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 18
Spectrum Management for Science in the 21st Century 2 The Earth Exploration-Satellite Service In 1960 the first weather satellite dramatically opened humanity’s eyes to the beauty and complexity of Earth’s atmosphere. Never before had anyone photographed a hurricane’s movement or cyclonic shape (see Figure 2.1) or observed the global form of atmospheric waves on a planetary scale. After proving the usefulness of orbiting weather observations, NASA and the National Oceanic and Atmospheric Administration (NOAA) began developing ever more sensitive and innovative space-based instruments that help people understand the natural world around them and their impact on it (see Box 2.1). Modern observation systems offer economically and societally important forecasts extending farther into the future than ever before, but these advances depend on protected radio frequency allocations. With the development of more advanced instrumentation, it quickly became clear that there were great opportunities to observe at wavelengths other than what is usually called light. In fact, visible light is now only a small part of the story. Most current satellite sensors also observe terrestrial emissions at infrared and/or radio wavelengths. These environmental applications have evolved over the past 50 years by combining radio astronomy and geophysical techniques to form the new scientific field known as microwave remote sensing. Human eyes evolved to detect visible light because Earth’s atmosphere allows solar radiation to pass through an “atmospheric window” at those wavelengths. In the same way, the “eye” of the satellite (the receiver) is designed to view Earth through atmospheric windows at other wavelengths, rather than observing reflected sunlight as human eyes do. Analogous to what infrared goggles (providing heat
OCR for page 19
Spectrum Management for Science in the 21st Century FIGURE 2.1 Hurricane Camille as it approaches the Gulf States in 1969, as photographed from the NASA Nimbus III satellite. Image courtesy of NASA/Nimbus III Satellite. vision) do, most satellite instruments detect the inherent emission of radiation (heat) from the atmosphere and terrestrial surface at wavelengths that reveal details invisible to human eyes. When the atmosphere itself is of interest, opaque wavelengths that do not pass through the atmosphere but are absorbed by it offer more information. Each window and opaque band responds differently to the various properties of the terrestrial surface and atmosphere, allowing those properties to be studied by a simultaneous analysis at multiple frequencies. The accuracy of these studies increases with the number of observed frequencies. The unique ability of passive microwave sensors to “see through” most clouds makes those sensors essential, particularly where clouds are persistent. The sensors are passive in that they do not transmit signals but instead only receive the natural background emission. Scientists can thus extract information from the radio spectrum on environmental properties as varied as atmospheric temperature and humidity, precipitation rate, soil moisture, ocean salinity, and ocean waves (and therefore surface winds and
OCR for page 20
Spectrum Management for Science in the 21st Century BOX 2.1 The Origin of Earth Remote Sensing Earth remote sensing techniques have developed over many years, evolving out of astronomy and accelerating as satellite technology became more robust. Before 1932: Use of optical astronomy (initial passive spectral observations of stellar and planetary surface and atmospheric temperatures and compositions, demonstrating basic methods). 1932: The first radio astronomy observations by pioneer radio astronomer Karl Jansky, revealing cosmic radiation. 1940-1945: Wartime studies of centimeter- and millimeter-wave atmospheric absorption spectra and passive radiation; the development of sensitive radiometry. 1968: Launch of the first passive microwave radiometer on the Soviet Cosmos-243 satellite—it observed sea surface temperature, land temperature, snow/ice cover, water vapor, and liquid water using four unscanned window channels at 3.5-37 GHz (unfortunately short-lived, operating only for weeks). 1972, 1975: The first long-lived satellites to image window-channel parameters (humidity over ocean, sea ice, ocean roughness and wind, snow cover, precipitation, land temperature, etc.) and atmospheric temperature profiles: Nimbus-E Microwave Spectrometer (NEMS; two window channels and three opaque channels) and Electrical Scanning Microwave Radiometer (ESMR) imaging at 19.36 GHz launched on the NASA Nimbus-5 satellite in 1972, and the Scanning Microwave Radiometer (a wide-swath imaging version of NEMS) and the dual-polarized ESMR imaging at 37 GHz launched on Nimbus-6 in 1975. 1978: The first operational weather satellites to incorporate imaging passive microwave spectrometers for temperature sounding (microwave sounding unit with four opaque-band channels at 50-58 GHz on TIROS-N and NOAA-6 and -7). 1987: The first operational satellites to monitor surface parameters and atmospheric water (Special Sensor Microwave/Imager with seven window channels at four frequencies, 19.35-85.5 GHz, first launched by the Defense Meteorological Satellite Program). Post-1987: Launch of continually improved research (NASA) and operational (National Oceanic and Atmospheric Administration and Department of Defense) passive microwave instrument types. ocean internal waves). The full global coverage provided by satellites enables scientists to monitor Earth’s environment far more accurately and completely than had been possible using traditional means such as weather stations and balloon sounders. Satellite data have also greatly improved the accuracy of weather forecasts and enabled sensitive, large-scale climate studies revealing, for example, the effects of ozone-modifying trace gases. Figure 2.2 presents a typical image of the abundance of water vapor over the oceans as observed by combining observations in multiple frequencies obtained by the Advanced Microwave Scanning Radiometer-Earth (AMSR-E) imaging passive microwave spectrometer.
OCR for page 21
Spectrum Management for Science in the 21st Century FIGURE 2.2 Advanced Microwave Scanning Radiometer-Earth (AMSR-E) data showing tropospheric water vapor abundance over Earth’s oceans, denoted by the colors in the image. Land is denoted by shades of gray, its shade depending on the elevation; sea ice is denoted by white. AMSR-E data are produced by Remote Sensing Systems and sponsored by the NASA Earth Science MEaSUREs DISCOVER Project and the AMSR-E Science Team. Data are available at www.remss.com. Today, the United States operates a suite of more than 30 satellites that measure Earth’s planetary environment and collectively represent many billions of dollars invested by U.S. taxpayers. The significance of the passive radio services is suggested not only by the substantial government investment in their development and operation, but also by their impact on the national economy. The environmental products facilitated by the passive services are critical for day-to-day, long-term, and severe weather forecasting and also for the Department of Defense (DOD) and for the energy, agriculture, and transportation industries.1 The U.S. investment in passive Earth observatories provides the nation with a high degree of economic leverage over environmental events. 1 The 2006 report Economic Statistics for NOAA states that “weather and climate sensitive industries, both directly and indirectly, account for about one-third of the Nation’s GDP in sectors ranging from finance, insurance, and real estate to services, retail and wholesale trade and manufacturing. Industries directly impacted by weather such as agriculture, construction, energy distribution, and outdoor recreation account for nearly 10 percent of GDP.” National Oceanic and Atmospheric Administration, Economic Statistics for NOAA, Washington, D.C., 2006.
OCR for page 22
Spectrum Management for Science in the 21st Century On a larger scale, Earth’s climate is deemed so important to humanity that the 2007 Nobel Peace Prize was awarded to the Intergovernmental Panel on Climate Change (IPCC) and Albert Gore, Jr., “for their efforts to build up and disseminate greater knowledge about man-made climate change, and to lay the foundations for the measures that are needed to counteract such change.”2 The prize was based on the laureates’ assessment that large-scale climate change would irrevocably alter living conditions in many places in the world and thus lead to wide-spread civil unrest. Consistent with this assessment of the importance of climate to humanity are estimates that the potential consequences of global change in its various manifestations (sea ice loss, global warming and drought, coral bleaching, tropical ecosystem collapse, and other interrelated environmental problems) would be associated with unprecedented societal costs to the United States and the world.3 These staggering costs demand that the most precise information on global environmental processes be made available to decision makers grappling with questions of environmental policy. The precision of this information and the overall understanding of climate change are driven by both observational science and improved understanding and models of the environment, which in turn are dependent on the availability of spectrum for use in environmental observation. At stake are potential measures including limits on emissions of greenhouse gases such as carbon dioxide and methane, limits on aerosols and chlorofluorocarbons, restrictions on deforestation and freshwater usage, and stiff requirements for agricultural and manufacturing practices and the transportation industry. It is also useful to note the educational value of government programs that apply radio science to environmental problems. These programs are largely conducted either through or in collaboration with universities and thereby train many graduate students at the cutting edge of both radio- and microwave-frequency technology and Earth science, thus contributing to economic sectors critical to U.S. global competitiveness and the defense of the nation. The importance of environmental radio services has increased in parallel with the use of public and commercial wireless and other electronics technologies (see discussion in Chapter 4). Collectively there has been a substantial increase in the number of human-made radio signals that can interfere with and corrupt needed scientific and operational passive observations of the environment.4 The commoditization of wireless and other electronics technology has significantly increased the pressure on the passive uses of the spectrum in terms of allocations and disruptive 2 Available at http://nobelprize.org/nobel_prizes/peace/laureates/2007/; accessed August 26, 2008. 3 Intergovernmental Panel on Climate Change, Working Group II Report, Impacts, Adaptation and Vulnerability, Geneva, Switzerland, 2007. 4 Scientific observations are those conducted for research purposes. Operational observations are conducted in consistent, repeated ways for use in products such as weather forecasts.
OCR for page 23
Spectrum Management for Science in the 21st Century interference. As quickly as techniques have been developed to mitigate human-made interference, they are eroded by other expanding active uses of the spectrum. Moreover, as the spectral efficiency of wireless technology improves, the interference that it produces increasingly resembles random noise, which is more difficult to identify and mitigate. These difficulties are compounded by the increased use of spectrum licenses that permit unlimited numbers of approved devices to be used, with decreasing means for enforcement or further mitigation. Section 2.5 discusses these difficulties in a variety of circumstances. Most active services can use coding techniques, better antenna systems, and higher-power transmitters to survive even high levels of interference, but these techniques are not applicable to passive services. There is a fundamental asymmetry between the spectral requirements of active communications services and passive environmental uses. Advances in wireless technology are rapidly increasing the abilities of competing communications services to share radio spectrum through agile time-frequency multiplexing, whereas the measurement precisions of the passive services are intrinsically limited by the strength of natural emissions, the reception bandwidth, and the observing time. The competition for radio spectrum also has global implications, as the U.S. environment is affected by environmental conditions in other nations and vice versa. Both U.S. and foreign environmental satellites fly over almost the entire globe and continuously observe within the same spectral bands everywhere; thus critical environmental radio bands need to be uncontaminated everywhere. The data from these diverse, Earth-orbiting, multinational assets are increasingly being shared in the global public interest, which parallels the separate national interests, and can be obtained by no other means. Furthermore, the national character of environmental services and the multidecadal times required for their development and use in space make them much less nimble than the private sector that can develop new radiating products in a period of months. It has therefore become clear that a new look at spectrum policies and regulations is necessary in order to protect the critical passive environmental observations by Earth observation satellites and to permit the passive and active services to coexist productively. This chapter discusses the reasons behind the need for new regulations, which are further elaborated on in Chapter 4. 2.1 SPECIFIC APPLICATION AREAS OF PASSIVE MICROWAVE REMOTE SENSING Earth remote sensing is critically important to the United States and to the advancement of human scientific knowledge about Earth and the environmental processes that support life and commerce. Microwave remote sensing, or the Earth Exploration-Satellite Service (EESS) in regulatory parlance, provides direct eco-
OCR for page 24
Spectrum Management for Science in the 21st Century nomic benefits to the nation by obtaining information that has economic value to both the public and private sectors. In addition, the collection of these data is a highly technical enterprise that strengthens the U.S. industrial, defense, telecommunications, and environmental sectors. The United States operates in a competitive, information-dominated economy that is dependent not only on having access to the passive spectrum for obtaining data for commercial, governmental, and public purposes, but also on having skilled engineers who are trained in the most sophisticated microwave engineering techniques. Passive and active remote sensing act in tandem to collect environmental information and ultimately to provide the benefits to society referred to above. Much of the data that lead to these benefits, however, is only available from passive microwave sensors, and these sensors have unique needs that must be met to enable the measurements that they make. For example, passive microwave remote sensing is indispensable for better numerical weather forecasting, large-scale monitoring of subsurface soil moisture, and so on, and improvements in weather forecasting are important economically and strategically. This section presents a sampling of applications in which passive access to the microwave spectrum is essential for the country. The discussion is organized in the following broad topics: weather forecasting and monitoring, severe weather and disasters, climate and global change, resource management, aviation, defense and pubic safety, international partnerships, and education and technology. The subsection on international partnerships includes discussion of a recent international effort, initiated by the United States and the Group of Eight (G8), to ensure that the nations of the world engaged in space remote sensing collaborate in exchanging data to benefit their societies. Weather Forecasting and Monitoring Satelliteborne passive microwave sensors are a critical part of the global weather monitoring system. Passive microwave sensors are particularly critical for measuring temperature, humidity, and precipitation profiles in the cloud-affected troposphere below approximately 10 km, where most economically important weather occurs, and in measuring sea surface winds and temperatures and soil moisture. Part of the reason for this importance is that weather radars measure only the reflectivity of water and ice droplets in the atmosphere but are insensitive to these other parameters. Even so, the extraction of useful information from radar reflectivity measurements relies greatly on knowledge of the droplets’ size distribution, which requires complex and costly multiband radar measurements to measure directly. Passive microwave radiometers, however, directly measure the total quantity of liquid water as well as water vapor and other variables. Such radiometers can herald impending weather events by measuring the presence of water vapor in
OCR for page 25
Spectrum Management for Science in the 21st Century advance of cloud formation and then detect the formation of liquid water droplets well in advance of the detection by rain radars. Moreover, when used in conjunction with weather radars, passive radiometers provide a high degree of precision in the measurement of the path- or area-averaged quantities being observed that serve to calibrate the radar’s signal. In this manner the radiometer is able to facilitate the radar’s capability to provide high resolution. Radars are thus useful in conjunction with radiometers but not as a substitute for them, as exemplified by the recent Tropical Rainfall Measuring Mission (TRMM) and the CloudSat and future Aquarius and Soil Moisture Active Passive (SMAP) missions. Modern weather forecasts are based primarily on numerical weather prediction (NWP) models run on massively parallel computers. These models use direct data assimilation (DDA), a powerful technique developed during the past two decades that incorporates all available data from satellites, balloons, radars, and surface stations to steer NWP models. Major worldwide centers developing and operating these models are located in the United States, Europe, Canada, China, Japan, and Australia. Their algorithms, from the beginning, have relied heavily on passive microwave measurements of relevant environmental variables, and they will continue to do so as spatial and temporal resolutions improve. Passive microwave data in the opaque temperature-sensitive bands above 50 GHz have been particularly helpful because of their insensitivity to most clouds; these observations probably constitute the single most valuable data source currently enabling 1-week weather forecasts. The demand for improved space and time resolution has been relentless since the inception of NWP modeling in the 1970s and is expected to continue for the foreseeable future, particularly as wireless devices enabled by the Global Positioning System increase the demands for ever more site-specific, personalized information on weather. In recent decades, the accuracy and usefulness of weather forecasts have increased tremendously because of progress in both NWP systems and satellite-based remote sensing systems. Figure 2.3 illustrates this progress in terms of the number of days for which forecasts of a given quality are obtained. For the highest-quality Southern Hemispheric forecasts, satellite data increase the forecast from 12 hours to 2 days—a factor of four—and for an anomaly correlation of 0.6 the forecast doubles from 3.25 to 6.5 days. The anomaly correlation is a common measure of forecast accuracy, with values above 0.6 generally considered to be significant. Much of the improvement in forecasting shown in Figure 2.3 is due to the direct use of passive microwave data on their own, and to the integration of microwave and infrared data that combine the best features of both sensor types. Surface wind data over the ocean derived from spaceborne microwave measurements have also been helpful. These improvements are particularly striking in the Southern Hemisphere where data from surface stations and balloon soundings are sparse, but they also extend forecasts in the Northern Hemisphere by roughly
OCR for page 26
Spectrum Management for Science in the 21st Century FIGURE 2.3 Anomaly correlation for days 0 to 7 for 500 hectopascal geopotential height in the zonal band 20°-80° South for August/September. The red and blue arrows indicate that the use of satellite data in the forecast model has doubled the length of a useful forecast (i.e., a forecast with anomaly correlation = 0.6). Image courtesy of NOAA. 25 percent. Passive microwave sensors are also useful for tracing the movement of water through normal weather cycles. For instance, surface soil moisture, snow cover, and snow-water-equivalent drive energy exchange with the atmosphere and therefore affect weather forecasts. The major impact of these surface variables on forecast accuracy is just beginning to be seen (Figure 2.4). Brief discussions of a few specific weather-monitoring topics follow. Soil Moisture Accurate knowledge of the parameters of soil moisture (SM) has been shown to improve forecasts of local storms and seasonal climate anomalies. In Figure 2.5, panel (C) shows the observed difference in rainfall between two extreme years, the flood year of 1993 minus the drought year of 1988, over the middle of the United States. Current atmospheric models tend to use sea surface temperatures (SSTs) as their primary boundary condition because so much of Earth’s surface is
OCR for page 27
Spectrum Management for Science in the 21st Century FIGURE 2.4 A depiction of the impact of observations of soil moisture (left) on 12-hour rainfall forecasts that use Weather Research and Forecasting models (for June 12, 2002). Panels at right: Forecasts with and without the Land Information System (LIS) providing improved soil moisture initial and boundary conditions. Image courtesy of NASA. FIGURE 2.5 The value of soil moisture data to climate forecasts. Predictability of seasonal climate is dependent on boundary conditions such as sea surface temperatures (SST) and soil moisture—soil moisture being particularly important over continental interiors. In the results of a simulation driven only by SST (panel A), the climate anomaly in panel C (observed difference in rainfall between the flood year of 1993 minus the drought year of 1988) is not reproduced. Results of the simulation driven by SST and soil moisture (panel B), however, accurately predict this seasonal anomaly. SOURCE: D. Entekhabi, G.R. Asrar, A.K. Betts, K.J. Beven, R.L. Bras, C.J. Duffy, T. Dunne, R.D. Koster, D.P. Lettenmaier, D.B. McLaughlin, and W.J. Shuttleworth, “An Agenda for Land Surface Hydrology Research and a Call for the Second International Hydrological Decade,” Bulletin of the American Meteorological Society, 80(10): 2043-2058 (1999). ocean. However, models just using SSTs do not do a good job of capturing seasonal climate anomalies in the middle of large continents. As seen from the results in Figure 2.5(A), the climate anomaly is not reproduced. However, if SM data such as those derivable from space-based 1.4 GHz passive microwave measurements are incorporated, atmospheric models can accurately predict the seasonal anomalies in
OCR for page 28
Spectrum Management for Science in the 21st Century FIGURE 2.6 Soil moisture data improve numerical weather prediction over continents by accurately initializing land surface states. In this example, 24-hour prior forecasts of a high-resolution atmospheric model of rainfall are shown without (A) and with (B) soil moisture input data. The observed data are shown in panel C. Provided by the National Snow and Ice Data Center. extreme weather, as seen in Figure 2.5(B). In the second example of the importance of SM data (Figure 2.6), NWP can be improved over the continental United States by more accurately initializing the land surface state with soil moisture data. Soil moisture is also a key parameter in forecasting relating to agriculture, drought, and flooding and for predicting vegetative stress and establishing related government policies. Passive microwave radiometers operating at frequencies of 10 GHz and lower are sensitive to variations in soil density, type, and moisture content and are needed for SM measurements. Radiometry in the 1-2 GHz range is arguably the best means for measuring subsurface soil moisture on a national or global basis. Sea Surface Winds Global sea surface wind data are critical for high-quality NWP forecasts, for developing data pertinent to tropical cyclone warnings, aircraft and ship operations, ship routing, and other civil and military operations. Sea surface wind data constitute one of the most important parameters in operational meteorological remote sensing. Space-based remote sensing of sea surface wind vector (SSWV) depends on precision measurements of polarimetric microwave emissions from the ocean surface. These measurements have been shown to improve the forecasting capability of NWP models significantly, thus contributing to maritime and coastal safety and commerce. The accuracy of the wind vector products obtained from the Naval Research Laboratory’s (NRL’s) WindSat retrievals to date has reached or exceeded the accuracy of the wind vector products available from active scatterometer systems such as QuikScat at moderate to high wind speeds. Also, the ability of microwave radiometers to measure simultaneously atmospheric and sea temperature properties motivates attempts to improve further the accuracy
OCR for page 77
Spectrum Management for Science in the 21st Century FIGURE 2.22 An example of interference to EESS observations of opportunity at 6.6 GHz. Passive microwave imagery at 6.6 GHz from the Scanning Multi-channel Microwave Radiometer (SMMR) from (A) 1979 and (B) 1987, showing no noticeable brightness temperature from radio frequency interference (RFI). In contrast, passive microwave imagery from the Advanced Microwave Scanning Radiometer-Earth (AMSR-E) on NASA Earth Observing System Aqua from 2003 (C) and 2004 (D) shows substantial RFI. The black spots represent high levels of anthropogenic emission that saturate the AMSR-E radiometer, primarily over regions of California and Arizona. The red spots over most of the remaining areas of the United States represent contaminated brightness-temperature measurements. AMSR-E data are produced by Remote Sensing Systems and sponsored by the NASA Earth Science MEaSUREs DISCOVER Project and the AMSR-E Science Team. Data are available at www.remss.com. large parts of the Middle East, Asia, and Japan, and even sophisticated statistical procedures could not adequately distinguish RFI from the background of natural brightness variability, nor filter it out in post-processing of the data.48 Because the 6.9 GHz RFI was so prevalent and difficult to identify and mitigate over the United States, this instrument channel was subsequently ignored in the global AMSR-E algorithm used for the production processing and data archiving of SM data. Reliance was instead placed on the higher-frequency AMSR-E channels that are less sensitive to SM. Over those parts of Europe and Japan where the 10.7 GHz channels were also affected by RFI, no AMSR-E soil moisture retrievals at all were possible. 48 Ibid.
OCR for page 78
Spectrum Management for Science in the 21st Century On a research basis (separate from the global production algorithm), it is still possible to use the 6.9 GHz brightness data for soil moisture retrieval over significant RFI-free global areas such as most of Africa, South America, and Australia. Extensive analysis of AMSR-E and WindSat data provides a clear picture and plausible explanation for RFI at C-band, but not in other parts of the spectrum. Other RFI surveys have been inconclusive, tied to a single location, and/or have not been able to provide much insight regarding the global status of potential RFI to EESS. The duty cycle, waveforms, emitter spatial distribution, transmitter power, and spectral utilization of the RFI need to be measured to effectively and optimally design RFI mitigation strategies into EESS radiometer systems and to further develop equitable spectrum usage policies. 49 In short, inadequate data on spectrum usage exist. The Federal Communications Commission’s (FCC’s) 2002 Spectrum Policy Task Force came to this same conclusion: More information, however, is needed in order to quantify and characterize spectrum usage more accurately so that the Commission can adopt spectrum policies that take advantage of these spectrum white spaces. Currently, no federal agency or other organization systematically measures temporal spectrum use.50 Finding: Better utilization of the spectrum and reduced radio frequency interference for scientific as well as commercial applications are possible with better knowledge of actual spectrum usage. Progress toward the goal of improved spectrum usage could be made by gathering more information through improved and continuous spectral monitoring. Such monitoring would benefit both the scientific community and commercial interests by allowing more efficient use of the spectrum for communications. Interference mitigation at C-band has been demonstrated on a limited basis and for particularly strong (and therefore relatively obvious) interference in airborne images of thermal emission at C-band.51 The radiometer and algorithm were designed to detect spectral variations that were not of natural origin by fitting the spectrum to a standard model, then rejecting channels that compromised the fit 49 J.R. Piepmeier, “Radio Frequency Survey of the 21-cm Wavelength (1.4 GHz) Allocation for Passive Microwave Observing,” in Proceedings of the 2003 International Geoscience and Remote Sensing Symposium (IGARSS), Toulouse, France, 2003, pp. 1739-1741; and presentation by Dennis Roberson, Illinois Institute of Technology, to the committee on September 29, 2007, in Irvine, California. 50 Federal Communications Commission, Report of the Spectrum Policy Task Force, November 2002, p. 10. 51 A.J. Gasiewski, M. Klein, A.Yevgrafov, and V. Leuskiy, “Interference Mitigation in Passive Microwave Radiometry,” Proceedings of the 2002 International Geoscience and Remote Sensing Symposium (IGARSS), Toronto, Ontario, Canada, June 24-28, 2002.
OCR for page 79
Spectrum Management for Science in the 21st Century FIGURE 2.23 Polarimetric Scanning Radiometer C-band maps from a swath segment observed during SP99 on July 14, 1999, over central Oklahoma: (A) raw calibrated brightness maps for front and back looks for four subbands and (B) interference-corrected maps using a spectral sub-band algorithm (A.J. Gasiewski, M. Klein, A.Yevgrafov, and V. Leuskiy, “Interference Mitigation in Passive Microwave Radiometry,” Proceedings of the 2002 International Geoscience and Remote Sensing Symposium [IGARSS], Toronto, Ontario, Canada, June 24-28, 2002). AMSR-E data are produced by Remote Sensing Systems and sponsored by the NASA Earth Science MEaSUREs DISCOVER Project and the AMSR-E Science Team. Data are available at www.remss.com. to this natural model. The techniques have proven effective at mitigating large-amplitude interference (Figure 2.23). However, they provide no guarantee that interference of amplitudes on the order of the system noise level can be detected and mitigated. Finding: There is currently inadequate protected spectrum in C-band and X-band for operational passive microwave observations of sea surface temperature, soil moisture, and ocean surface wind speed and direction. Finding: While unilateral radio frequency interference mitigation techniques are a potentially valuable means of facilitating spectrum sharing, they are not a substitute for primary allocated passive spectrum and the enforcement of regulations. Finding: Important scientific inquiry and applications enabled by the EESS are significantly impeded or precluded by radio frequency interference (RFI). Such RFI
OCR for page 80
Spectrum Management for Science in the 21st Century has reduced the societal and scientific return of EESS observatories and necessitates costly interference mitigation, which is often insufficient to prevent RFI damage. Potential Future Radio Frequency Interference and Its Impact on EESS Observations Ultrawideband Devices and Anticollision Radar (1-24 GHz) A major concern for future EESS observations is the proliferation of ultrawideband (UWB) devices that radiate over wide bandwidths at low power, typically in the 2-10 GHz and 22-27 GHz ranges. Automotive collision-avoidance radars that employ the entire 22-27 GHz range have recently been included on new vehicles and are becoming widespread. In particular, the FCC’s 2002 approval of the use of UWB devices in the 3-10.6 GHz band and of anticollision radar operation as Part 15 devices near 24 GHz has alarmed the EESS community.52,53 These sources produce broadband signals that resemble thermal noise, making them difficult to distinguish from natural emissions. The potential for large-scale market penetration of such devices further exacerbates the problem, particularly if they are permitted to radiate across protected frequency bands (particularly in the protected 1.400-1.427 GHz and 23.6-24.0 GHz bands). Emissions from UWB sources in these protected spectral bands present a serious problem, and action will need to be taken to prevent such emissions and limit the numbers of such devices. Scenarios involving RFI to EESS systems from multiple low-level emitters within the passband and footprint of EESS measurements must be analyzed on a cumulative basis as outlined in Appendix C. In these scenarios the maximum output power of each transmitter and their number per square kilometer are critical factors in EESS compatibility studies. Examples include UWB at 6 GHz and point-to-point transmitters near 57 GHz (see V-band scenarios later in this section). A study analyzing the impact of losing the protected 23.6-24.0 GHz channel suggests that although the ideal level of RFI in the band is zero, 0.03 K might be established as its maximum permissible value, which is equivalent to −126.84 dBm of RFI within a 500 MHz band.54 More serious is the fact that unless the RFI level is 10 K or more, the NWP applications cannot reliably flag the data as 52 See the Glossary in this report for a definition of a Part 15 device. 53 FCC Press Release, “New Public Safety Applications and Broadband Internet Access Among Uses Envisioned by FCC Authorization of Ultra-Wideband Technology,” February 12, 2002, available at http://www.fcc.gov/Bureaus/Engineering_Technology/News_Releases/2002/nret0203.html; accessed January 7, 2010. 54 S. English, Assessment of the Requirement for 23.6-24.0 GHz Observations for Weather Forecasting, Forecasting Research Technical Report No. 440, Exeter, U.K.: Met Office, 2006.
OCR for page 81
Spectrum Management for Science in the 21st Century erroneous, thereby degrading forecasts within and downstream of any regions where intermediate-level RFI is present. Such intermediate-level interference is difficult to detect with any confidence except in locations where its effects become extreme. Since automobiles are nearly ubiquitous over land and especially within populated regions where forecasts have the greatest economic value, the problem is endemic to users who rely the most on forecast data. This final point is sufficient to support the exclusion of all intended emissions near the protected EESS band, consistent with the intent of the original regulation. In addition, there is great concern for the future of EESS measurements of opportunity at C-band. This band covers much of the spectral region commonly used by EESS for measurements of sea surface temperature and soil moisture on an as-available basis. These measurements are critical for accurate weather forecasting, severe weather prediction, and drought prediction, among other applications. The wide proliferation of low-level UWB devices within C-band is a significant concern of the EESS operational and scientific communities (see Appendix C for the density of interferers analysis). Since RFI in EESS operations is cumulative, there is no protection from the impact of a high density of low-level emitters resulting from the strong market penetration of unlicensed products. In these scenarios, all mitigation techniques for AMSR-E and WindSat data would be rendered useless, and important future C-band observations would not be possible without mandatory bilateral mitigation strategies (as described in Chapter 4 of this report). It is instructive to contrast the scenarios at C-band for EESS, where a large number of emitters contribute to RFI within a single pixel of AMSR-E and WindSat data (especially over populated areas), with the RFI scenario outlined in Appendix D. In the latter case, the impact of RFI on EESS measurements from one or more radars is considered. For cases where only a few high-level emitters in adjacent bands are present (for example, in L-band radar RFI), the measured brightness temperatures are increased by spurious and/or OOB emissions. Such emissions contribute directly to the maximum allowed in-band emissions for EESS; however, the RFI is the result of a single emitter rather than the cumulative effect of many in-band emitters. Although current regulations—if enforced—could preclude the effects of cumulative in-band emissions on EESS systems operating in allocated bands (e.g. 1.400-1.427 GHz and 10.6-10.7 GHz), they are largely ineffective in their present form in limiting OOB and spurious emissions. In considering these scenarios, it should be noted that the present specifications on OOB and spurious emissions were established decades ago, before heavy use was made of bands adjacent to where critical EESS measurements are now conducted and prior to major advances in microwave signal processing and filtering technology. Considerations of new technologies must be made in reassessing the effects of and in regulating OOB and spurious emissions.
OCR for page 82
Spectrum Management for Science in the 21st Century Ground-Based Atmospheric Sounding (23.8 GHz, 31.5 GHz, 50-60 GHz, 89 GHz, 183 GHz) Ground-based microwave radiometers are being used increasingly for the temperature, humidity, and cloud liquid profiles in the lower troposphere for both nowcasting and forecasting. Thus, they are being incorporated into weather observing networks as a replacement and augmentation of the global radiosonde network. It is expected that RMS instrument errors in oxygen-band temperature profiling radiometer measurements will be as low as 0.2 K (or lower) in the future and that the nominal tolerable RFI level for these systems will be 0.02 K. For a typical five-channel 22 GHz to 30 GHz upward-looking water vapor profiling radiometer, 1 K of RFI in a channel near the center of the 22.235 GHz water vapor line can induce a 10 percent error in retrieved water vapor abundance in the lower and mid-level troposphere. This error is comparable to the current performance of such a profiler, and the tolerable RFI level is therefore about 0.1 K. It is expected, however, that the absolute accuracy of ground-based systems will increase as the models and instruments improve, possibly attaining an absolute accuracy of 0.2 mm of precipitable water vapor (PWV). Since each millimeter of PWV produces approximately 1.4 K of signal at 23.8 GHz, RFI must be less than 0.03 K, assuming a maximum tolerable interference of 10 percent of the sensitivity of the instrument. Higher RFI levels of up to 1 K can be tolerated for observations of integrated liquid water in clouds and rain. Wideband anticollision radars are being licensed and produced in the 22-26 GHz region of the 22-30 GHz wave band, which spans the radio astronomy reserved quiet band at 23.6-24 GHz. These active sources are difficult to discriminate from thermal noise, even with elegant and costly detection methods, and are expected to be an ever-increasing problem to ground-based water vapor (humidity) profiling. Ground-based radiometers receiving around 89 GHz are important in that they are used to discriminate between cloud liquid water and ice. The transitions between the ice-liquid-vapor phases of water drive the thermodynamic energy transport cycles of the atmosphere and are therefore important for monitoring and predicting weather. Knowledge of these three phases is also critical to understanding planetary albedo and planetary radiative transfer, and therefore climate change and global warming, as well. There is a protected primary radio astronomy band at 86-92 GHz, but as mentioned elsewhere in this report, it is difficult to enforce against intrusions by spurious and out-of-band transmissions. Active technologies up to 110 GHz are being developed, in part due to military interest in and funding for active radars around 94 GHz. The growing availability of these high-frequency technologies in this wave band will undoubtedly result in problems from RFI for EESS observations.
OCR for page 83
Spectrum Management for Science in the 21st Century The strong water vapor line centered at 183 GHz is observed for water vapor profiling in dry climates such as high-altitude astronomical observatories and arctic and desert regions. Because of the level of technology required at these high frequencies, little interference in this region is foreseen in the near future. Other Concerns SST Measurements at C-Band and X-Band (5-10 GHz) Of particular future concern is RFI affecting continuous all-weather microwave sea surface temperature measurements in littoral regions that are critical for severe storm forecasting and weather and climate studies (see Figure 2.8). These measurements rely principally on observations at 5-10 GHz, which are generally sensitive to surface temperature changes while being insensitive to clouds. Active services using spectrum adjacent to and within the EESS allocation at 10.6-10.7 GHz can make SST measurements difficult or impossible at this band. UWB devices that radiate in the 2-10 GHz range could be particularly problematic in the future. It is also important to note that 10.6-10.68 GHz is shared with the Fixed Service, and in several areas worldwide, significant interference has been measured and continues to increase. Several EESS satellites have improved on TMI’s 10 GHz measurements of SST by including observations of C-band microwave brightness temperatures, typically near 6.8 GHz. These measurements specifically improve the accuracy of all-weather SST measurements in cold regions and are less prone to being affected by heavy clouds and precipitation. However, uncontaminated measurements of environmental parameters near 6 GHz are becoming more difficult to obtain owing to the high usage of the C-band spectrum and the lack of any EESS allocation adequate to support SST measurements. While the problem of contamination of 5-10 GHz SST measurements exists over all of the global oceans, it is particularly an issue in littoral regions where severe weather is economically important and population density (including ship traffic) is high (see also §2.1). V-Band (50-64 GHz) A number of currently operating space-based instruments use the atmospheric oxygen absorption band (50-64 GHz) to estimate profiles of atmospheric temperature and moisture. These measurements are central to NWP, severe weather forecasting, and climate analysis. International frequency allocations provide a shared “primary” status to EESS in the 57.0-59.3 GHz range, and these frequencies are currently used by several space-based radiometers, including the Advanced Microwave Sounding Unit and the Special Sensor Microwave Imager/Sounder. Both of these sensors operate on multiple satellites to provide full global coverage every few hours (see Table 2.2). AMSU sensors operating in the 50-59 GHz band may
OCR for page 84
Spectrum Management for Science in the 21st Century be the single most important data source enabling useful global weather forecasts up to 7 days in advance. In response to a growing interest in the active use of this part of the spectrum, EESS scientists have begun analyzing the potential for future interference to remote sensing measurements at V-band. The wide bandwidth available and small device sizes that can be manufactured make this potentially fertile ground for commercial interests.55 A recent FCC notice of public rule making (NPRM) requested an allowance for increased power emission levels for sources operating within 57-64 GHz, which includes the ITU-protected 57.0-59.3 GHz portion used for weather-related sensing by many satellites and weather forecasting services.56 Unfortunately, the FCC NPRM included no analysis of the potential impact of these increased power levels on essential EESS passive measurements from AMSU or related instruments, even though it is currently envisioned that wireless systems operating near 60 GHz will become ubiquitous consumer devices for applications such as local DVD broadcasts and personal networking. While atmospheric absorption limits the range of active users’ transmissions, attenuation from the surface to the top of the atmosphere is not complete (as shown in Figure 1.2). A sufficiently high spatial density of low-power emitters on the ground can affect spaceborne microwave observations. Members of the EESS passive community raised this issue in comments filed in response to the FCC’s NPRM, and the FCC’s decision is still forthcoming as of the time of this writing.57 The community is also interacting with IEEE standards organizations to determine the possible impact of such wireless systems on future EESS observations.58 It is clear that RFI degradation of EESS measurements and weather forecasting services appears to be likely if widespread unlicensed transmissions in these bands begin. Consideration should be given to limiting the strength and density of transmitters in this band (see Appendix C) in order to address the concerns of EESS. It may well be that no practical limit exists if such devices are sold as unlicensed 55 B. Bosco, “Emerging Commercial Applications Using the 60 GHz Band,” IEEE Wireless and Microwave Technology conference (WAMICON) 2006, proceedings; B. Razavi, “Gadgets Gab at 60 GHz,” IEEE Spectrum, February 2008. 56 In the Matter of Revision of the Commission’s Rules Regarding Operation in the 57-64 GHz Band, Notice of Proposed Rulemaking, 22 FCC Rcd 10505 (2007). 57 IEEE Geoscience and Remote Sensing Society, “Comments to the proposed revision of the Commission’s Rules Regarding Operation in the 57-64 GHz Band,” available at http://fjallfoss.fcc.gov/prod/ecfs/retrieve.cgi?native_or_pdf=pdf&id_document=6519741794; accessed June 9, 2009. 58 It is noted that while considerable resources are often available to be applied toward legal filings by active users of the spectrum, the nongovernmental scientific community has had little or no financial support for pursuing such legal matters. Virtually all responses from the nongovernmental EESS and RAS communities to NPRMs are the result of either voluntary efforts (in the case of university personnel) or are in direct reaction to threats to the viability of the passive services (in the case of industry personnel).
OCR for page 85
Spectrum Management for Science in the 21st Century and thus potentially used without limit. However, there is no apparent technical reason why the wider band 59.3-64 GHz could not alternatively satisfy essentially all commercial requirements for ubiquitous devices since such bandwidths in a single device far exceed the capacities of most home fiber and cable systems that offer hundreds of television channels and other services. High Frequencies (>100 GHz) In order to improve the understanding of the chemistry associated with stratospheric ozone depletion, it is necessary to observe the global distributions of a wide array of trace gases.59 Measurements are made by observing narrow spectral line emissions. The frequency requirements of those measurements are dictated by molecular quantum transitions of the gases under consideration. Trace gases of particular interest include ozone, chlorine, hydrogen, bromine, and water vapor. NASA’s Microwave Limb Sounder (MLS) and associated follow-on instruments have been designed for trace gas observations.60 The EOS version of MLS operates in five primary spectral bands near 118, 190, 240, 640, and 2500 GHz.61 The specific passbands and minimum detectable signals for MLS are listed in Table 2.3. RFI should be kept at or below one-tenth of the minimum detectable signals levels noted in the table. While no RFI has been reported to date, it is envisioned that the bands above 100 GHz may become commercially useful to the active services in the coming decades. In the near term, the Submillimeter Infrared Radiometer Ice Cloud Experiment (SIRICE) mission is being designed to measure cloud ice water path using passive channels above 100 GHz. SIRICE is currently in pre-Phase A development at NASA. Design studies have identified three channels (including frequencies, bandwidths, and rms measurement errors) for SIRICE required to retrieve IWP with the necessary accuracy and precision. The spectral requirements are summarized in Table 2.4. RFI contamination of SIRICE observations should be at or below one-tenth of the NEΔT levels noted in the table if the scientific integrity of the IWP retrievals is to be maintained. 59 S. Solomon, “Stratospheric Ozone Depletion: A Review of Concepts and History,” Reviews of Geophysics, 37(3): 275–316 (1999). 60 J.W. Waters, W.G. Read, L. Froidevaux, and R.F Jarnot, “The UARS and EOS Microwave Limb Sounder (MLS) Experiments,” Journal of Atmospheric Science, 56: 194-217 (1999). 61 J.W. Waters et al., “The Earth Observing System Microwave Limb Sounder (EOS MLS) on the Aura Satellite,” IEEE Transactions on Geoscience and Remote Sensing, 44(5): 1075-1092 (2006).
OCR for page 86
Spectrum Management for Science in the 21st Century TABLE 2.3 EOS Microwave Limb Sounder Instrument Spectral Coverage and Sensitivity for Measurement of Trace Gases in the Upper Atmosphere Passband (GHz) Minimum Detectable Signal (K) 115.3-122.0 0.1 177.2-206.2 0.03 221.4-240.5 0.1 606.7-657.5 0.1 2481.9-2506.0 0.1 SOURCE: J. Waters, R.E. Cofield, M.J. Filipiak, D.A. Flower, N.J. Livesey, G.L. Manney, H.C. Pumphrey, M.L. Santee, P.H. Siegel, and D.L. Wu, “An Overview of the EOS MLS Experiment,” NASA EOS MLS DRL 601 (part 1), ATBD-MLS-01, JPL D-15745/CL#04-2323, ver. 2.0, January 7, 2005. TABLE 2.4 Submillimeter Infrared Radiometer Ice Cloud Experiment (SIRICE) Instrument Spectral Coverage and Sensitivity Requirements for Measurement of Ice Water Path Center Frequency ± Double Sideband Offset (GHz) Bandwidth (GHz) NEΔT (K) Polarization 183.31±1.5 1.4 0.7 Vertical 183.31±3.5 2.0 0.6 Vertical 183.31±7.0 3.0 0.5 Vertical 325.15±1.5 1.6 1.8 Vertical 325.15±3.5 2.4 1.4 Vertical 325.15±9.5 3.0 1.3 Vertical 448.00±1.4 1.2 2.3 Vertical 448.00±3.0 2.0 1.8 Vertical 448.00±7.2 3.0 1.5 Vertical 642.90±6.7 2.8 1.9 Vertical 642.90±6.7 2.8 1.9 Horizontal 874.40±4.5 6.0 1.9 Vertical 2.6 SUMMARY OF THE IMPORTANCE OF AND RISKS TO CONTINUED CONTRIBUTIONS OF THE EARTH EXPLORATION-SATELLITE SERVICE IN THE FUTURE The Earth Exploration-Satellite Service (EESS) provides critical and unique measurements that support (1) day-to-day weather and other environmental operations, (2) climate research, and (3) model development and other scientific advances in Earth observation. EESS measurements are currently impacted by RFI at all key frequencies up to 19 GHz, and likely at 24 GHz and higher frequencies soon. There is also potential for significant future interference to EESS systems operating at 50-60 GHz. This interference occurs whether the band of concern is assigned to the passive services exclusively, shared with other services, or not assigned to EESS but
OCR for page 87
Spectrum Management for Science in the 21st Century has unique physical properties that demand observation when interference is absent. Unless these issues are addressed in a timely manner, the effectiveness and utility of EESS will likely be increasingly compromised, particularly as wireless services and unlicensed devices proliferate. Most problematic are future ubiquitous unlicensed ultrawideband consumer devices that can proliferate without limit. Box 2.3 illustrates a sporadic record of achievement in appropriately allocating spectrum and/or coordinating technology development between EESS and competing active services. A technology advisory body, incorporating members from all relevant services, could help mitigate such failures. Such an entity would link EESS and other relevant active and passive communities in an early identification of issues and opportunities regarding competing spectral needs and shared standards development. Such a holistic body would supplement the more adversarial and segmented bodies that currently provide most such advice. BOX 2.3 Illustrative Examples of Successes and Failures in Frequency Coordination That Affect the Earth Exploration-Satellite Service (EESS) Successes European and Japanese transition to 77 GHz band for automobile radar, avoiding 23-24 GHz. The development of airborne sub-band-based radio frequency interference (RFI) mitigation methods that delete single strong interference signals, although not weak or diffuse interference. The International Telecommunication Union trade-off of allocations to obtain stronger protection at more important bands at 50-57 GHz. The migration of new instrument specifications toward protected bands (Advanced Technology Microwave Sounder, Special Sensor Microwave/Imager, Special Sensor Microwave/Imager Sounder, Conical Microwave Imager Sounder, and Microwave Imager/Sounder). Failures The lack of engagement between the auto radar community, Earth Exploration-Satellite Service (EESS), and regulators during the technology’s early development. The lack of accepted remedies when unlicensed devices producing limited EESS interference multiply in numbers so as to collectively damage EESS and other services. The lack of global exclusive EESS allocations at 18.7 and 10.65 GHz; critical bands experiencing RFI. No allocation of a protected band at C-band. The difficulty in effectively employing lower-frequency bands (e.g., 1400-1427 MHz) owing to RFI; apparent inadequate protection for EESS operation in the exclusively passive 1400-1427 MHz band.