, and AIRS ments: TIM, SIM, SOLSTICE, and XPS. because a coating was sanded and should data. Orbital Sciences is to provide the space- not have been! craft and launch (set for July 2002). PICASSO-CENA will be able to handle SORCE is to operate for five years at 600- In the future, ADEOS 2 with SeaWinds multilayered clouds—AMSR-E data will km orbital altitude. will provide both speed and directions of fill in for optically thick low-level clouds. the winds. AMSR with a 2-m antenna will Frank Wentz (Remote Sensing Systems) have improved area resolution. Wentz “The Latest TOMS Results” were reported reported on “Early SST Results from expects to see 0.2 K accuracy with 3-day by P. K. Bhartia (Goddard Space Flight TRMM.” He said that this work is averaging. Center) who started by showing the ozone

14 July/August 1999 • Vol. 11 No. 4

hole as determined from TOMS for of cloud condensation nuclei; treat water clouds, radiative fluxes, and heating. The September 30, 1998. uptake; and treat cloud/aerosol interac- link between cloud physics and cloud tions. optics is missing, and this project is trying TOMS primary products are reflectivity to bridge this gap. (380/340 nm), total column ozone, and Ghan then discussed MIRAGE—the volcanic columnar SO2. Derived products Model for Integrated Research in Atmo- CloudSat is a multi-sensor mission to are tropospheric column ozone, aerosol sphere Global Exchanges. The model has quantatively evaluate the representation optical thickness and back scatter coeffi- four modes of internally mixed aerosols. of clouds and cloud processes in global cient, and surface UV flux. Chemical processes in the model are models, and is co-manifested with relatively simple. There is some agree- PICASSO-CENA for a March 2003 launch. Bhartia noted that surface UV reflectivity ment, but not everywhere, between It requires tight formation with PICASSO- for July, climatologically, varies from 4% to MIRAGE and AVHRR on global optical CENA (~ 30 sec). It will carry an advanced 8% percent over the globe, with little depths. (He used data for August 1994.) millimeter cloud-profiling radar that can difference between land and ocean. The He has found that anthropogenic sulfur see through thin and thick clouds plus TOMS ozone record for 1980 to 1999 is in forcing is greatest at northern precipitation. It will measure both Liquid good agreement with ground measure- midlatitudes. Water Content (LWC) and Ice Water ments. The spring maximum of ozone at Content (IWC) profiles. The radar will midlatitudes is about the same as it was in In future work with MIRAGE, Ghan will operate at 94 GHz with 500-m vertical 1979. Northern hemisphere polar ozone be looking at additional sensitivity resolution. There will also be an oxygen A- has been decreasing until the last two experiments, correcting relative humidity band spectrometer to determine atmo- years. Comparing 2-D model results for bias, correcting droplet number bias, and spheric scattering. It will serve as a total ozone in the south polar region taking into account the effects of biomass particle sounder to give profiles of both shows fairly good agreement. A 3-D burning. cloud and aerosol layers. Combining the chemistry and transport model also shows data from the radar (active) and the Graeme Stephens (Colorado State good agreement with observations. Shifts spectrometer (passive) will give determi- University) presented “CloudSat Science in the region of minimum ozone show the nations of both the liquid water path and Objectives.” CloudSat is a mission of effects of El Niño and biomass burning. the particle sizes. opportunity that will provide “A new dimension in the global observations of Bhartia reported on ozone studies with CloudSat validation makes use of multi- clouds.” This is a USAF partnering TOMS. Sulfate effects on UV reflectivity agency, multi-platform internationally mission with DOE’s ARM heavily are different from those of absorbing integrated programs. Finally, Stephens involved with the validation aspects. aerosols such as smoke and dust. The said that CloudSat “is the opportunity to effects of aerosol attenuation on UV flux at Stephens stressed that cloud feedback is test new ideas that will lead to new and the surface can be as much as 50%. one of the largest areas of uncertainty in important information.” climate prediction. He showed a feedback The Triana mission is to have an EPIC loop from circulation to cloud properties The final speaker of the second day of the camera with TOMS and MODIS channels. to heating to circulation. IWG meeting was Chris Bretherton An image will be available once each hour (University of Washington) on “Current with 8-km resolution. The project includes numerical weather Understanding of Stratus Clouds.” prediction (NWP) and field observations. Bretherton discussed the effects of cold Steve Ghan (Battelle Pacific Northwest Stephens noted that progress in modeling advection vs. warm advection. Warm Laboratory) reviewed the “Use of Satellite is now surpassing the ability to validate advection can lead to fog. Many types of Measurements to Evaluate Physically- the models. boundary-layer clouds are observed. He Based Models of Aerosol Radiative showed a chart of annual stratus cloud Forcing.” He began by presenting the Cloud processes are fast. CloudSat is to amounts over the globe and another chart requirements to reduce the uncertainties: provide the information needed to of boundary-layer cumulus cloud treat all radiatively important aerosol improve the physics that is used in models amounts. particles; treat all important components for NWP. It is critical to have profiles of

15 THE EARTH OBSERVER

Radiative effects of boundary-layer stratus is more heterogeneous than the models DeFries gave examples of potential include albedo enhancement and cooling can use. She went on to give the U. of improvements for the continuous field of the atmosphere. Where there is more Maryland approach to meeting this method and pointed out its applicability sun, stratus cools the surface; where there problem. to biogeochemical models such as Steve is less sun (as in polar regions), stratus Running’s BIOME-BGC and CASA. She warms the surface. The presence of As background she noted that a vegeta- showed how the continuous field can be boundary-layer clouds is correlated with tion map from AVHRR shows thirteen used to derive NPP from Running’s lower tropospheric stability. Many different vegetation classes, but most model. In CASA, NPP can be allocated to processes are at work in cloud-topped pixels in reality have mixtures of vegeta- wood, leaf, and root. (Running com- boundary layers. tion types. There are many disagreements mented that this approach could improve among models as to the vegetation types. GCM representations.) Bretherton reviewed problems with Maps are static and do not respond to models. There are differences of factors of changes. The talk by Ranga Myneni (Boston two between ISCCP data and CCM3 University) was concerned with “Global cloud-cover fractions. The CCM3 model In the Maryland approach each pixel has Products of Leaf Area and Absorbed optimizes the land boundary layer, but proportional coverage for each vegetation Photosynthetic Radiation of Vegetation does not take all conditions into account type. Leaf forms, leaf type, and leaf from Terra.” Myneni said that his group over water. Coupled models in particular longevity are described in continuous does algorithm development. His purpose have problems. The 26o C isotherm goes fields for each pixel. DeFries uses a linear- in this talk was to inform the IWG too far eastward in the Pacific. In an mixture model approach to set propor- members about his group’s activities experiment wherein aircraft followed a tions for each type. Prototypes for regarding LAI and FPAR data, both from column of air, 2-D eddy-resolving models continuous fields require global multi- MODIS (1-km daily coverage) and MISR tended to agree with observations. spectral, multitemporal data, which (16-day coverage). His algorithm assumes currently come from AVHRR. The model six structural vegetation types. Important process uncertainties include: includes percentages for three growth entrainment—this is not well agreed forms: woody, herbaceous, and bare He pointed out that determinations of LAI upon; 3-D radiative transfer is not well ground. Leaf longevity percentages are can be ambiguous when they are based on reported; microphysics is inadequate, and given for evergreen and for deciduous pairing of NIR and red signals. However, aerosol feedbacks need better representa- trees. Leaf type percentages are for with more bands or look angles the tion. needleleaf and broadleaf. uncertainties can be resolved. He has an algorithm that gives the distribution of Finally, there is a need to know more DeFries said that her approach has LAI with specified uncertainties. The about aerosols and drop radii over the applications to the Kyoto treaty where it is addition of MISR angular measurements globe and their relation to shallow cloud necessary to establish carbon stocks for a improves the retrieval. Myneni showed precipitation. baseline year. It is necessary to set the some SeaWiFs data that indicated the spatial extent of forests despite the capability of his method. He noted that Thursday, June 17 Morning Plenary Session (Terrestrial different definitions. Tucker’s values of NDVI have paralleled Carbon Science Program) Hansen’s ground temperatures for 18 Following her discussion of the method, years of springtime data. He ended by This session was chaired by Steve she gave several examples of its applica- showing the LAI/FPAR validation sites Running (University of Montana) tion. Examples included: percent tree organized by biome. cover globally, independent of forest type, Ruth Defries (University of Maryland) and a comparison of this result with a The scheduled talk by C. David Keeling described “Deriving Continuous Fields of Landsat image—it looked good; an area of (Scripps Institution of Oceanography) Global Vegetation Properties and Imple- dramatic deforestation where there was “Evidence of Interannual Variability in mentation for Biospheric Modeling.” fair comparison with Landsat—the Terrestrial Carbon Fluxes Deduced from

Defries said that biospheric models need comparison was made difficult because Atmospheric Isotopes of CO2” was given accurate land cover, and actual land cover she used 8-km AVHRR data. by Steve Running from Keeling’s notes as

16 July/August 1999 • Vol. 11 No. 4

Keeling was unable to attend the meeting. uptake early in El Niño periods followed tion and Biospheric Monitoring.”

Keeling has looked at global trends in CO2 by losses. Ocean uptake followed by high Hollinger said that he is the director of the fluxes from 1978 to the present. Measure- tropical biomass burning may cause this AmeriFlux project. As indicated on the ments are based on 13C/12C ratios. pattern. Web, AmeriFlux is the Carbon Dioxide Flux Measurement Network of North, Keeling’s look at the fluxes for 20 years vs. Schimel’s key finding is that a warming South, and Central America. It is a latitude has shown that latitudes above trend in high latitudes leads to increases in program of long-term CO2 flux measure- 47o are biosphere sources of carbon; primary productivity. He remarked that ments using the eddy-covariance tech- latitudes from 23o to 47o N are a biosphere the temperature sensitivity of respiration nique. Measurements come from different sink; and latitudes between 23o S and 23o is greater than that of primary productiv- ecosystems and include carbon dioxide, N are a tremendous biosphere source. ity. The temperature changes can trigger water vapor, and energy exchange. Over the last five years the global bio- ecosystem responses on multiple lagged AmeriFlux assesses fluxes at spatial scales spheric flux has risen sharply. time scales. The temporal structure of the up to square kilometers in area, continu- lags is consistent with feedbacks through ously. Keeling has data for La Selva, Costa Rica, the nitrogen cycle. a tropical zone that is a CO2 source. He FLUXNET is a “partnership of partner- reports on a flux increase vs. tree-growth Schimel said that there are challenges for ships,” formed by linking existing sites anomaly, wherein a growth reduction the remote-sensing community: and networks of which AmeriFlux is one. implies an increase in CO2 respiration. He Accordingly, it “provides researchers also finds that increases in biospheric flux • inability to directly observe from space access to consistent and integrated correlate with temperature rises, but the respiration or nitrogen-cycling measurements of carbon dioxide, water flux increases when the rain decreases. changes; vapor, and energy fluxes and associated Keeling’s summation of this research is • detection of year-to-year changes in site vegetation, edaphic, hydrologic, and that the tropics are the key source of the phenology is possible—MODIS will meteorological characteristics. Fluxes and global CO2 rise. help, but simple inferences about links ancillary information are unified into between climate and observed consistent, quality-assured, documented, Dave Schimel (NCAR) reported on the phenology may be unwarranted readily accessible data sets via the World

“Interannual Variability of Terrestrial CO2 because of lag or carry-over effects; Wide Web.” Exchange: A New Inverse Approach.” He • large residual fluxes may be associated showed records of continuous fluxes from with year-to-year changes in land-use Hollinger gave the advantages of the 1980 to 1995 with the trends removed and practices. eddy-covariance technique: pointed out the influences of the El Niño and the El Chichón and Mt. Pinatubo Other comments were that the effect of • minimal impact—simply requires volcanoes. temperature changes does depend on the putting up a tower; time in the growing season, and that the • high time resolution; The standard inverse modeling approach nitrogen cycle may not be the sole reason • gives net exchange; for CO2 fluxes gives estimates of the for the lags in CO2 flux release after the • provides area averaging; magnitudes and spatial distribution of the initial temperature response. In fact, in the • operates automatically; and fluxes. Schimel’s new k-model approach k-model, the principal lagged effects arise • some ability to partition flux into soil looks only at interannual variability. The through plant metabolism of nitrogen respiration and canopy uptake output gives the global relation between (which fertilizer studies show to persist changes in temperatures and changes in for several years) and not from the Hollinger gave several examples of net

CO2 fluxes. The model takes into account ‘ecosystem’ nitrogen cycle. ecosystem exchange (NEE) measurements. the role of nitrogen uptake in plant Some of these were hourly measurements growth. Dave Hollinger (University of Montana) for July and others were daily measure- and Steve Running (University of ments. They indicated a peak summer Residuals from the model of ecosystem Montana) addressed “The Role of uptake. physiology (k-model) suggest anomalous FLUXNET and AmeriFlux in EOS Valida-

17 THE EARTH OBSERVER

Characteristic scales of tower measure- ments are 104 to 106 m2 . A large gap at 1010 SOlar Radiation and Climate to 1012 m2 is related to political boundaries, and this is where remote sensing could Experiment (SORCE) Science Team bridge the gap. Meeting

FLUXNET has both a science component — Robert F. Cahalan ([email protected]), NASA Goddard Space Flight and a data and information system (DIS) Center, Greenbelt, MD. component. The science component sees to the intercalibration of the flux measure- ment sites, and the DIS component maintains a long-term archive at the Oak Gary Rottman (U. Colorado-LASP), Thomas Woods (U. Colorado), George Lawrence (U. Ridge DAAC. EUROFLUX has all Colorado), Jerold Harder (U. Colorado), William McClintock (U. Colorado), Julius London managed forests. Some of the NASA (U. Colorado), Oran White (NCAR), Peter Fox (NCAR), Ray Roble (NCAR), Peter Pilewskie Aeronet sites are at the FLUXNET sites. (NASA/Ames), Judith Lean (NRL), Dominique Crommelynck (Inst. for Meteorology), Belgium Paul Simon (Inst. for Aeronomy, Belgium), Claus Fróhlich (PMOD, Switzerland) Steve Running said that the AmeriFlux net allows the study of geographical and seasonal patterns of carbon fluxes and will Goddard has been authorized to proceed tions from magnetograms and other solar be used for validation of MODIS land data with the implementation of SORCE by image data. including both gross primary productivity memo from the Associate Administrator of (GPP) (effectively photosynthesis) and Earth Science on August 31, 1999. A cost Goddard Project Scientist Bob Cahalan NPP. There is an attempt to get tower sites cap of $112M has been specified, which discussed dependence of Earth’s albedo to deliver their micrometeorology data includes science data processing. July 2002 and atmospheric absorption on TSI and weekly. With those data, Running could is the baselined launch readiness date. solar spectra, and Peter Pilewskie of Ames compute MODIS values of GPP and Research Center described measurements compare them with DAO values. The The first SORCE Science Team meeting of solar spectra within Earth’s atmosphere. tower data would be used to eliminate was held August 26-27 in Taos, New DAO errors. In effect, he would be able to Mexico. Gary Rottman, the SORCE PI Rottman has formally requested that test the MODIS algorithms independent of from University of Colorado-LASP gave Crommelynck, Frohlich, Lean and the DAO. an overview of the mission, followed by Pilewskie be added to the SORCE Science presentations by each of the LASP Team which, together with previous Many of the towers will have soil moisture instrument scientists - George Lawrence members from SOLSTICE and TSIM, in addition to their regular data. There are (TIM), Jerry Harder (SIM), Bill McClintock brings the team to 13 members, 10 from five ongoing studies of the possibility of (SOLSTICE), and Tom Woods (XPS). the USA, 2 from Belgium and 1 from making doppler lidar measurements of Dominique Crommelynck of the Royal Switzerland.

CO2 flux from space. NASA is now Met. Inst of Belgium described instrument considering the technology that would be intercomparison efforts, and Claus A series of Space Shuttle flights, approxi- necessary to bring down the costs of the Fróelich of PMOD, Switzerland discussed mately one per year beginning in 2000, is tower sites and thus expand their cover- the past record of TSI (Total Solar Irradi- being explored to intercompare TSI age. ance). measurements by TIM with similar ones by ACRIM and SOLCON, to narrow the Final speaker of the session was Chris Judith Lean of NRL talked about detection uncertainty in absolute calibration of TSI. Field (Carnegie Institute of Washington) of long-term solar variations, and implica- A working group is being formed to reporting on the “Integration of the Global tions for SORCE. Dick White and Peter handle the implementation and science Carbon Cycle and Challenges for Imple- Fox of NCAR described algorithms for objectives of the intercomparison, orga- menting Kyoto Protocols.” Field outlined computing TSI and solar spectral varia- nized by Bill Ochs and Bob Cahalan of Goddard. (Continued on page 35)

18 July/August 1999 • Vol. 11 No. 4

Why are there no clear standards for SAR CEOS Working Group on Calibration image quality at the user level? and Validation Meeting on Digital In the course of the papers and discussion a number of issues were raised that Elevation Models and Terrain suggest action at the level of the CEOS Plenary. One pressing problem resulting Parameters from the acquisition of global data sets, such as from SRTM, is how to obtain a — Ian Dowman ([email protected], University College London) — Jeff Morisette ([email protected], University of Maryland) well-distributed coherent set of ground — Chris Justice ([email protected], University of Virginia) control points. It was also noted that more — Alan Belward ([email protected], Joint research Centre, European Commission) international collaboration was required in the area of data fusion and synergistic use of sensors, the development of test sites and control data, and on an agreed Introduction Synthetic Aperture Radar (SAR) data. statement of sources of error and quality Andy Sowter discussed the automatic requirements. On May 26 and 27, 1999 the Committee on generation of DEMs from Stereo SAR and Earth Observing Satellites’ (CEOS) noted that a key issue is to produce an A number of actions were agreed for the Working Group on Calibration and affordable product that satisfies user Terrain Mapping subgroup to take Validation held a meeting at University needs and minimizes use of ground forward: College, London. The WGCV provides an control. He also stressed the need for international forum for calibration and calibration/validation of orbit data. • prepare recommendations for the validation information exchange, co- Laurent Polidori gave a paper on accuracy establishment of a global GCP ordination, and co-operative activities. considerations for DEMs over equatorial network; The working group promotes the ex- forests where well-planned collaborative • consider how orbit validation could be change of technical information and projects, which make use of all available developed; documentation, joint experiments, and the information, can give a very wide range of • update current status of sensors; sharing of facilities, expertise, and information. Christian Staetter empha- • produce a statement of DEMs avail- resources among its members as appropri- sized this point in discussing high fidelity able; and ate. This pooling of international effort DEM generation using PAN and multi- • produce a DEM requirements docu- helps the CEOS WGCV to promote spectral image data. Wolfgang Noack ment with a science rationale, taking standards and allows all to benefit from reported on the use of X-SAR on SRTM into account the output from SRTM. hard-won experience. It helps avoid and gave a number of examples of the duplication and overlap, while encourag- quality that can be obtained and how Terrain Parameters ing synergy. cartographic techniques such as hill- shading can aid visualization. Other The focus of the May meeting was on The review and exchange of calibration papers discussed DEMs for hydrological production and validation of both Digital and validation information for terrain modeling (Alec Walker), the comparison Elevation Models (DEMs) and terrain parameters covered a broad spectrum of of slope estimates from low resolution parameters, as derived from spaceborne issues, ranging from reviewing established DEMs (Nick Drake), and segmentation of sensors. These two major components are land cover accuracy assessment results multitemporal SAR (Douglas Corr). A summarized below. For more information (Alan Belward on IGBP landcover general paper by Polidori asked a number on the meeting see http://wgcv.ceos.org. validation and John Taylor’s results from of fundamental questions on the use of the MARS project), to current plans for DEMs SAR data: What is the most useful new validation of higher order biophysical research needed for SAR? Can we trust products (EOS/MODIS validation plans The DEM sessions covered a variety of the technical literature concerning the by Chris Justice and Jeff Morisette and topics, but with an emphasis on using operational status of SAR algorithms? validation activities in Australia presented

19 THE EARTH OBSERVER

by Fred Prata). In addition to specific • improving access to validation data related international coordination such as validation activities there were presenta- sets. the Global NPP activity of the Terrestrial tions related to planned missions Observation Panel for Climate (TOPC). (IKONOS by Nick Kladian and Envisat by There is an opportunity to use the CEOS Yves-Louis Desons) and sampling GOFC project as a test-bed for developing (Note: Following the meeting Jeffrey strategies with various global tessellations improved land product validation Privette, EOS MODIS Land Validation (Tim Richards). The Global Observations coordination. This would not only Program, NASA Goddard Space Flight of Forest Cover (GOFC), presented by contribute to the credence of the GOFC Center, and Stefan Dech, German Aero- Frank Ahern in the context of the Inte- products but also allow the WGCV to space Center (DLR), German Remote grated Global Observing Strategy (IGOS), assess gaps, overlap, and redundancy in Sensing Data Center (DFD), emphasized the importance of the global observation requirements arising from the Oberpfaffenhofen, agreed to co-chair the perspective in thematic product valida- validation implications of the developing new WGCV sub-group. NASA and DLR tion. CEOS GOFC provides an operational IGOS. This opportunity could be ad- have kindly agreed to support them in pilot activity for the provision of space- dressed by a CEOS calibration/validation their roles as joint co-chairs. ) borne data for various aspects of forest initiative on product validation for cover monitoring (http://www.gofc.org/ selected GOFC products, e.g.: WGCV has made important contributions gofc). One important focus for the GOFC to the process of international co-opera- project is to secure the provision of data • Leaf Area Index (LAI); tion ensuring long-term confidence in the sets needed to quantify forest cover extent • Fraction Absorbed Photosynthetically accuracy and quality of Earth Observation and rates of change to address outstand- Active Radiation (fAPAR); data and products. Users benefit from ing questions of the global carbon cycle. • Active Fire, Burned Area; better quality data, better documentation • Land Cover Type and Characteristics and sustained commitments to provide Through the presentations and related (continuous fields, land cover change); calibration updates. Developments such as discussions on terrain parameters, the • Forest Biomass; and IGOS (and the GOFC pilot project) raise WGCV reached the conclusions that • Net Primary Productivity. the stakes even higher. A forum for WGCV should initiate a Validation Sub- progress review, for verifying the quality Group within CEOS WGCV, which would This would build on current and planned of terrestrial products, for proposals for focus on validation of land products. The land product validation initiatives. future developments and to optimize primary goals of the sub-group would exploitation of the products is needed, and include: Spring 2000 was set as a target date for a meetings such as that proposed for spring first meeting of the Validation Sub-group. 2000 represent excellent opportunities to • promoting the quantification and The focus will be on validation of prod- strengthen the international co-operation characterization of satellite land ucts associated with GOFC. The meeting on which global scale activities so clearly product accuracy; would include a generic session of invited call. • sharing land product validation past papers on validation concepts/approaches experience and lessons learned; and presentations on the GOFC product Conclusion • moving towards the generation of suite (with a focus on standardizing ‘standardized products with known products and product validation). It was As the land remote sensing community accuracy’ from similar sensing also recommended that the meeting starts to pay increased attention to data systems in the context of data continu- include break-out sessions to discuss: a) in quality and accuracy through product ity; situ measurement protocols, b) validation validation, an international validation • establishing relationships between like sites, and c) data access for various coordination mechanism has considerable products - e.g. V.I.s; product suites (such as the selected appeal. The space agencies and land • developing traceable in situ validation products listed above). Invitees should remote sensing community will benefit measurement standards and proto- represent the different moderate and high from sharing experience and resources cols; resolution, as well as SAR, instruments associated with the collection and sharing • coordinating international validation contributing to GOFC. Appropriate activities; and linkages would be made with other (Continued on page 38)

20 July/August 1999 • Vol. 11 No. 4

NASA EOS Validation via SAFARI 2000: Kasama

ZAMBIA Preparing the Way Nacala Mongu Lusaka MALAWI Katima Mulilo Harare — Michael King ([email protected]), NASA Goddard Space Victoria Falls MOZAMBIQUE Kasane Flight Center Okaukuejo ZIMBABWE NAMIBIA Maun — Tim Suttles ([email protected]), Raytheon ITSS Corporation Beira BOTSWANA Swakopmund Walvis Bay Windhoek Pietersburg

Gaborone Pretoria Maputo Johannesburg SWAZILAND During July 8-24, 1999, a team of U.S. and (ii) the Department of Electri-

South African scientists visited southern cal & Electronic Engineering, Masen SOUTH Durban Africa as part of preparatory work leading University of Stellenbosch, and AFRICA LESOTHO Figure 1. Map of up to the Southern Africa Regional Science (iii) met with a radio reporter from southern Africa with key Stellenbosch Initiative 2000 (SAFARI 2000). SAFARI the South African National Public locations of significance Cape Town to SAFARI 2000 marked 2000 is an international regional science Radio. initiative being developed to explore, Meetings study, and address linkages between land- were held at atmosphere processes and the relationship the Univer- of biogenic, pyrogenic, or anthropogenic sity of Cape emissions and the consequences of their Town with deposition to the functioning of the Profs. Chris biogeophysical and biogeochemical Reason and systems of southern Africa. This initiative Mathieu is being built around a number of on- Rouault. going, already funded activities by NASA, Prof. Reason the international community, and African is an oceano- nations in the southern African region. graphic The largest component of SAFARI 2000 is modeler a regional field experiment to be con- interested in ducted in southern Africa from August- the retrore- September 2000. flection of the Figure 2. Table Mountain towering over Cape Town, South Africa. Agulhas Figure 1 shows the locations of principal Current and activities during SAFARI 2000, many of its interaction with the Benguela Current described his interests in rainfall variabil- which were visited during a tour of the off the Namibian coast. Michael King ity, convective systems embedded in region in July 1999 by U.S. and South demonstrated the QuikScat Web page southern Africa, and validation of regional African scientists. showing how to obtain SeaWinds mea- atmospheric models. Rouault and his team surements from NASA’s recently launched are interested in and are currently a. Cape Town and Stellenbosch, satellite to validate the winds produced by working with TRMM data. Republic of South Africa (RSA) (July Reason’s model. Reason was also inter- 8-9) ested in sea surface temperatures, pres- A special visit was made to the University ently produced by AVHRR and to be of Stellenbosch to get an update on results During July 8-9, Michael King and Tim produced later this year by MODIS, and from South Africa’s first satellite Suttles visited (i) the Department of ocean altimetry that is produced by (SUNSAT), designed and built by students Oceanography, University of Cape Town, TOPEX/Poseidon. Mathieu Rouault at the University of Stellenbosch and

21 THE EARTH OBSERVER

launched by NASA earlier this year. On tour details for Botswana, Zimbabwe, (South African Weather Bureau), Bob the day of the visit, the University Zambia, and Namibia; and the Swap (University of Virginia and U. S. managed to acquire their first image from planned Gaborone SAFARI 2000 Coordinator of SAFARI 2000), and the satellite, and they provided a high implementation workshop. Prof. Harold Annegarn (WITS and RSA quality print of it to take back to Dr. coordinator of SAFARI 2000). Ghassem Asrar. They are having commu- • Met with the Prof. Leila Patel, Deputy nication problems with the satellite, so Vice Chancellor of WITS, along with • Discussed aircraft overflight clearance, they are only able to bring down low- senior faculty and staff. customs waivers and use of Carnet de data-rate, quick-look images of a small Passage, shipping of consumables to portion of a scene from any given orbit. If • Discussion with WITS faculty and RSA (radiosonde balloons), and they like the scene that was acquired, they students involved in SAFARI 2000, current schedule. can download the entire scene (60 × 60 including Prof. Charlie Mather km), but this can only be accomplished (Geography; image processing), Prof. Visit to Department of Transport over a 7-day period! From an educational Spike McCarthy (Geology; expert on • Meeting with Johann Bierman (Section point of view, however, this was a good Okavango Delta), and Juliao Manager, Air Transport Regulation) project for Stellenbosch. Ninety-six staff Cumbane, Coleman Dube, and Abel who is responsible for all bilaterals members and students worked on the Sakhau (students). and licensing of foreign aircraft project and 46 M.S. theses were written on operations over RSA. SUNSAT. What is lacking is onboard or • Lunch at Hofmeyer House staff club preflight calibration, which is a major with Profs. Mary Scholes (Life • Bierman expressed willingness to impediment to achieving quantitative Sciences), Helda Marks (Chemistry facilitate aircraft flight operations of science from the mission. and Deputy Dean of Science), June the NASA ER-2, Proteus, and Univer- Meeuwis and Peets Wolvaardt sity of Washington CV-580 over RSA, Finally, while in Cape Town, King and (Geography & Environmental Sci- and support of SAFARI 2000 in Suttles met with John Richards, a reporter ences, Rand Afrikaans University), Dr. general. from South Africa’s National Public Radio. Gerhard Held (Eskom TRI–research They discussed NASA’s Earth Observing labs of electric utility), and Luanne Visit to Satellite Applications Centre System, global change, ship tracks, the Otter (CSIR). (SAC), CSIR Mikomtek Earth’s radiation balance, and finally • Visited the SAC outside Pretoria, to get NASA’s interest in southern Africa (i.e., Visit to St. John’s College a better understanding of the role of SAFARI 2000). Richards taped interviews • Michael King and Tim Suttles gave this organization within RSA, which with King and Suttles separately, which lectures to the faculty and students of now includes receiving, archiving, and were subsequently broadcast throughout St. John’s College, a prestigious all- distributing data from Earth observa- South Africa. boys private high school in tion satellites throughout southern Johannesburg (see Figure 3). The b. Johannesburg and Pretoria (July lectures described EOS, global change 12-13) (land cover and land use change, ozone depletion, global warming), and During July 12-13, King and Suttles visited SAFARI 2000. Johannesburg and Pretoria, and had a busy schedule that included: Visit to U. S. Embassy of South Africa • Meeting with Atim Ogunba (Environ- Visit to the University of Witwatersrand mental Science & Technology Officer, (WITS), in Johannesburg U. S. Embassy), Luanne Otter and Ike • Met with Prof. Harold Annegarn and Lombard (scientists and international Stuart Piketh on aircraft and landing coordinators from CSIR), Ferdi van der Figure 3. Michael King, Tim Suttles, Sandy facilities in Kruger National Park and Walt (International Officer, National Heather, and Lorraine Tandt at St. John’s College Pietersburg, South Africa; inspection Research Foundation), Nico Kroese following lecture to faculty and students.

22 July/August 1999 • Vol. 11 No. 4

Africa. The SAC has a large satellite U. S. Embassy d. Maun and Okavango Delta, tracking dish, which could be used for • Met with Robert Krueger (U. S. Botswana (July 16-17) direct broadcast of Terra and PM data, Ambassador to Botswana), Scott and has in the past been used to DeLisi (Deputy Chief of Mission), and On July 16, Michael King, Tim Suttles, receive Landsat 4 and 5 TM data. The David Franz (Political-Economic Harold Annegarn, Bob Swap, and Steve Pretoria station has been used for Officer), to discuss SAFARI 2000 and Platnick departed Gaborone via the SAWB global launch support and tracking for requirements for ground-based Aerocommander 690A on a tour of NASA, ESA, CNES, DLR, and Russian measurements, data processing, and proposed SAFARI 2000 measurement Space Agency spacecraft. SAC has an airborne operations in Botswana. sites. En route to Maun, Botswana, the extensive archive of data from group overflew the Makgadikgadi Salt Meteosat (1977-1995), NOAA (1984-), U. S. Agency for International Develop- Pan of Botswana. At Maun, a meeting was Landsat MSS (1989-), Landsat TM ment (USAID) held with the assistant aircraft manager to (1989-), Spot (1989-) and ERS-1 and -2 • Met with Ed Spriggs (Director), assess the feasibility of operating the CV- (1994-). Anthony Vance (Deputy Director), 580 from this airport; though feasible, it Candace Buzzard (Project Manager, was later determined that it was too far c. Gaborone, Botswana (July 14-15) Natural Resources Management from the primary sampling sites near Project), Raymond Morton (Chief, Mongu, Zambia. A trip was made via 4- On July 14, King, Suttles, and Bob Swap Agriculture & Natural Resources), and wheel drive vehicle to the Maun tower (cf. traveled from Johannesburg to Gaborone Robert Buzzard (Program Manage- Fig. 4), which is operated by the Max via the South Africa Weather Bureau ment Specialist), to discuss SAFARI Planck Institute in Germany on real estate (SAWB) Aerocommander 690A aircraft. 2000 and ways in which USAID could owned by the University of Botswana. The group was met at the airport by a contribute immeasurably to the This is a possible site for locating an Professor of Physics at the University of success of SAFARI if they (i) sup- AERONET sunphotometer as well as Botswana (UB) and driver, and immedi- ported a data center and training additional solar radiation measuring ately transported to UB for a scheduled center for the southern Africa region equipment. seminar. King and Suttles gave a com- at UB, and (ii) supported scientists bined seminar entitled ‘New Remote from southern Africa north of South Sensing Data Coverage for Southern Africa who want to participate in Africa—The SAFARI 2000 Experience,’ SAFARI 2000 but who have no with King describing EOS and global financial support to participate. change science while Suttles followed with a description of SAFARI 2000 and • Visit with Mr. A. V. Lionjanga, current plans for southern Africa. The Permanent Secretary, Ministry of seminar was attended by 60 students and Works, Transport and Communica- faculty members, and was hosted by Prof. tions, who suggested the University of Susan Ringrose of UB. Washington might prefer to operate its aircraft from Maun, Botswana, rather Additional meetings held in Gaborone than Victoria Falls, Zimbabwe; his included: ministry approves overflight and • Visit with Mr. Masego Mphathi, operation of foreign aircraft in Director of the Department of Crop Botswana. Production and Forestry, Ministry of Agriculture. • Visit Mr. I. Muzila, Principal Hydro- logical Engineer, Department of Water • Dinner with Prof. Sharon Siverts, Vice Affairs, Ministry of Minerals, Energy, Chancellor of the University of and Water Affairs, to brief him on Botswana, and senior administrative goals of SAFARI 2000. staff of UB. Figure 4. Tower in Maun, Botswana.

23 THE EARTH OBSERVER

Over the weekend a visit was made to the extremely high amount ($283 for a and the parking area is small and would Okavango Delta, the largest inland delta single). have to be set aside for the CV-580 to have in the world, to experience the wildlife of clear rights to a regular parking space on southern Africa (Bob Swap and Tim • Phones did not work for most of the the tarmac. This is being pursued. Suttles were charged by a hippopotamus, day we were at Victoria Falls, and it Due to a tight schedule, a visit was not which fortunately turned back at the last took over an hour to use a Visa card to made to Katima Mulilo. However, it was moment). get $20 U.S. converted to Zimbabwe learned later from the airport manager of dollars; the use of phones, faxes, and e- Walvis Bay, Namibia, who is Director of all e. Victoria Falls, Zimbabwe and mail would more than likely be regional airports in Namibia, that it may nearby Zambesi River, Zambia (July difficult to use reliably in Zimbabwe. be a good operating base for the CV-580. 18-19) The Director indicated he could guarantee f. Kasane, Botswana and Katima a hangar and office space, and that the After briefly landing in Victoria Falls and Mulilo, Namibia (July 20) runway was quite long and the town very clearing customs, King, Suttles, Annegarn, nice. It is located on the Zambesi River and Platnick flew on the Aerocommander With the realization that Maun was too far across from Zambia and due south of over Victoria Falls and along the Zambesi from Mongu to effectively operate the CV- Mongu. Hence, it is also a practical place River to Mongu, the site of one of the EOS 580 during the central part of SAFARI to operate from. Unfortunately, a couple of flux towers (Jeff Privette, Goddard Space 2000 (cf. Fig. 1), and with the additional weeks after this visit, there was a separat- Flight Center, is principal investigator). difficulties in Zimbabwe associated with ist uprising in Katima Mulilo, which is The environment includes the miombo excessively high per diem rates, poor located in the Caprivi Strip of Namibia woodland of Mongu, nearby grassland, phone lines, and an unwelcoming between Botswana and Zambia, and 16 and riverbed of the nearby Zambesi. This government to foreign visitors, the SAWB people were killed. Before operating from site will be overflown quite often by the Aerocommander pilots recommended two this location, the political situation would ER-2 and the CV-580, and will be coordi- alternative locations for the CV-580 need to be re-evaluated. nated with many Terra overflights during (Kasane, Botswana and Katima Mulilo, SAFARI 2000. Many, many fires were Namibia). g. Namibia (July 20-22) observed en route to Mongu, along with a well-mixed boundary layer with a A flight was made from Victoria Falls to After leaving Kasane, the group flew to uniform, plane-parallel, haze layer Kasane to check out the airport facilities. Ondangwa, Namibia (near the Angola overlying essentially all of western Unfortunately this visit occurred on a border) to clear customs, and then flew to Zambia. Before departing Zimbabwe, a Botswana holiday and the airport man- Etosha National Park. Enroute the aircraft meeting was held with Jerry Ndlovu, ager was not available. Nevertheless overflew the Etosha Pan, the largest salt Director of the Victoria Falls airport. airport facilities appeared to be quite pan in Africa (110 × 60 km), to assess its Although Ndlovu was quite enthusiastic adequate, only 50 km or so west of uniformity as a ground calibration target about SAFARI 2000 and hosting the CV- Victoria Falls, and much cheaper to for MODIS and MISR and as a uniform 580 for two weeks during the experiment operate from than Victoria Falls. In ecosystem for making measurements of (September 2000), the group felt that addition, this is a most interesting location the bidirectional reflectance factor from Zimbabwe was not a good area to operate next to the Chobe River and Chobe the CV-580. An AERONET sunphotometer any U.S. aircraft. This opinion is based on National Park, which has the largest will be located at Namutoni on the eastern the following considerations: elephant herds on Earth. The Botswana edge of the pan during the experiment, government, including the U.S. Embassy, along with a sodar from the University of • There is a $30 (U.S.) entry visa required is more than willing to host activities in Virginia (Mike Garstang). to enter Zimbabwe, and a $20 depar- Botswana, and the landing fees and ture tax. parking fees are $50 in Botswana, com- Wilfred Versfeld, Director of the Etosha pared to $80 in Zimbabwe and $40 in Ecological Institute, met the group at • The hotel charges in Zimbabwe were Namibia. The disadvantage is that there is Okaukuejo, the largest guest camp at based on the country the guest was really no office space available at the Etosha, which means ‘Great White Place’. from, and U.S. guests were charged an airport, unlike Maun and Victoria Falls, The next day (July 21), a meeting was held

24 July/August 1999 • Vol. 11 No. 4

with Johann le Roux, who is the satellite The group then flew from Windhoek to • Terra launch and deploy (with remote sensing expert at Etosha. He has Walvis Bay, Namibia overflying the red Australia background, rather than long been collecting AVHRR 1-km data sand dunes of the Namib Desert (the Vandenberg; Reto Stockili, VAL). over northern Namibia, and showed the tallest sand dunes on Earth, at 1000 feet), seasonal variability of reflectance of the at a height of 100 to 200 feet above the • Landsat band combination illustration pan, which is much brighter in the west dunes. At Walvis Bay, a meeting was held with San Francisco, including digital (near Okaukuejo) than in the east (near with the airport manager, Mr. D. F. elevation of the mountains in the Bay Namutoni). In addition to the biomass Booysen, who was exceedingly supportive area (Jesse Allen, VAL). burning smoke that is well mixed in the of the CV-580 operating from this field. He • Landsat 7 ETM+ flyover of Cape boundary layer and which originates from is the one who pointed out the virtues of Town, RSA (Jesse Allen, VAL, based the north (Angola and western Zambia), Katima Mulilo as well, which would on data and animation work done at there is a persistent low-level wind from simplify the letter agreement from NASA the Science Visualization Studio the east during the day that picks up dust to Namibia if the UW aircraft were to [SVS]). from the pan and blows it west over the operate at two bases in Namibia as well as Atlantic Ocean. King expressed an interest Pietersburg, RSA. The airfield was quite • Africa 1 km land use classification still in setting up an AERONET site at this long, and there were adequate facilities for (Yuhong Tian and Ranga Myneni, location in the west, as well as placing Si- operation for the final component of Boston University). Chee Tsay’s surface radiation network at SAFARI 2000, which involves overflights this location. of the Namibian stratus clouds that occur • Global 1 km AVHRR imagery for both over the Benguela Current off the ocean and land (National Geographic In the afternoon, a trip was made to the Namibian coast in September. and JPL). pan itself, where it was observed that the pan is nonuniform, and has most unusual h. Johannesburg (July 23-24) • NDVI seasonal coverage of the U.S. reflectance characteristics. It was generally (left screen) and Africa (right screen) quite dark looking in the direction of the On July 23, the group flew from Walvis animated over a 1 (U.S.) and 2 (Africa) sun (no sunglint or Fresnel reflection), and Bay to Johannesburg. That evening year period, showing similarity in quite bright looking away from the sun Michael King gave an invited Prestige greening up and drying out of both (especially bright in the backscattered Lecture at the National Awards Ceremony continents in series (Mike Manyin, direction). None of the group had ever of the Associated Scientific and Technical VAL). seen any natural surface with these Societies (AS&TS) of South Africa, an characteristics, and hence collected amalgamation of 94 professional societies • 17 years of monthly NDVI averages samples of the pan and returned them to of science and engineering throughout for Africa (Jesse Allen, VAL, based on Goddard so that Jim Butler could make South Africa, consisting of some 40,000 data and animation work done at the measurements of the BRDF in his state-of- members. The lecture was entitled ‘NASA SVS). the-art diffuser laboratory. Earth Observation Programs: Past, Present, and Future.’ For this formal black • TRMM miracle swath containing The tour continued with a flight from tie and highly publicized event in RSA, Precipitation Radar (PR) data over 4 Etosha to Windhoek for a meeting at the King used the electronic theater developed hurricanes (Greg Shirah, SVS). U. S. Embassy with Harmony Caton, by Fritz Hasler, manager of the Visualiza- Third Secretary/Vice Consul and Major tion & Analysis Laboratory (VAL), at • El Niño with SST, surface topography, Nichols, Air Attaché. They were very NASA Goddard. Staff of the VAL, includ- and vector winds (Greg Shirah, SVS). supportive of SAFARI 2000 and likely to ing Jesse Allen, Barbara Summey, Mike be of much value in helping to work out Manyin, and Reto Stockili, supported this • MODIS Airborne Simulator (MAS) the aircraft overflight and basing permis- activity and prepared the following cycling of 50 bands over a biomass sions in Namibia and in helping to execute animation sequences: burning fire in Brazil, comparable to a letter agreement to operate the CV-580 in what is expected during SAFARI 2000 Namibia. • Meteosat colorized still (Merit Jentoft- in Zambia (Rob Simmon, Earth Nilsen, VAL). Observatory/GSFC DAAC).

25 THE EARTH OBSERVER

• MAS flyover of fire in Brazil (Barbara NASA Earth Science Enterprise educa- including the presence of Jesse Allen. Summey, VAL). tional materials were distributed to the Please would you pass this on to the teachers for use in the classroom. The powers-that-be that took all the trouble to This 50-minute lecture was extremely educational materials included copies of make this all happen? powerful, and had a very significant each of the EOS science theme posters (7 impact on the diverse audience of engi- in the series), along with fact sheets and It may interest you and your colleagues to neers, scientists, and spouses, and was other science brochures. The EOS posters, know that your lecture kicked off a very well received. Credit goes to Jesse first produced in 1994, are still being wonderful initiative. In other words, the Allen, Fritz Hasler, and Barbara Summey requested today over the Web at a rate of importance of it provided the occasion for for pulling this material together for 400 per day! The EOS Project Science many different people to meet who would presentation in a forceful and effective Office has just completed an educational never normally do so, and, from that manner. CD-ROM based, at least in part, on the occasion, to realize that there is so much to poster series but much expanded, with an do in this country - but altogether, and for The following day, prior to returning to accompanying teacher’s guide. This will a major common cause - namely, the use of the U.S., Michael King was invited to give be sent over to South Africa shortly. science, engineering and technology in the one more lecture, this time for 1.5 hours study and improvement of the environ- and to 500 public high school teachers After returning from the visit, Michael ment for the betterment of all mankind. from Gauteng province (particularly King received the following message from The repercussions for this country were Soweto, Johannesburg, and Pretoria). Prof. Rosemary Falcon, president of and are continuing to be many fold, and Much of the animation from the previous AS&TS, the society that hosted the event not just in the technical aspects! I hope to evening was used (cf. Figure 5), eliminat- in Johannesburg: be able to report to you more on this in the ing only the MAS flybys of Brazil as too near future. All in all, your lecture was technical. This was followed with presen- ‘I must tell you that you left a trail of indeed a major step, in many directions.’ tations on ocean processes (SST, vector enormously excited people here, many of winds, ocean color, and topography) and whom cannot wait until you get back! You In conclusion, this was a most effective land processes (especially hydrology) had such a marvelous manner of presenta- trip to southern Africa in many regards, from the EOS poster series using tion, delivered with such skill, conviction and leads to the expectation that the vugraphs, and concluded with a descrip- and obvious enjoyment. For this may I experiment being planned for southern tion of ship tracks as observed by AVHRR now thank you so very much, and also for Africa next year will be a great success. and visualized by the RC-10 camera on the all the trouble that was taken to mount the The region is poised to work with NASA NASA ER-2 aircraft. special lecture and the Electronic Theatre, and the international scientific community.

Figure 5. Michael King showing El Niño animation during lecture to 500 public school teachers in Johannesburg (photo courtesy of Antony Cowey, Editor, Information and Communications Division, Mintek.)

26 July/August 1999 • Vol. 11 No. 4

the data available to the scientific commu- nity at a single electronic location - the SHADOZ (Southern Hemisphere SHADOZ website in the Atmospheric ADditional OZonesondes): Chemistry and Dynamics Branch at Goddard: http://code916.gsfc. nasa.gov/ A New Ozonesonde Data Set for the Data_services/Shadoz/ Earth Science Community shadoz_hmpg2.html 2. Where are SHADOZ sites? — Anne M. Thompson ([email protected]), and Jacquelyn C. Witte (SMAC) NASA Goddard Space Flight Center, Atmospheric Chemistry & Dynamics Branch Stations in SHADOZ include four islands in the Pacific: Fiji, Tahiti, Galapagos and American Samoa. Two sites are at and in 1. Why and What is SHADOZ? hemisphere tropical sites have operated the Atlantic: Natal (Brazil) and Ascension balloon-borne ozone instrumentation Island. Three other sites span Africa In the past 5 years, new tropical tropo- (ozone sondes) in the past decade, but (Nairobi and Irene, near Pretoria, South spheric ozone data products have been sampling frequencies vary and opera- Africa) and the Indian Ocean (Réunion developed from TOMS and other satellites tional periods are often sporadic. When Island and Watukosek in Java, Indonesia). [Fishman and Brackett 1997; Hudson and Earth-Probe TOMS was launched in mid- Table 1 lists coordinates and the SHADOZ Thompson 1998; Ziemke et al. 1998]. Also, 1996, only four of these sites were opera- Co-Investigator responsible for each site. during this period, many global chemical- tional. Figure 1 shows the clickable map that transport models have been developed for appears on the SHADOZ website. All interpretation of satellite data and to In 1998, the Goddard Space Flight Center SHADOZ sites were operational at the predict future ozone levels in the tropo- with its Wallops Flight Facility and initiation of the project. A guiding sphere and stratosphere. NOAA/CMDL (Climate Monitoring and principle is to enhance sonde launches Diagnostics Laboratory), began a 2-year and data at facilities on a cost-share basis The lack of independent ozone profile project to collect a consistent set of with an international partner. It is hoped measurements for evaluation of satellite measurements by augmenting that wider dissemination of data and data and models is critical in some ozonesonde launch frequencies at south- interaction with users and field projects regions. At least a dozen southern ern hemisphere tropical sites and making will leverage local funding to maintain

SHADOZ Sites

The SHADOZ homepage gives technical information for each station, a clickable site map, addresses and email for Co-Investigators. Co- Investigators are responsible for original data processing and should be consulted for details of method and appropriate references to their work. San The website is: http://code916.gsfc.nasa.gov/ Data_services/Shadoz/shadoz_hmpg2.html. Cristóbal Nairobi Natal Am. Samoa Questions about data should be directed to the Ascension Watukosek datakeeper and webmaster: Jacqueyn Witte: La Tahiti Is. Fiji [email protected]. Questions about Irene Reunion SHADOZ should be directed to Anne Thompson: [email protected]

Figure 1. SHADOZ stations as they appear on the website.

27 THE EARTH OBSERVER

SHADOZ Sites Lat./Long. (deg) Co-I by Dawn Holdren and Tom Northam (SESI) at Wallops Fight Facility. Pago Pago, A. Samoa -14.23 -170.56 Samuel Oltmans (NOAA/CMDL) Papete, Tahiti -18.00 -149.00 Samuel Oltmans (NOAA/CMDL) 3. Multiple Applications for San Cristóbal, Galapagos -0.92 -89.60 Samuel Oltmans (NOAA/CMDL) SHADOZ Observation Natal, Brazil -5.42 -35.38 Volker Kirchhoff (INPE) SHADOZ observations are already being Ascension Island -7.98 -14.42 Francis Schmidlin (NASA/WFF) used by modelers and researchers Irene, South Africa -25.25 28.22 Gerrie Coetzee (SAWB) preparing satellite ozone data sets (Figure Nairobi, Kenya -1.27 36.80 Bruno Hoegger (Aer. Sta., Payerne) 3). SHADOZ is designed to meet other La Réunion -21.06 55.48 Jean Leveau (Univ. Réunion) needs of the Earth science community as well: Watukosek, Indonesia -7.50 112.60 Toshihiro Ogawa (NASDA/EORC)

Suva, Fiji -18.13 178.40 Samuel Oltmans (NOAA/CMDL) (1) Provide the first climatology of tropical ozone along the equatorial Table 1. zone for the study of atmospheric infrastructure and operations. Originally Current sampling at all stations is once- processes like the wave-one pattern in planned for the 1998-1999 period, most per-week or twice monthly, usually, but total ozone [Shiotani 1992; Kim et al. sites will contribute SHADOZ data not always, mid-week. Balloon-borne 1996]. Until SHADOZ, no comparable through 2000. ozonesondes are coupled with a meteoro- data set existed with extensive logical radiosonde for data telemetry, coverage through the Pacific, Indian Note: transmitting air pressure, air and pump and Atlantic regions. SHADOZ data sets are products of evolving temperatures, relative humidity, and research by the site Co-Investigators and ongoing community collaboration. As you work with the ozone to a ground receiving station at 1- (2) Supplement field project observations. data, please keep us and site Co-Is posted on issues second intervals. Figure 2 gives an For example, the 1998-1999 sondes at that will help us improve the value of the data. In American Samoa, Fiji and Tahiti came 2000, the 1998-1999 SHADOZ data will be example of a typical sounding from available on a compact disc. Ascension Island (8°S, 14°W), which was from the GTE/PEM -Tropics (Global taken by the US Air Force and processed Tropospheric Experiment/Pacific Exploratory Mission) experiment. The Ascension Island (8.0S, 14.4W) Irene site will launch sondes more Ozonesonde Profile for 01/27/99 frequently than usual as a SAVE Total Ozone (SBUV) = 267(49) DU (Southern African Validation for EOS) component during SAFARI-2000. 10 (3) Improve ozone profile climatologies Temperature (deg C) for TOMS, SBUV and SAGE II satellite retrievals. The current tropical data are based on a very limited number of soundings. SHADOZ will also guide 100 Ozone Partial Pressure (nbar) algorithm development for future

Pressure (mb) Pressure satellite instruments, e.g., SAGE III, QuikTOMS, OMI, SCHIAMACHY, HIRDLS, TES.

(4) Serve as a resource for scientists and educators in southern hemisphere 1000 tropical locations, some of which are -100 -50 0 50 100 150 developing more scientific research Figure 2. Typical ozonesonde profile viewable on SHADOZ website shows ozone and temperature profiles from the surface to 7 mb. Ozone is measured by the electrochemical concentration cell (ECC) method. capacity.

28 July/August 1999 • Vol. 11 No. 4

Tropo. Ozone from Ascension ozonesondes Mauritius, as the Ronald Brown continued vs TTO, July 1997-June 1998 y eastward from Africa on its way around 80 Figure 3. Ascension Island integrated tropospheric the world. column ozone (stars) with tropospheric ozone derived 5. Summary from TOMS by the 60 modified-residual method. Column-integrals in SHADOZ attempts to set an example for Dobson Units (DU). timely public distribution of much needed Adapted from Thompson and Hudson [JGR, in radiosonde and ozonesonde data, foster 40 press, 1999]. Real-time further interaction amoung modelers, maps of tropical tropospheric ozone by the satellite investigators, SHADOZ Co- modified-residual method Investigators, students and educators. By appear on http:// 20 demonstrating the value and desire for metosrv2.umd.edu/~tropo. At the same website, ozone data among the Earth science twice-monthly maps of community, SHADOZ hopes to inspire tropospheric ozone (20N- 0 20S, 1x2 degrees) for participant countries and agencies to July Sept Nov. Jan. Mar. May July 1979-1992 are available in maintain their sites for scientific research, 1997 1998 digital form. monitoring, and education. At the conclusion of the SHADOZ project, the sonde data will become part of the 4. Additional SHADOZ Data Sets - are taken from the first North America to WOUDC (World Ozone and Ultraviolet Experiments of Opportunity Africa cruise with ozonesondes. James Data Center) archive in Toronto. Johnson of the NOAA/Pacific Marine From time to time, intensive tropical Environmental Lab in Seattle added five campaigns are making data available to soundings between Cape Town and SHADOZ. The first half of 1999 added data from several special field projects. GSFC supplied some of the ozonesondes R/V R.H. Brown Cruise Track launched daily in the Maldives (5N, 73E) during the INDOEX (Indian Ocean Experiment) in January through March 1999. The data will become part of the SHADOZ archive, as will half a dozen Norfolk North sondes launched by SOWER (Strato- Atlantic spheric Ozone and Water in the Equatorial Region) at Christmas Island in the Pacific (2N, 157W) during the NASA PEM- Tropics experiment in February and March 1999. Perhaps the most unusual augmen- tation of a SHADOZ data set is from the NOAA Research Vessel Ronald H. Brown, South on which 22 sondes were launched by Atlantic SHADOZ Principal Investigator Anne Cape Town Thompson between Norfolk, Virginia, and Cape Town, South Africa, in January and February 1999 (see Figure 4 and 5). 17 January - 06 February 1999 Ozonesondes aboard ship have been launched between Europe and South Figure 4. The Atlantic segment of the R. H. Brown ship track. The solid circles show the location of each America and Africa, but SHADOZ data of the 22 ozonesonde launches.

29 THE EARTH OBSERVER

Acknowledgments

SHADOZ is supported by NASA’s ACMAP Program and the TOMS project. Day-of-flight photos Individual SHADOZ sites are also were taken on-board the supported by in-country agencies and Ronald H. Brown universities, NOAA, NASDA (Japan), research vessel during the WMO, and the Swiss Meteorological Atlantic leg of the cruise Agency. (January-February 1999).

References

Fishman, J., and V. G. Brackett, 1997: The climatological distribution of tropospheric ozone derived from satellite measurements using version 7 Total Ozone Mapping Spec- trometer and Stratospheric Aerosol and Gas Experiment data set. J. Geophys. Res., 102, 19275- 19278.

Hudson, R. D., and A. M. Thompson, 1998: Tropical tropospheric ozone (TTO) from TOMS by a modified-residual method. J. Geophys. Res., 103 , 22129-22145. (c) Kim, J-H., R. D. Hudson, and A. M. Thompson, 1996: A new method of deriving time-averaged tropospheric column ozone over the tropics using total ozone mapping spectrometer (a) (TOMS) radiances: Intercomparison and analysis using TRACE-A data. J. Geophys. Res., 101, 24317-24330.

Shiotani, M., 1992: Annual, quasi-biennial and El Niño-Southern Oscillation (ENSO) time-scale variations in Equatorial total ozone. J. Geophys. Res., 97, 7625-7634.

Thompson, A. M., and R. D. Hudson, 1999: Tropical Tropospheric Ozone (TTO) Maps from Nimbus 7 and Earth-Probe TOMS by the Modified-Residual Method: Evaluation with sondes, ENSO signals and trends from Atlantic regional time series. J. Geophys. Res., in press.

Ziemke, J. R., S. Chandra, and P. K. Bhartia, 1998: Two new methods for deriving tropo- spheric column ozone from TOMS measure- ments: The assimilated UARS MLS/HALOE and convective-cloud differential techniques. J. Geophys. Res., 103, 22115-22128. (b)

Figure 5. (a) The electrochemical concentration cell (ECC) ozonesonde is connected to the ozone test unit to condition it, check response times, and record an average background value of ozone. (b-c) Before launching, the ozonesonde is sealed inside a weatherproof polystyrene box. A radiosonde is then attached to the outside of the box for data telemetry transmitting meteorological parameters and ozone to a ground receiving station on-board the ship. (c) The final launch set-up shows a small parachute and payout reel connecting the ozonesonde box to the balloon.

30 July/August 1999 • Vol. 11 No. 4

conducted by researchers at Goddard and the University of Colorado. In March 2000, Pre-Launch Validation Activities for the UCSB will carry out snow-mapping MODIS Snow and Sea Ice Algorithms validation with Airborne Visible-Infrared Imaging Spectrometer (AVIRIS) flights at King’s river basin, California.

— Dorothy K. Hall ([email protected]), NASA Goddard Space Flight Center Greenbelt, MD Snow-cover products validation — Shusun Li ([email protected]), Geophysical Institute/University of Alaska, Fairbanks, AK — Anne W. Nolin ([email protected]), NSIDC/University of Colorado, Boulder, CO Validation activities in the MODIS snow and — Jiancheng C. Shi ([email protected]), ICESS/University of California at Santa Barbara, Santa Barbara, CA ice team at NASA/GSFC.

Field and aircraft experiments have been conducted in order to assess the accuracy of snow-cover mapping using the MODIS snow-mapping algorithm. Experiments Introduction conducted in Alaska, Saskatchewan, and the mid-western United States in continu- Pre-launch validation efforts have been of ous snow cover, and New York and New conducted for the at-launch Moderate Alaska project is Hampshire in patchy snow cover, have Resolution Imaging Spectroradiometer responsible for validation of the provided a framework for determination (MODIS) snow and ice products. Daily MODIS sea ice products, and the UCSB of expected, global snow-cover mapping and 8-day composite snow-cover and sea and University of Colorado teams are errors with MODIS. Using the EROS Data ice products will be produced at 500-m conducting validation activities for the at- Center (EDC) land-cover maps of North and 1-km spatial resolution, respectively, launch snow cover, and post-launch snow America and Eurasia (Loveland and following the launch of Terra. A post- albedo products, respectively. Belward 1997), the Earth’s land surface launch, daily, global snow albedo product has been classified into seven different will also be produced, and is still under Numerous field programs have already classes as follows: forest, mixed agricul- development. been conducted in support of validation ture and forest, barren/sparsely veg- efforts for the snow and ice products (Hall etated, tundra, grassland/shrublands, It is anticipated that the MODIS at-launch et al. 1993, 1998a and b; Klein et al. 1998a; wetlands, and snow/ice (Hall et al. 1998a). snow-cover maps will be used, along with Li et al. 1999; Nolin and Stroeve 1999a and Snow-mapping errors, derived using other data, by hydrological modelers to b; Zhou and Li 1999). MODIS prototype algorithms for most of make predictions of water supply and the seven classes, were determined from flooding, and by general circulation Four field programs are planned for 2000. previous field, aircraft and satellite studies modelers and climatologists to study the The first is a cruise in the Southern Ocean (Hall et al. 1993, 1998a; Klein et al. 1998b). Earth’s energy balance and climate. In during which the University of Alaska addition, the MODIS snow and ice will perform field validation with concur- Average monthly snowline positions for products will be used as input to other rent Terra overflights, and is scheduled for the Northern Hemisphere were obtained MODIS algorithms to produce other February and March 2000. Another from the NOAA National Environmental products, e.g., the MODIS cloud mask. Southern Ocean cruise is proposed to NSF Satellite, Data and Information Service for the austral spring in September and (NESDIS) (Matson et al. 1986). The Pre-launch validation activities are being October 2000. These sea ice cruises will be snowline positions were registered to conducted by the MODIS snow and ice conducted in concert with other work land-cover maps, and the percent of each team at Goddard, as well as by the funded by NSF. Also in February 2000, of the seven land covers north of the University of Alaska, the University of there will be a field program in Wisconsin continental snowline was calculated Colorado, and the University of California and Minnesota for validation of the snow- monthly for North America and Eurasia. at Santa Barbara (UCSB). The University cover and snow albedo algorithms Using the information on snowline

31 THE EARTH OBSERVER

position and snow-mapping error shown in Table 1. It is at these times that on the ER-2 (Shi and Li 1998; Li and Shi estimates in the various land covers, snow covers most or all of the boreal 1998), have been used. Thus, a large hemispheric snow-mapping errors were forests and many of the forests located in amount of the ground truth for snow calculated for each land-cover type. It was the mid-continent regions. The expected, covered areas can be used to validate the assumed that errors are uniform within aggregated Northern Hemisphere snow- accuracy of the snow mapping algorithm each class. While this is not always a valid mapping error is about 7.5%. Error under a variety of viewing and illumina- assumption, it allows us to make a estimates will be refined after we acquire tion conditions, and land cover, atmo- preliminary estimate of Northern Hemi- the first year of MODIS global snow-cover spheric and terrain conditions. The sphere snow-cover mapping accuracy. data. greatest possible error of the MODIS algorithm is expected to occur in rugged Snow-mapping errors are expected to be Other efforts by the MODIS snow and ice alpine and forest regions, especially under greatest in forested areas. In forested team have focused on determining relative patchy-snow conditions. In the pre-launch areas, even where snow cover on the accuracy of snow cover products derived time frame, validation has been performed ground is reported to be 100%, snow cover from different sensors (Hall et al. in press; using 67 AVIRIS scenes (10 × 12 km) for observed from space may not be seen as Tait et al. submitted), and on the develop- which the VNIR color photos are avail- complete because the area covered by tree ment of a snow-cover product using both able. These data are mainly from the Sierra stems, branches and canopies is often non- visible/near-infrared and passive- Nevada and South Cascades Mountain snow covered. In addition, less snow in microwave data. The spatial resolution Ranges from April to July, during snow forests will be mapped at off-nadir view and bands of the sensor are key to melt season from 1994-1997. angles than at nadir. When viewed from accurate snow mapping; however, the an angle, the tree stems, branches and ability to map snow through cloud cover The validation was performed in two trunks block the view of the snow more and darkness is also important. A combi- ways: 1) total snow fraction at the AVIRIS than when the forest is viewed from nadir. nation of visible, near-infrared and scene scale - about 10 × 12 km, and 2) When viewing a snow-covered forest from passive-microwave sensors can provide a pixel-based validation. The former the air, little or no snow cover may be snow-cover product that takes advantage represents the case for climatic studies seen, especially if there is no snow in the of good spatial resolution, and also has the where we are only interested in the snow tree canopy. ability to map snow through clouds and fraction information at a grid but do not darkness. Algorithms that will combine care where they are inside the grid, such The greatest monthly errors in snow MODIS and Advanced Microwave as GCM input. The latter represents the mapping can be expected from about Scanning Radiometer (AMSR) data, case for hydrological studies, especially at November through April (9-10%) as following the launch of the PM-1 platform the drainage basin scale, where we need in 2000, are under development in information, not only on the total snow- collaboration with Al Chang, NASA covered area, but also on locations within Month North America Eurasia GSFC, of the AMSR team. the grid. In terms of estimating the total snow-covered area (SCA) fraction from an January 9 9 Activities for snow-cover product validation at AVIRIS scene, the MODIS algorithm has a February 9 9 March 9 10 UCSB. slightly larger error in the large fraction April 9 10 range, with over-estimation in most cases, May 6 8 Progress has been made in testing the than its estimation at the small fraction June 5 6 MODIS snow cover mapping algorithm range. The overall accuracy for the July 5 5 August 5 5 under the most difficult circumstances – fraction cover in a scene is 12.1% with a September 5 5 patchy snow cover in mountainous areas. maximum error of 31.6%. For a grid of October 6 7 GCM models at the regional to global November 9 8 The focus of the UCSB investigation has scale, it is generally at a scale coarser than December 9 8 thus far been on alpine watersheds. For the AVIRIS coverage. The error is much validation of the MODIS snow-cover less than the current techniques that are Table 1. Maximum expected errors (in %), by algorithm, digitized, high resolution (1-4 m) used to generate the snow extent as the month, in Northern Hemisphere snow mapping using EOS/MODIS data. VNIR color photos taken simultaneously GCM input, such as that obtained from

32 July/August 1999 • Vol. 11 No. 4

passive microwave sensors. Therefore, this Activities for the future snow albedo product included atmospheric optical depth, work shows that MODIS will significantly validation at the University of Colorado. diffuse and total irradiance, and snow improve the accuracy of the snow extent physical properties. Ground-based snow input data for climatic and regional or Snow surface albedo will be one of the spectral reflectance measurements were global hydrological studies. post-launch products to be generated from made using an Analytical Spectral Devices data acquired by the MODIS instrument. FieldSpec FR field spectrometer. Airborne For pixel-based validation, however, the In particular, the angular reflectance data were successfully acquired on March overall RMSE from all scenes is 25.1%. characteristics over snow are needed to 10, 1998. MAS data were collected over the When a binary classifier is used, the convert satellite measurements of bidirec- study site, concurrent with the ground- probability that a “mixed” pixel will be tional reflectance to albedo. MISR also based measurements of atmospheric and classified as snow is related to the fraction requires characterization of the angular snow properties. The MAS data have now of snow in a pixel. The spatial distribution reflectance characteristics over snow, and been calibrated and agree well with of snow cover in a mountain area depends thus this work will be useful for character- ground-based snow reflectance measure- upon elevation, surface orientation, and izing the surface albedo needed for MISR ments. Now that the BRDF conversion the intensity and direction of wind. It is, in Top-of-Atmosphere products. This scheme has been validated, anisotropic general, characterized as full, large validation effort has a three-pronged reflectance factors have been derived for fraction, and small fraction at high, approach consisting of: MODIS (channels 1-19 and 26) at an middle, and low elevation zones, respec- angular resolution of 5 degrees. The same tively. A common problem with a binary • validation of a model to convert from will be done for MISR channels in the near classifier is that its estimation will over surface reflectance measurements to future. estimate or under estimate the snow surface albedo; extent in different elevation zones. • development and validation of a A critical concern is the current lack of an Because the snow melting rate differs at scheme for converting narrowband to aerosol retrieval method that provides the different elevation zone, the sensitivity broadband albedo; and accurate estimates over snow-covered of the snow-melt prediction to snow • sensitivity analyses to determine how surfaces. This prevents accurate atmo- mapping accuracy differs at each elevation atmospheric components, especially spheric correction of MODIS and MISR zone. Either high resolution image data aerosols, affect the accuracy of the data over snow. Through a combination of (i.e. ASTER or ETM+) or a more sophisti- albedo retrievals. model simulations and field validation cated snow mapping technique (such as experiments (both pre-launch and post- estimating snow fraction at each pixel) The DISORT (DIScrete Ordinates Radia- launch), this research will validate an may be required for hydrological applica- tive Transfer) model is used as the basis existing scheme for converting measure- tions at the drainage basin scale, in for converting spectral reflectance to ments of snow bidirectional reflectance to mountainous areas. spectral albedo. Data from a pre-launch snow albedo for both MODIS and MISR. validation campaign (held in March 1998 To date, all the modeling sensitivity For post-launch validation, we have at a site near Mono Lake, California) have analyses have been completed. These runs investigated the feasibility of using high- confirmed the efficacy of this model for were made using the 6S model, which was resolution image data—ASTER and characterizing the bidirectional reflectance modified by adding a snow spectrum for ETM+—to derive ground truth using 1) distribution function (BRDF) of snow. surface reflectance. Snow BRDF data have existing binary classification algorithms, With extensive support from JPL scientists now been added to the 6S model so that and 2) the spectral linear unmixing James Conel and Mark Helmlinger of the anisotropic effects are considered. Results technique. The latter can provide very MISR validation team, the University of from the 6S runs show that reflectance for accurate snow mapping (Shi et al. 1999a Colorado team was able to collect a all channels decreases as aerosol optical and b). In addition, the data from AVIRIS, unique data set of snow bidirectional depth increases. Snow is brighter than MAS, and EO-1 will be used to obtain reflectance using a continually rotating path radiance and aerosols have a net ground truth data to validate the MODIS radiometer known as PARABOLA. Data darkening effect, which is roughly linearly algorithm under different conditions. from this instrument have now been a function of aerosol optical depth. Urban calibrated and are in close agreement with aerosols have the greatest effect and DISORT results. Additional measurements maritime aerosols the least effect. In post-

33 THE EARTH OBSERVER

launch efforts, we will quantify how the latter has a diameter of 20-30 m. The reflectance, two precision spectral different estimates of atmospheric aerosols mixture of ice and water with various ice pyranometers for the derivation of surface change our estimates of albedo from concentrations in the marginal ice zone total albedo, two broad band thermal IR MODIS and MISR. may pose difficulties for mapping ice radiometers and 20 automatic data coverage and estimating ice concentra- logging thermisters for measurement of The University of Colorado team is tions using conventional satellite optical surface temperature and brightness currently using ASD field data and sensors that only have a small number of temperature, and a sun photometer for DISORT model output (convolved to the wavelength channels, Riggs et al. (1999) estimation of the atmospheric water vapor MODIS and MISR channels) to derive have successfully used the MODIS content and correction of atmospheric relationships between narrowband and Airborne Simulator (MAS) to map sea ice effects on spaceborne remote sensing data. broadband albedos. Such a scheme will extent and classify sea ice type in the Future cruises in the region have been allow intercomparison of broadband marginal ice zone in the Bering Sea in planned for field validation with concur- albedo retrievals between MODIS and April. The Southern Ocean would be ideal rent Terra satellite flights, with one MISR. While still in the preliminary stage, for further demonstration of the useful- scheduled for February and March 2000, this work will be a focus of efforts in the ness of the sophisticated MODIS sensor and the other proposed to NSF for an upcoming year. Plans are also being made for mapping sea ice, especially for austral spring season in September and for a post-launch validation campaign in mapping the vast marginal ice zone in the October 2000. February 2000 at Lake Mendota, Wisconsin. region. References Sea ice products validation Second, the Terra satellite instruments will map the Southern Ocean in a solar Foster, J. L., A. T. C. Chang, and D. K. Hall, 1994: Snow mass in boreal forests derived from Activities for the sea ice cover and ice-surface illumination condition considerably a modified passive microwave algorithm, temperature validation at the University of different from that for the Arctic Ocean. Multispectral and Microwave Sensing of Forestry, Hydrology, and Natural Resources, E. Mougin, K. Alaska. Because the snow and ice surface reflec- J. Ranson and J. A. Smith (ed.), September 26- tance is anisotropic, lower solar illumina- 30, 1994, Rome, Italy, pp. 605-617. The University of Alaska Fairbanks (UAF) tion angles can introduce a major error Hall, D. K., J. L. Foster, J. R. Irons, and P. W. group is responsible for validation of source in spring and autumn when the Dabney, 1993: Airborne bidirectional radiances MODIS ice product in the Southern Ocean. solar illumination angles in the region are of snow-covered surfaces in Montana, U.S.A., Annals of Glaciology, 17, pp. 35-40. An independent validation in the South- low at local solar time. This situation ern Ocean is necessary for two main requires collection of field data in the Hall, D. K., J. L. Foster, V. V. Salomonson, A. G. Klein, and J. Y. L. Chien, 1998a: Error analysis reasons. Southern Hemisphere to assess the for global snow-cover mapping in the Earth performance of the MODIS snow and ice Observation System (EOS) era, Proc. of First, sea ice in the Southern Ocean is product in the region. IGARSS’98, Seattle, WA, pp. 1524-1526. more dynamic than that in the Arctic in Hall, D. K., J. L. Foster, D. L. Verbyla, A. G. terms of the annual variability in its With support from both NASA and the Klein, and C. S. Benson, 1998b: Land and forest cover classification using aircraft and satellite spatial coverage. In the Arctic, sea ice National Science Foundation (NSF), the data in snow-covered areas in central Alaska. coverage varies from a March maximum UAF group has conducted two pre-launch Remote Sens. of Environ., 66, 129-137. of approximately 15 × 106 km2, to a field campaigns in the Ross Sea aboard the Hall, D. K., A. B. Tait, J. L. Foster, and A. T. C. × 6 2 September minimum of about 8 10 km . U.S. Research Vessel Nathaniel B. Palmer, Chang, in press: Intercomparison of satellite- In the Antarctic, a maximum extent of with one in the austral winter in May and derived snow-cover maps. Annals of Glaciology. approximately 20 × 106 km2 occurs in June 1998, and one in the austral summer Klein, A. G., D. K. Hall and G. A. Riggs, 1998a: × 6 September while a minimum of 4 10 in January and February 1999. The group Improving snow-cover mapping in forests km2 occurs in February (Parkinson and was able to calibrate field remote sensing through the use of a canopy reflectance model, Hydrological Processes, 12, 1723-1744. Gloersen 1993). In addition, the more instruments and to obtain field data sets stormy condition of the Southern Ocean during the cruises. The field remote Klein, A. G., D. K. Hall, and K. Seidel, 1998b: creates a vast marginal ice zone where the Algorithm intercomparison for accuracy sensing instruments tested in the cruises assessment of the MODIS snow-mapping main sea ice types are pancake ice and include a fine resolution spectroradio- algorithm, Proc. of the 5th Eastern Snow Confer- cake ice. The former has a size of <1 m and meter for measurement of surface spectral ence, Jackson, New Hampshire, June 2-3, 1998, pp. 37-45.

34 July/August 1999 • Vol. 11 No. 4

Li, S., X. Zhou, and K. Morris, 1999: Measure- multiple datasets for snow cover mapping, Field listed “sticky” technical issues: ment of Snow and Sea Ice Surface Temperature submitted to Remote Sens. of Environ. and Emissivity in the Ross Sea, Proc. of IGARSS’99, IEEE No. 99CH36293, II, 1034-1036. Zhou, X. and S. Li, 1999: Summer and winter • leakage problem; snow and sea ice surface spectral directional • many sources problem—there are Li, Z., and J. Shi, 1998: Establishment of a reflectance and albedo measured in the Ross database for development and validation of Sea, Proc. of IGARSS’ 99, IEEE No. 99CH36293, thousands of power plants worldwide; snow mapping algorithms. To be published in I, 104-106. • delayed responses problem; and Proc. of Fourth Remote Sensing in Hydrology, Santa • finite commitment periods. Fe, Nov. 1998.

Loveland, T. R., and A. S. Belward, 1997: The “Sticky” scientific issues include accuracy, IGBP-DIS global 1 km land cover data set, presentation of relevant processes, spatial DESCover: first results. Int. J. of Remote Sens., 18, 3289-3295. resolution, etc. Assessing of Kyoto sinks requires looking at inventories, a book- Matson, M., C. F. Ropelewski, and M. S. Varnadore, 1986: An atlas of satellite-derived keeping land-use model, biogeochemical northern hemisphere snow cover frequency. models, and inversions and eddy flux. The National Weather Service, Washington, D.C., 75 real “crux” is land-use/land-cover change. pp. (Continued from page 18)

Nolin, A. W., and J. C. Stroeve, 1999a: Validation In the characterization of 1990, sinks are of the MODIS snow albedo product. In Proc. of Minutes of The Fifteenth assessed only in relation to the 1990 the Annual Meeting of the Asso. of Amer. Geogr. Earth Science Enterprise/ baseline. EOS data will improve the Nolin, A. W., and J. C. Stroeve, 1999b: Valida- Earth Observing System understanding of earlier AVHRR/Landsat tion studies and sensitivity analyses for (ESE/EOS) Investigators data retrievals of snow albedo from EOS AM-1 instruments. NASA 1999 Annual Report, 16 pp. Working Group (IWG) Meeting Field said that long-wavelength SAR can Parkinson, C. L., and P. Gloersen, 1993: Global be used to determine carbon stocks, and sea ice coverage, in Atlas of Satellite Observations Related to Global Change, R. J. Gurney, C. L. his talk which would cover the require- he deplored the dropping of such instru- Parkinson, and J. L. Foster (eds.), Cambridge ments posed by the protocol; “sticky” mentation early in the EOS program. He University Press, pp. 371-383. issues involving both technology and stressed the need to develop a way to Riggs, G. A., D. K. Hall and S. A. Ackerman, biogeochemistry; contributions of EOS- sense soil carbon remotely. Enforcement is 1999: Sea ice extent and classification mapping type data; and enhancing the relevance of the key to success of Kyoto, but it is not with Moderate Resolution Imaging Spectro- radiometer Airborne Simulator Data. Remote EOS data. spelled out. Sens. of Environ., 68, pp. 152-163. The FCC is the Framework Convention on Shi, J., and Z. Li, 1998: Validation of MODIS Lively discussion followed this presenta- snow mapping algorithm using AVIRIS image Climate Change, dated June 19, 1993. The tion with opposing sides saying that EOS and Near-infrared color photo, in Proc. of “Protocol” is a document still under would or would not be able to play a role SPIE’98, Ed. R. O. Green and Q. Tong, The Int. Society for Optical Eng., 3502, pp. 243-250, development. The ultimate objective of the in the necessary science to carry out Beijing, China protocol is to stabilize greenhouse gases Kyoto. Shi, J., 1999a: Estimation of Snow Fraction within an ‘appropriate’ time frame. The Using AVIRIS Simulated ASTER Image Data. To protocol is based on the ‘precautionary’ be published in Proc. of Eighth Airborne Geos. principle. AVIRIS Workshop.

Shi, J., 1999b: Estimating snow fraction using Article 3 of the protocol says that green- AVIRIS simulated ASTER image in alpine house gas emissions must be five percent regions, Proc. of IGARSS ‘99, IEEE No. 99CH36293, III, pp. 1795-1797. below those of the 1990 base period. The change in carbon stocks from 1990 is a key Tait, A. B., D. K. Hall, J. L. Foster, and A. T. C. Chang, 1998: Detection of snow cover using measure. Article 5 has provisions for millimeter-wave imaging radiometer (MIR) monitoring and trading emissions. There data. Remote Sens. of Environ., 68, pp. 53-60. can be various agreements among Tait, A. B., D. K. Hall, J. L. Foster, A. T. C. countries to accomplish ‘clean’ develop- Chang, and R. L. Armstrong, 1999: Utilizing ment.

35 THE EARTH OBSERVER

at http://www.strategies.org/ Earth Science Education Program conference.html. The conference will be held November 14-17, 1999 in Austin, Update Texas, and is a first—bringing together the principals from NASA Earth science education programs from across the — Nahid Khazenie ([email protected]), Education Program Manager, Office of Earth Science, NASA Headquarters country. Over 140 participants have — Steve Graham ([email protected]), EOS Project Science Office, registered for the conference from Raytheon ITSS programs targeted at all levels and audiences.

NASA Student Involvement The Institute for Global Environmental design model, are delivered over the Program (NSIP) 2000 Strategies (IGES) in Arlington, VA, Internet, and feature student-centered, announces an opportunity to participate knowledge-building virtual communities. NSIP is a national competition for in the Earth System Science Education Successful proposers will participate in an students in grades 3-12 that links student Alliance (ESSEA), an innovative profes- intensive four-day training workshop to research and projects to NASA’s strategic sional development program. learn how to deliver the courses. enterprises. Two of the areas of competi- tion are related to NASA’s Earth Science ESSEA is a partnership among IGES, the The ESSEA program, with the on-line Enterprise: Watching Earth Change (for Center for Educational Technologies (CET) Earth system science courses as it’s grades 5-8 and 9-12) and Earth Systems in at Wheeling Jesuit University, and centerpiece, will not only include training my Neighborhood (for grades 3-4). sponsored by NASA’s Earth Science teachers but will develop an active and Entries for this year’s competition are due Enterprise. ESSEA has the following long-term mechanism for delivering February 1, 2000, and the National objectives: professional development on a national Symposium will be held in Washington, scale. DC May 6-10, 2000. For more information • promote the growth of knowledgeable about NSIP, including the 1999-2000 and well-equipped K-12 Earth system IGES intends to award three-year grants to Program Announcement and copies of science educators through on-line entities with unique and innovative Educators’ Guides for each competition, professional development courses; proposals to facilitate and offer the Earth see the project website at http://www. • demonstrate the effectiveness of the system science on-line courses developed nsip.net. World Wide Web in the promotion of and tested by CET. It is anticipated that professional development of K-12 each grant will be in the range of a total of University Earth System Science (UNESS) Project Earth system science educators; and $45,000 over a three year period. Plans are • respond directly to the need to to issue up to six grants in 1999; up to Solicitation Number: AO-99-OES-02 prepare more teachers to meet the eight new grants in 2000; and an addi- Response Date: Dec 01, 1999 demand of a growing U.S. student tional eight new grants in 2001. population. The National Aeronautics and Space A copy of the full announcement and Administration (NASA) Office of Earth To meet these objectives, ESSEA will proposal guidelines is available at Science announces the opportunity to support universities, colleges, and science http://www.strategies.org/essea.html conduct innovative spaceborne Earth education organizations in offering K-12 system investigations in the form of Earth system science (ESS) on-line Earth Science Education Forum complete spaceflight missions or second- graduate courses that have been devel- Program ary payload instruments through the oped within the CET at Wheeling Jesuit University Earth System Science (UnESS) University for NASA’s Earth Science The preliminary program for the first Project. This Announcement of Opportu- Enterprise. The Earth system science NASA Earth Science Education Forum nity (AO) is intended to foster the devel- courses use an innovative instructional is available on the conference WWW site

36 July/August 1999 • Vol. 11 No. 4

opment of the next generation of Earth system scientists, engineers, managers, EOS Scientists in the News educators, and entrepreneurs through significant and meaningful hands-on — Emilie Lorditch ([email protected]), Raytheon ITSS student involvement in Earth observation space missions at the university level. The hands-on student involvement should include helping prepare the proposal through analysis and distribution of the data to the scientific community. This AO will give equal weight to the scientific and “Acclimating to a Warmer World,” Science “Carbon Dioxide In Air Can Stem Soil student/applications involvement aspects News (Aug. 28) by Richard Monastersky. Erosion,” Reuters (Aug. 11). Chris Field of the proposal. Cynthia Rosenzweig (NASA GISS) says (Carnegie Institution of Washington.) has that the study seeks to analyze how found that atmospheric carbon dioxide The AO is at the Earth Science Enterprise climate change will affect the 19.6 million helps bind soil particles together, which Home Page: http://www.earth.nasa.gov/ people living in the New York metropoli- can help in places vulnerable to soil under “Research Opportunities” or via tan area. Rosenzweig is part of the U.S. erosion. Field discovered that carbon anonymous ftp at ftp.hq.nasa.gov/pub/ National Assessment of the Potential dioxide stimulates changes in soil struc- ese. Consequences of Climate Variability and ture making conditions perfect for the soil Change and leads the New York study. to stick together. This announcement will be open for the period from August 2, 1999, through “Sunspots May Play Role in Global “La Niña Keeps Hanging Around,” December 1, 1999, and proposals may be Warming,” Christian Science Monitor (Aug. Environmental News Network (Aug. 10). submitted at any time throughout that 26) by Robert C. Cowen. James Hansen William Patzert (NASA JPL) examines La period. (NASA GISS) discusses the possibility that Niña’s impact on North America’s climate aerosols and dust in the atmosphere, this fall. Using data from the TOPEX/ Point of Contact: generated partly by human activity, have Poseidon mission, Patzert observed how Mr. Gordon Johnston substantially counteracted expected La Niña caused contrasting ocean levels UnESS Project Executive greenhouse warming. and temperatures on opposite sides of the Phone: (202) 358-4685; Fax: (202) 358-2769 north Pacific Ocean. E-mail: [email protected] “Getting to the Roots of Carbon Loss,” http://nais.nasa.gov/EPS/HQ/ Science (Aug. 20) by Jocelyn Kaiser. Chris “Case Grows for Climate Change,” date.html#AO-99-OES-02 Field (Carnegie Institute of Washington) Chemical & Engineering News (Aug. 9) by Ruth DeFries (Univ. of Md.), and Steven Bette Hileman. James Hansen (NASA Running (Univ. of Mont.) aim to sum up GISS), Stephen Wofsy (Harvard Univ.), V. all the carbon added to the atmosphere Ramanathan (Scripps), and Kevin since humans began to leave a significant Trenberth (NCAR) are searching for mark on the Earth. evidence that proves a human influence on the global climate and the connection “Research to Benefit from Cheaper between greenhouse gas emissions and Landsat Images,” Nature (Aug. 19) by global climate change. Tony Reichardt. Samuel Goward (Univ. of Md.) explains how even a modest land “Carbon Dioxide Shakes Off Its Pursuers,” use study can result in expensive bills for Science News (July 24) by Richard images. The cost of a Landsat image is Monastersky. Inez Fung (Univ. of Calif. $475 a scene, which make the images Berkley) challenges a team led by much more affordable than most satellite Princeton University researchers that images. reported that the forests and fields in the

37 THE EARTH OBSERVER

United States are acting as a carbon been a decline in carbon dioxide emis- Goddard Space Flight Center, Greenbelt, dioxide sink with her own study. Fung sions. This finding suggests that the MD 20771, Tel. (301) 441-4031; fax: (301) says the biggest problem with the study is atmospheric carbon dioxide won’t double 441-2432; e-mail: elorditc@pop900. that North America appeared to absorb by 2050 as previously thought, giving gsfc.nasa.gov. tons of carbon dioxide while Europe and humanity more time to cope with global Asia absorbed next to nothing. climate change.

“Flood Cooled Atmosphere in Ice Age,” “Global Warming —Turning Up the Los Angeles Times (July 22) by Chris Kahn. Heat,” ABC’s World News Now (July 15). Richard Alley (Penn State Univ.) explains As the heat index soars, 25 deaths have Cynthia how a giant flood can trigger a worldwide been blamed on the heat. (Continued from page 20) freeze in a matter of decades. Rosenzweig (NASA GISS) says that developing nations appear to be more “Ocean Fever Heralds African Epidem- vulnerable to global warming. CEOS Working Group on ics,” Science News (July 17) by Richard Calibration and Validation Monastersky. Kenneth J. Lithicum (Walter “Flights Could Be Helping Make Earth’s Meeting on Digital Reed Army Inst. of Research) and Assaf Skies More Unfriendly,” Dallas Morning Anyamba (NASA Goddard) discuss their News (July 12) by Alexandra Witze. Elevation Models and research about the 1997 Rift Valley Fever Patrick Minnis (NASA Langley) has Terrain Parameters epidemic that killed tens of thousands of studied how contrails can thin into cirrus- livestock and hundreds of people in like clouds that trap more of the sun’s of in situ validation data. In the next few Kenya. Lithicum and Anyamba’s research heat. Scientists think the situation will get years there will be an increase in the also appeared in the Associated Press and worse and predict that by 2050 the United availability of moderate resolution (c. 1 Science. States will have nearly three times more km) sensing systems and a proliferation in contrail coverage than in 1999. data products. With the high cost of in situ “Does a Globe-Girdling Disturbance Jigger data collection relative to the resources El Niño?,” Science (July 16) by Richard “Purer Air May Bring More Heat / available for the non-space segment of Kerr. Michael McPhaden (NOAA) and Sulfates Now Block Some of Sun’s Rays,” most observation programs, the potential Compton Tucker (NASA Goddard) Houston Chronicle (July 1) by Seth benefits from international coordination explain the unpredictable atmospheric Borenstein. James Hansen (NASA GISS) are considerable. The May 1999 CEOS oscillation that may help transform an cautions that determining how sulfur, WGCV meeting and its recommendations ordinary El Niño into a monster for clouds, and other variables factor into provide a point of departure for such forecasters. climate change is too complicated to international coordination and an ex- predict outcomes confidently. panded role for the CEOS Calibration and “Rift Valley Fever,” National Public Radio, Validation Working Group. Science Friday (July 16). Compton Tucker “A Better View of Earth,” Popular Science (NASA GSFC) reports on the Rift Valley (July 1999) by Mariette DiChristina. fever outbreaks in East Africa from 1950- NASA’s first Earth-observing satellite, 1998 and how these outbreaks followed Terra, will capture a better view of the periods of abnormally high rainfall. complex processes that govern the Earth’s climate. Terra will look for changes in “New Evidence Heats Up Climate long-term land-use patterns, snow and ice Debate,” Christian Science Monitor (July 15) cover, and climate variations. by Robert C. Cowen. James Hansen (NASA GISS), V. Ramanathan (Scripps), EOS researchers: Please send notices of and Charles Kennel (Scripps) discuss the recent media coverage in which you have decline by 25 percent of heat-trapping been involved to: Emilie Lorditch, EOS greenhouse gases although there has not Project Science Office, Code 900,

38 July/August 1999 • Vol. 11 No. 4

— 2000 — June 12-14 EOS Science Calendar Sixth Circumpolar Symposium on Remote January 9-14 Sensing of Polar Environments, Yellowknife, American Meteorological Society 2000, Long Northwest Territories, Canada. E-mail: Oct. 12-14 Beach Convention Center, Long Beach, CA. Call Landsat Science Team Meeting, NASA/GSFC. [email protected], tel. (867) 920- Contact Darrel Williams, e-mail: darrel@ (202) 682-9006; Fax: (202) 682-9298; e-mail: 3329, website at http://www.gov.nt.ca/RWED/ ltpmail.gsfc.nasa.gov. [email protected]. rs/circumpolar2000. February 17-22 October 19-21 American Association for Advancement of July 16-23 AIRS Science Team Meeting, University of Science (AAAS), Washington, DC. Call (202) International Society for Photogrammetry & Maryland, Baltimore County. Contact H.H. 326-6736, website at http://www.aaas.org. Remote Sensing (ISPRS) 2000, Amsterdam. Aumann, e-mail: [email protected]. Call for Abstracts. Contact organizing March 7-10 secretariat, tel. +31 20 50 40 203; Fax: +31 20 Oceanology International 2000. Call for October 25-27 50 40 225; e-mail: [email protected] Joint TOPEX/POSEIDON-Jason Science Papers. Contact Christine Rose, Conference Executive, Oceanology International 2000, Working Team Meeting, St. Raphael, France. July 24-28 Contact Yves Menard, e-mail: menard@ Spearhead Exhibitions Ltd, Ocean House, 50 IGARSS 2000 International Geoscience and imhotep.cst.cnes.fr, or Lee-Lueng Fu, e-mail: Kingston Road, New Malden, Surrey KT3 3LZ, Remote Sensing Symposium, 20th [email protected]. UK. Tel. +44 (0) 20 8949 9222; Fax: +44 (0) 20 Anniversary, Hilton Hawaiian Village, Honolulu, 8949 8186/8193; e-mail: christine.rose@ Hawaii. Call for Papers. Website at http:// spearhead.co.uk; website at http://www. www.igarss.org. Global Change Calendar spearhead.co.uk. July 24-29 March 14-15 International Radiation Symposium (IRS- Adaptive Sensor Array Processing Workshop, October 24-28 2000), Saint Petersburg State University, St. Geological Society of America. Denver, CO. MIT Lincoln Laboratory. Call for Papers. Petersburg, Russia. Contact conference Contact Tammy White, e-mail: twhite@ Contact Edward J. Baranoski, e-mail: kballos@ coordinator, Evgenia M. Shulgina, St. geosociety.org, website at http://www. ll.mit.edu, website at http://sam2000.uconn. Petersburg State University, Research Institute geosociety.org. edu. of Physics, 1 Ulyanovskaya, 198904, St. Petersburg, Russia; Fax: +7 (812) 428-72-40; November 2-5 March 16-17 e-mail: Evgenia.Shulgina@ pobox.spbu.ru; or The CEOS Global Observation of Forest Cover IEEE Sensor Array and Multichannel (SAM) [email protected]. (GOFC) Meeting on Fire Mapping and Signal Processing Workshop, Cambridge, MA. Monitoring, the Joint Research Center, Ispra, Call for Papers. Contact Edward J. Baranoski, Italy. Contact Frank Ahern, e-mail: Frank. e-mail [email protected], website at http:// [email protected] .gc.ca. sam2000.uconn.edu. March 27-31 December 6-10 28th International Symposium on Remote The Fourteenth William T. Pecora Memorial Sensing of Environment, Cape Town, South Remote Sensing Symposium & the Land Africa. Call for Papers. For abstracts Satellite Information in the Next Decade III submission: [email protected], or Publications Available Conference & Exhibition, Doubletree Hotel, see website at http://www.isrse.co.za, Fax: +27 Denver, Colorado. Contact: ASPRS, The 21 883 8177; tel. +27 21 886 4496 (ask for EOS/AM-1 Digital Elevation Model Imaging & Geospatial Information Society, Deidré Cloete); postal: The 28th ISRSE (DEM) Data Sets: 5410 Grosvenor Lane, Suite 210, Bethesda, Technical Committee, P.O. Box 452, DEM and DEM Auxiliary Datasets in MD 20814-2160; e-mail: [email protected]; Stellenbosch, 7599, South Africa. website at http://www.asprs.org. Support of the EOS/Terra Platform May 22-26 December 13-17 ASPRS: The Imaging and Geospatial Order from: The 1999 American Geophysical Union (AGU) Information Society, 2000 Annual Conference, Thomas Logan Fall Meeting, Moscone Center, San Francisco, May 22-26, 2000. Washington, DC. Call for Jet Propulsion Laboratory CA. Contact: AGU, 2000 Florida Avenue, N.W. Papers. For abstracts submission see website E-mail: [email protected] Washington, D.C. 20009. Tel. (202) 462-6910; at http://www.asprs.org/dc2000; tel. (410) Fax: (202) 939-3229. Phone: (818) 354-4032 208-2855; Fax: (410) 641-8341; e-mail: [email protected].

39 Code 900 Bulk Rate Mail National Aeronautics and Postage and Fees Paid Space Administration National Aeronautics and Space Administration Goddard Space Flight Center Permit G27 Greenbelt, Maryland 20771

Official Business Penalty For Private Use, $300.00

The Earth Observer

The Earth Observer is published by the EOS Project Science Office, Code 900, NASA Goddard Space Flight Center, Greenbelt, Maryland 20771, telephone (301) 614-5559, FAX (301) 614-5620, and is available on the World Wide Web at http://eospso.gsfc.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 Dave Olsen at (301) 441-4245, 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]) Renny Greenstone ([email protected]) Jim Closs ([email protected]) Design and Production: Winnie Humberson ([email protected]) Distribution: Hannelore Parrish ([email protected])

Printed on Recycled Paper