Click for next page ( 88


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 87
5 Elements of the Mission to Planet Earth SYSTEMS DEFINITION The grand themes can be addressed adequately only by a Mis- sion to Planet Earth that includes an integrated and interrelated set of satellite and surface observations. These will involve polar and geosynchronous orbiters, special purpose orbiters, data relay satel- lites, ocean drifting instruments, pop-up buoys, tethered buoys, ocean bottom instruments, optical fiber links to islands and buoys (up to 1000 km seems to be feasible), automated ground stations, and simple ground stations such as corner reflectors, rain gages, and tide gages. Microchip, computer, and low-power technology can make these sea and land observations extremely powerful in their capabilities, and satellite relay links will make it possible to collect data from a large area of the Earth's surface. In addition, modules can be developed for ships and aircraft that can be linked into the global data network. The measurement strategy for the implementation of an ob- serving system to address the grand themes involves a system that is composed of about 8 orbiters, 1000 ocean systems, 1000 auto- mated land stations, thousands of simple surface stations and a variable number of itinerants (ships, planes, balloons). The ele- ments of the entire system are summarized in Table 5.1, including 87

OCR for page 87
88 TABLE 5.1 Elements of a Mission to Planet Earth Quantity Element 1-2 5 2-6 18-24 2-3 1,000 100 100 100-1,000 1,000-10,000 Space Station Special-purpose satellite missions Geosynchronous orbiters Polar orbiters GPS constellation Tracking and data relay satellites Floating buoys, "pop-upe" Moored buoys Ocean bottom stations Smart ground stations Simple ground installations Aircraft Balloons Rockets Ships, research Ships, opportunity the various components of PLATO. The association of a variety of measurements with satellites and in situ stations is summarized in Table 5.2. A1BO included in the latter table are the types of instrumentation required to obtain the measurements. Several of the subsystems and experunent packages are well known and described in NASA documents. Others such as ocean drifting instruments, pop-up buoys, and smart ground stations are relatively new concepts. Some of these would be called land- ers, penetrators, unmanned stations, or rovers if they were being developed for other planets. A part of the strategy for such an interlocked system is given in the NASA Earth Observing System (EOS) report, but the subsystems described therein are just a portion of the whole earth experiment the task group is proposing. As has been repeatedly emphasized, it is important that these measurements be globally complete, simultaneous, and continu- ous. The next task is to suggest what sort of observing systems might satisfy these requirements. The full specification of such an implementation in a period as remote as 1995 to 2015 would take an expertise in instrumentation and space systems not present in this task group. The task group can, however, use reports appli- cable to the shorter time range of the 1990s and make some rea- sonable conjectures about further advances in measurement and

OCR for page 87
89 TABLE 5.2 Summary of Instruments and Measurements for the Mission to Planet Earth Measurement Systems Polar Geosyn- Other Aircraft Ocean Ground Measurement Orbiter chronous Satellites Balloons Stations Stations Magnetic field G G G G G Gravity field (geoid) G G. A G G G. A G. A Stratospheric chem. C, P C P P Aerosols M, P M T T T Winds A, P A T T T Severe storms R. M R. M D T T T Clouds M, P T T T Precipitation R. M, P R. M M, D T. S T Particulate matter M, P M T T T Tropospheric them. C, P C C C Ocean currents M, A M, A A, D A Ocean chlorophyll M M O O Ocean salinity M M O Lake levels A O Sediments M M D O Sea state R R D O Sea ice R. A, M R. A, M A O Glaciere R. M, A A A Snow R. M R. M A Topography A A A Surface temperature M M O O T Albedo M M T Surface geochemistry M M Geological features R. M R. M Cultural features R. M R. M Vegetation M M T. C Soil moisture M M Soil erosion M M Surface strain A A, D A Seismic wave velocities D S S Tectonic deformation D, L S S NOTE: Instrument categories are as follows: M Multispectral imaging R Radar imaging A Altimetry ranging P Vertical profile remote sensing D Data links T Meteorological instruments O Oceanography instruments S Seismographs- acoustic detectors G Gravimeter-magnetometer C Chemical composition instruments L Locations, precise geodetic data processing capabilities, and then suggest the broad outlines of a Mission to Planet Earth after the turn of the century. The two reports on which this chapter is partially based are Earth Observing System (NASA Goddard Space Flight Center, Tech. Memo. 86129, August 1984 (2 volumes)~; and Earth Systems Science Committee Working Group on Imaging and Tropospheric

OCR for page 87
go Sounding Final Report (Caltech Jet Propulsion Laboratory D- 2415, January 1985~. This second report is, to an appreciable extent, a critique of the first. ELEMENTS OF THE SYSTEM Satellite `'Earth Obeer~nng Systems (EOS) The data base that has been built by operational land and meteorological satellite systems since the early 1970s must be preserved and continued. These data should be placed and main- tained in forrr~s that optimize their use in combination with the data generated by the missions already specified for project starts in the period 1986 to 1995, and with the EOS information system described below, to be implemented in the m~-1990s. Earth observables can be characterized in two broad cate- gories: quasi-staticproperties varying slowly enough that map- pings at intervals of several years suffice (e.g., rock types, vegeta- tion regimes, magnetic field); and dynamic properties varying on time scales ot mmutes to seasons (e.g., precipitation, vegetation state, snow cover). EOS is primarily directed to quantities in the second category. The required resolutions and instrument sensi- tivities anticipated in the 1990s are such that the most efficient orbit appears to be sun-synchronous at altitudes of 600 to 1000 km: i.e., inclination about 95, circular, about 14 orbits per day. Such an orbit sees the atmosphere and surface below it always at the same local time, which greatly eases data analysis; 2:00 p.m. is recommended by the EOS report. Some parameters can be oW served from geosynchronous altitudes at 36,000 km. An equatorial geosynchronous orbiter can observe effectively to latitude 60, and a set of five such orbiters (e.g., the current operational weather system) gives good global coverage within these latitude bands. Principles important to any observing system are global syn- opticity, nested coverage (coordination of lower and higher resolu- tion systems), quality controlmost importantly, calibration by in situ measurementsand integration with data interpretation and data continuity. The instruments given in Appendix A are currently recommended for EOS. The altitude of 600 to 1000 km is based primarily on the swath width appropriate for the MODIS, HMMR, and LASA (see Appendix A) instruments. All of this instrumentation would not be carried on the same polar-orbiting . . ~

OCR for page 87
91 sun-synchronous spacecraft. For example, the radar altimeter will be carried on the non-sun-synchronous TOPEX satellite. The report does not address the needs for observations from geosyn- chronous orbits, specialized orbiters, or nonorbiting devices, all of which are important aspects of a total earth observing system. As this report and many others have emphasized, the global nature of atmospheric, oceanic, and land processes makes it nec- essary to view Earth as a single interactive system in order to describe, understand, and predict significant trends in its state. At the same tune, the complexity of the internal processes occur- ring in each component of the system makes it prudent to study these processes separately. We therefore should pursue a strategy of assessing the capabilities of the entire spectrum of remote sens- ing techniques, identifying the directions in which real progress can be made. The ultimate goal is to develop an observational system that can relatively quickly provide an improved state of knowledge about planet Earth and its main components. This leads to the following recommendations: 1. A major effort needs to be devoted to amalgamation of remote sensing data from several spectral regions and from more than one source in order to improve the accuracy of derived pa- rameters. In addition, space and in situ measurements must be combined to provide optimum measurement systems for obtaining the basic ciata. 2. In view of the long-term efforts that used to be devoted to the derivation and utilization of remote sensing data, there is a critical need to develop acceptable calibration methods to guaran- tee long-term stability of the measured data and of the retrieved geophysical parameters. In particular, we need to establish an ac- ceptable approach for providing Relative calibration" of the solar channels on Landsat, AVHRR, and HIRS. 3. Space and in situ conventional data must be related in order to understand the physical processes producing remotely sensed signals and to provide the instrument verification and cali- bration needed to maintain the integrity of the data sets. 4. The requirements for the radiation budget at the top of the atmosphere, surface flux, sea surface salinity, and surface pressure should be added to the set of important parameters. In addition, there is a strong sense of urgency regarding the need for rapid development and unplementation of observational tools for soil

OCR for page 87
92 moisture, precipitation, land surface temperature, and spectral emlsslvlty. 5. EOS should add high-spectral-resolution visible and in- frared instruments for accurate determination of important sur- face and atmospheric parameters. Active lidar systems should also be advanced and tested because of the potentially unique contributions they offer. In general, the task group notes that we must design future remote sensing instruments as integrated systems with comple- mentary sensor packages. This includes coordinated data analysis algorithms, calibration, and validation methods that wiB permit us to observe the planet Earth in "all its dimensions." Because the requirements for earth science research are so demanding in terms of accuracy and precision, and are of such long-term nature, we cannot rely on placing together the needed data from fragmented sets of observations. The EOS space-based infrastructure should be designed to ex- pedite deployment of new instruments using onboard data process- ing and data compression, with overlapping data from consecutive sensors. EOS should also consider the deployment of instruments that include, as an essential part of the system, elements deployed on the surface. As one example, the task group recommends the development and deployment of a system of ocean stations to collect surface and subsurface information. In order to accomplish the Mission to Planet Earth we will need development of new technology to measure some of the more elusive parameters, combine data from the ground network of stations, buoys, and pop-ups with those Corning from satellites, and develop further the data processing, merging, selection, and distribution capabilities. For instance, satellite data collection from autonomous drifting platforms at sea has become routine in the past decade through the French System ARGOS, carried by the TIROS/NOAA A-] polar-orbiting satellites. Fewer than 1000 platforms are now monitored in this way. As new global programs aimed at understanding climate and global change become fully Operational, the task group expects to see an increase of a factor of 10 in ciata rate. Moreover, the collection of oceanographic and seismic data by sea floor and island stations that use fiber optics and other high-data-rate collection techniques is expected to place a demand on the data system at least equal to that of the drifting

OCR for page 87
93 platforms. Thus the task group expects that the l990s will see an urgent need for an increase of at least one order of magnitude in global satellite data collection and transmission capabilities. The task group recommends that planning begin now for a follow- on to ARGOS that ~~! be capable of handling the data from the approximately `04 platforms, and the high (10 times present rates) data rate platforms that are expected in the mid-1990s. Smart Ground Stations A smart ground station (SOS) would have a modular design and an on-site computer capable of a variety of "intelligently deci- sions. For example, a microcomputer based on a 68020 Motorola chip would match the computer power of a VAX/780. Filtration, averaging, some quality control, and scheduling of individual mea- surements could take place on site. These stations could also have a built-in recording system, such as a large-capacity cassette or magnetic tape unit, but the primary mode of data transmission would be satellite telemetry. It Is anticipated that there would be 100 to 1000 SGS installations worldwide. With the computing power described above, the SGS could support many simultaneous measurements, including seismic, geo- detic, meteorological, hydrological, and soil properties. These are examined below. Seismic Measurements An SGS could measure the structure of the Earth's interior, earthquake source mechanisms, strong ground motion in seismic areas, volcanic activities, and tidal forces. A three-component, broad-band seismic station would be capable of resolving 10-9 m/s2 at a period of 25 s. Strong motion instruments should also have broad-band capabilities and remain on scale for accelerations above 1 g. Each channel would be sampled continuously at a rate of 20 samples per second using up to 3 bytes per sample, although a reduction through data compression might be feasible. Strong motion instruments would be triggered if acceleration were to exceed 0.01 g. The sampling rate would be 200 samples/s and the burst data rate would be 1800 bytes/s. Data could be stored in an on-site buffer and transmitted at a lower rate. In volcanically active areas, an SGS could monitor a local

OCR for page 87
94 network of short-period seismometers. The network could have continuous in situ recording with the telemetry activated upon detection of an eruption. Geodetic Measurements Stations making geodetic measurements could monitor ground motions, strains, uplift, volcanic activities, fault zones, plate tec- tonics, and post-seismic and post-glacial rebound. Instrumenta- tion would be dependent on tectonic setting, but it can be expected that emphasis would be on distances comparable to fault lengths and depths of minimally significant earthquakes. For example, in- vestment will be in position-di~erence instruments. Strainmeters and tiltmeters would be deployed in tectonically active regions, with sensitivity and type depending on their proximity to the fault zone. A typical sampling rate might be one sample per 10 s. Depending on the anticipated level of tectonic deformation, a GPS receiver could be permanently deployed, in which case several measurements per day could be performed, or it could be brought to the site at monthly to yearly intervals. In any case, the on-site computer could perform all the necessary signal processing at an expected accuracy of 1 to 3 cm. Electro-optical ranging devices may nonetheless continue to be more precise than the GPS at distances less than 10 km. However, mobile GPS receivers will be needed to obtain spatial resolution complementary to the temporal resolution of the fixed instruments. Meteorological and Hydrological Measurements The purpose of obtaining long-term calibrated data in these areas is to understand the coupling of radiative, dynamical, and chemical processes in the atmosphere; to improve the accuracy and extend the range of weather forecasting; to assess the influences of changes in sea surface temperature, ocean surface currents, and sea and land ice cover on climate; and to determine what factors control the hydrological cycle. The SGSs will be indispensable in obtaining the 10 to 50 years of continuous observations that are needed to collect the requisite amount of data. Atmospheric measurements are necessary for the following parameters: temperature and humidity profiles, surface winds, aerosols, minor constituents, long-wave and shortwave radiation,

OCR for page 87
9s and the amount, height, emissivity, albedo, and water content of clouds. With regard to hydrology, measurements must be made of ice and snow extent, thickness and dynamics, precipitation, evapotranspiration, water runoff, and ocean and land surface tem- peratures and albedos. The reader should consult A Strategy for Earth Science from Space in the 1980's and 1990's, Part If: Atmo- sphere and Interactions with the Solid Earth, Oceans, and Biota (National Academy Press, 1985) for the recornrnended measure- ment requirements for each of these parameters. Measurement of Soil Properties Smart ground stations wiD help determine the relationship between climate, vegetation, soil moisture, and topography. They will also assist in our understanding of the effects of changes in land surface evaporation, albedo, and roughness on local and global climate. Long-term data sets for the following soil-related parameters wall be required: soil types and areal extent, moisture of the surface and root zones, texture, color, elemental storage, temperatures, infrared emissivity, and albedo. The recommended measurement requirements are set forth in the National Academy Press publication cited above. Simple Grownd ~staBatiom In addition to those measurements requiring the high-data- rate facilities of the smart ground stations, simpler, more tradi- tional measurements at a large number of sites will still be nec- essary. There may be as many as 10,000 of these installations, as suggested in Table 5.1. One example is given here. Magnetic Field The magnetic field has been monitored at several dozen sites around the world for more than a century. However, this distri- bution is very nonuniform, and consequently inferences of lateral heterogeneities in the rate-of-change in the magnetic field have been quite uncertain. This problem would be appreciably relieved by a magnetic monitoring satellite, a relatively high altitude (up to 1000 km) spacecraft with a lifetime of decades. Because of temporal fluctuations generated by ionospheric currents, however,

OCR for page 87
96 there will still exist aliasing of internally generated changes in- ferred from a single orbiter. Hence if, as proposed here, there is established a global set of ground stations, it would be desirable to include at least 100 magnetometers as uniformly spread as feasible (i.e., at 200~km intervals). This modest investment would greatly enhance the value of a magnetic monitoring satellite. Ocean Bottom Stations In this version of the ocean bottom geophysical observatory, the signals would be transmitted by a fiber optic cable to the nearest island, from which the data would be transmitted by satellite. Whether the station processor is deployed on the ocean bottom or on the island will depend on the power requirements. Generally, the capacity of the processor should be similar to that of an SGS, except that if power consumption considerations should so indicate, part of the necessary computations could be performed at the data collection center. A primary purpose of this type of observatory would be to monitor seismic activity on the ocean floor, including the struc- ture of earthquake source mechanisms, strong ground motions in seismic areas, volcanic activities, and tidal motions. The seismic stations would have three-component, broad-based instrumenta- tion capable of resolving 10-9 m/s2 acceleration at a period of 25 s. Experimental measurements indicate that seismic noise in a deep ocean borehole is comparable to the noise level at quiet sites on land. There is therefore no reason to lower the standards established for the land observations. Strong-motion instruments deployed in seismically active areas should also be broad-band type and remain on scale for acceleration above 1 g. Data rates should be identical to those specified for SGS. Additional uses of these stations could be to measure a variety of ocean bottom characteristics such as pressure, temperature, and chemistry. The data rate for these could be 1 sample per 10 s, which would pose a negligible contribution to the data flow. Moored Buoys Moored buoys provide the means for measuring properties of the ocean at a fixed point. They can be combined with the ocean bottom stations described above. As with the stations, and

OCR for page 87
97 depending on power requirements and proximity to land, there may be direct relay of the data by satellite, or a link to the nearest land mass may be established by fiber optic cables. The buoys could be used primarily for oceanographic pur- poses, as they are at present. They could characterize processes below the sampling depths of satellites, vertical mixing rates of the surface layer, exchanges between the surface and deep water, and mesoscale features. Surface measurements would continue to be used for "ground truth" verification of satellite data. Seisrn~c measurements, sirn~lar to those of the ocean bottom package, could be performed as well. Floating Buoys These devices, in addition to the oceanographic capabilities described in connection with moored buoys, allow for mapping of ocean currents. Because their position changes with tune, they must be able to be easily located. This can be accomplished with an appropriate space-based system such as ARGOS. Drifting buoys can be at the surface, in which case they measure surface air and sea parameters as well as the ocean currents, or they can float at a given density level below the surface, monitoring tempera- ture and currents. The subsurface buoys are programmed to come to the surface periodically to report their position and data they have collected previously hence they are called "pop-up" buoys. While a full seismic package cannot yet be deployed, a hydrophore could provide information useful in locating earthquakes. Techni- cal details still need to be developed. Ships: Research and Voluntary Observing There are about 300 oceanographic and marine geophysics re- search ships operating today, and there are many others that are willing to take measurement on a not-to-interfere basis (voluntary observing ships). They are usually on special research expeditions that collect data and that may or may not be linked into the satellite data system. These can be considered additional sources of observations that can provide such basic data as temperature, salinity, precipitation, winds, wave data, and sea state. Modules can be designed for shipboard use that automatically collect and

OCR for page 87
98 transmit these data, or they can be treated as manned observato- ries. Global Geodetic Observations at the Centaneter [eve] The Global Positioning System (GPS) that is being deployed by the DOD provides tremendous opportunities for accurate geode- tic measurements on a worldwide bash. For ordinary surveying purposes, inexpensive instruments will greatly reduce costs after 1988. For the geophysical purposes of intent here, the limiting factor is currently tropospheric resolution. However, by 1995 it can be expected that water vapor radiometers, or substitutes for them, will be sufficiently inexpensive to enable widespread cam- paigns with the GPS system. For the primary purpose of providing data on geodetic displacements in areas of tectonic activity where most earthquakes occur, a dense network of control points at in- tervals of 1 to 30 km (dependent on site) is indispensable if the unraveling of complicated strain rate patterns ~ to be feasible. Accuracies of a few parts in 108 (e.g., a function of 1 mm over 10 km) should be achieved by the turn of the century. The GPS system can also provide data on inflation of volcanoes in order to instrument the volcanic eruption hazard on a worldwide basis. Secondary applications on land include measurement of reclining tectonic plate motions, mass loss of glaciers and ice caps, and post- glacial rebound. Observations on the oceans and their boundaries can provide data on global changes in sea level associated with changes in ice volume or other causes, ocean tides, and ocean currents. There is a need for permanent stations, including moored buoys, that monitor geodetic positions continuously and also for mobile stations that can monitor position on closely spaced net- works in areas of concern, such as the San Andreas fault in Califor- nia. In addition, the GPS system will provide accurate locations for satellites, aircraft, and ships, and thereby greatly improve their measurement capabilities. The development of precise GPS receivers is revolutionizing the field of geodesy. A large number of high-accuracy measure- ments of vector position and baseline length can be made economi- cally with such receivers. Given a sufficient number of constraining measurements, major progress in understanding local and regional

OCR for page 87
99 crust al deformation and perhaps plate driving forces should be possible. Receiver performance and uncertainty in the orbit ephemeris currently limit GPS accuracy, but both will improve significantly over the next several years. The limiting error source in the near future is thus expected to be uncertainties in signal delay related to propagation through a variable atmosphere. Improvements in water vapor radiometers (WVR) and improved atmospheric mod- els are required before ultra-high-precision GPS geodesy ("super GPS") becomes a reality. It may well be that a series of GPS and WVR observations stretching over several years in a variety of climatic zones will be required to separate systematic errors from plate motion and to fully understand atmospheric effects on GPS location techniques. Continuous observations would be necessary in order to characterize the dynamic effects of weather with periods of several days, in addition to longer period seasonal atmospheric effects. Permanent GPS facilities at key sites would therefore be required. The geodetic application of GPS to obtain differences of po- sition interferometrically is a by-product use not anticipated by the DOD. Eventually, it can be expected that a system with an optin~zed signal and a spacecraft easier to model for radiation pressure will become desirable for geodetic purposes. For certain areas, such as the San Andreas fault region, it may be useful to have some GPS receivers operating frequently or continuously at fixed points in an automatic or semiautomatic mode, similar to tide gages now. This mode would reduce errors due to local disturbance of monuments and imprecision of site reoccupation. But the major application of GPS to geodynaniics will be by mobile systems, reoccupying a network of sites at in- tervals of months to years. For this mode, it may be that a more economical system of comparable accuracy will be laser ranging from spacecraft to retrorefiectors on the ground. This technique should be developed, and then the costs should be compared with those for the GPS mode. It has now been confirmed from analysis of satellite ranging involving several tectonic plates that relative motions over an 8- year time base agree within 1 cm/yr, with plate motion rates based on sea floor magnetic lineations an a 105-year time scale. It can be expected that this picture will be greatly refined by 1995. Hence,

OCR for page 87
100 effort in the period 1995 to 2015 will be much more focused on regions of known deformation in the Holocene era. A more speculative aspect of geodetic measurements for geo- dynam~cs research is the ability to monitor the position of sea floor geodetic points. This could be done acoustically using three or more precision transponders mounted permanently to the sea floor, and a central surface platform such as a buoy for transponder interrogation and travel time measurement. The surface platform must be positioned relative to the nearest land via GPS, so that in effect the platform would act as an air-water transfer point for positioning. For several possible deployment platforms, particu- larly smaller, more economical buoys, the GPS unit would need to operate untended for the duration of the experiment. The ability to operate for longer periods in an untended mode would reduce the number of expensive ship visits for data transfer. The ma- jor limitation on the operational life of such an experiment would then be the lifetime of battery packs for the sea floor transponders, currently 2 to 3 years for such low duty cycle applications. Thus, a number of considerations suggest that a key part of NASA's program in solid earth geophysics should include devel- opment of a GPS-based geodetic system capable of operating at remote sites for long periods (months to years) in an untended, fully automated mule. The benefits that would accrue from such a development include unproved atmospheric models and accu- racy, reduced cost of field operations, a better description of the temporal spectrum of plate motion, and the ability to carry out precise location of surface buoys in support of a sea floor geodetic program. Land GPS stations could be deployed advantageously with the global digital seismic array as well as with many other types of instrumentation because they can share a common data transfer system. While GPS measurements will become more focused on tec- tonic problems that are more precisely defined by a context of geologic and seismic surveys, a need for monitoring of global-scale motions will remain. For this purpose, VI,BI is becoming preem- inent because it has all-weather capability and a multiplicity of sources. Lunar laser ranging wall continue to be of value to moni- tar systematic error sources and study the Moon's orbit. V~BI Is also able to provide orientation and precise differences of location to timing and factors determining GPS orbits.

OCR for page 87
101 The PLATO System Once in place, the Permanent Large Array of Terrestrial ON servatories (PLATO) system can be viewed as a gigantic terres- trial observatory, which, in addition to its normal functions, can be treated as a telescope or accelerator. It can be reconfigured for special preplanned experiments and preprogrammed for "targets of opportunity," such as volcanic eruptions or tsunami tracking, that require, for example, higher data rates for certain periods of time. The system and potential experiments can be upgraded as technology improves. Many of the technology needs for PLATO have been identified in previous sections and in other studies, particularly the NASA report, Earth Obsermng Systems. To summarize, these needs will require developments in two broad areas: data handling and gen- eral systems design. Among the improvements in data handling that will be necessary are random access mass storage devices of considerably greater capability than currently exists, better data relay capabilities, and "user friendly" data centers. These require- ments are discussed in more detail below. Systems design efforts will have to focus on a variety of unmanned remote observatories and their instrumentation, as discussed above. DATA MANAGEMENT AND ANALYSIS Implementation of the scientific objectives and measurements of this strategy will give rise to a new set of data problems. Such problems will be the result of the generally increasing complexity of measurements and magnitude of data over the period for which the strategy is directed, and the global nature of the strategy. Adoption of this strategy will impose the significant requirement that the data chain, from observation to interpretation, must be well conceived and elective. An overview of the measurement requirements specified in the discipline areas indicates a number of common data issues. It is apparent that over the next 20 to 30 years the increase in data volume will be due largely to meeting measurement requirements for high spatial and spectral resolutions, repetitive measurements, measurement of long-duration phenomena, and measurement of many short-duration phenomena with large-scale effects. The ex- pected availability of advanced instrumentation, especially those

OCR for page 87
102 that will provide more spectral bands and those that will produce large amounts of data, such as synthetic aperture radar, will add to the overall data volume. As the science represented in these strategies matures, it is reasonable to expect that the complexity of measurement require- ments and data interpretation will also mature. New sets and kinds of data will be requested, especially for the intercomparison of measurements of common sites or the comparison of widely sep- arated sites. In this regard, the task group wishes to call particular attention to the area of data interpretation. The present ability and tune required to translate space observations into analysis and understanding lag well behind the technology to build measure- ment devices and to collect data from space. To a large degree this lag is the result of overemphasis on an empirical approach to re- mote sensing. The effect has produced great delays not only in the interpretation and scientific understanding, but also in our ability to guide future measurements and instrument development based on the results of interpretation techniques. For this strategy, a ma- jor effort must be made in the development of new interpretation techniques in remote sensing for geology if the primary scientific objectives are to be met. Beyond the need to reduce remote sensing data and to archive them in a common format is the need to apply a variety of anal- ysis and image-processing techniques, to select among the data sets, and to combine the most appropriate data in a mult~spectral approach. An ongoing research effort should be maintained to develop optimum interpretational techniques for specific geologic problems. In oceanography, a similar research program in data interpretation techniques should be supported for new data ac- quisition systems. The development of such techniques must be accelerated to ensure that space measurements for earth science will produce the maximum scientific return within reasonable time periods following data acquisition. The science objectives put forth here place no severe demands on U.S. capability to obtain data. There is no apparent technolog- ical barrier to the acquisition of data implicit in this strategy. The new precedent for data demands lies rather in the organization and management of data acquisition and analysis systems, and these new requirements are due in large measure to the now global extent of the earth sciences. Driven by science requirements for ever more accurate and

OCR for page 87
103 more detailed remote sensing observations of the Earth, NASA has historically invested the lion's share of its resources in devel- oping sensor, instrument, and platform technologies. Individual scientists have developed whatever methods they required and have acquired analysis toob from developers on an ad hoc basis. The scope and complexity of current and proposed remote sensing instruments, such as unaging spectrometers and synthetic aper- ture radar systems, promise to dramatically change the situation with respect to the science data analysis needs. Today's scientist must not only cope with a greater variety and complexity of data, but must also attempt to understand and utilize computer hard- ware and software technologies, which are themselves undergoing a phenomenal growth in capabilities. It is unreasonable to expect individual scientists to develop strategies and evaluate available resources for managing and analyzing all these data. If they have to do this, they wiD spend most of their time attempting to keep up with technology and have little time left to do science. The earth science community feels that the tune has come for NASA to play a significant role in providing the scientists with a program that will yield a coherent integration of data analysis and data management technologies. Although current support within NASA for data analys~s/management technology is not insignificant, it Is severely fragmented among a large number of different programs. In many cases the needs for data analysis are seen as only secondary to the goal of a particular program or project. NASA currently has a strong ongoing program in space science data systems and in spigot data systems" within its Information Systems Office. NASA also supports the NASA Ocean Data Sys- tem (NODS) at JPL. As a result, many of the relevant issues such as data base management, storage, and communications technol- ogy are being addressed. The NASA EOS report recommends a further strengthening of these activities to include research, de- velopment, and implementation in the data analysis technology areas of (1) algorithm development and applied mathematics for analysis of large multispectral, mult~sensor, remote sensing image data and (2) utilization of the emerging technology of concurrent processing, with particular emphasis on high performance/cost ratio approaches, which hold the potential for putting large com- putational resources directly within reach of individual scientists. .

OCR for page 87
104 An interagency study is one mechanism by which the require- ments of a global data management for the next 25 years can be deterrn~ned. This study should include at least the following ele- ments: data processing, distribution, storage, and retrieval; data analysis needs; and computer capabilities. The task group suggests that this strategy for earth science and the CODMAC reports be utilized as a science requirement guideline for the study. In the interim, the task group urges the relevant agencies to begin to ar- ticulate the data issues and define the framework for establishing management and organization systems. INTEGRATION OF RESULTS FROM OBSERVING SYSTEMS The many experiments and instruments recommended in this chapter are difficult to relate directly to the major issues and themes set forth in Chapter 3. In this section the task group briefly reviews these leading themes and suggests how the mea- surement programs would contribute to unproved insight. In a few cases, the programs will generate specific diagnostics for certain questions. But in most instances their contribution is to give a richer context to th_ ongoing scientific thinking about the Earth, which stimulates important new ideas the most important often being the most unpredictable. Studies of the solid Earth will be advanced most significantly by the geodetic and seismological systems proposed in this chapter. Detailed complexes of geodetic measurements and seismometers will give a mapping of crustal motions in zones of deformation, such as the San Andreas, necessary to a greatly improved insight into the occurrence of earthquakes and other processes that de- termine the structure of the lithosphere. Practical by-products of their insight will be effective earthquake prediction and ciarifi- cation of the tectonic contexts of mineral formation. On a more globally interconnected scale, seismological networks with satellite gravimetry will produce more detailed maps of density variations throughout the mantle. These maps, coupled with computer mod- els, will determine the nature of mantle convection, the spectrum or length scales of its flows, the degree of interaction among dif- ferent parts of the mantle, and so forth. The programs of space measurements and associated obser- vations will give a more detailed description of a wide range of phenomena at both regional and global scales. The descriptions

OCR for page 87
105 will enable greatly improved short-term predictions and insights into ongoing processes. In several respects, the models for these processes will be largely separate. In particular, this separation ex- ists between solid earth processes on the one hand, and oceanic and atmospheric processes on the other. The coupling of these different subsystems occurs on a much longer time scale: millions of years. The nature of the hydrosphere and atmosphere depends on the outgassing of the mantle. Hence, a more fundamental problem is the long-term evolution of the mantle, which will depend strongly on convection that in itself was, going back in time, increasingly different from the contemporary convection whose manifestations are measured by current geodetic and seismological systems. The global data sets to be collected through EOS and PLATO, together with modeling on high-speed supercomputers, will allow for the first time the integration of processes on many time scales and from different disciplines. Only in this way can we address the wide disparity of time and space scales represented by geophysics, geochemistry, fluid dynamics, and biological processes on Earth.