4
Ensuring Continuity

The meteorological requirements for polar and geostationary satellite data were discussed in Chapter 3. Based on these requirements, the committee assessed the number of each type of satellite that is required to be operating in orbit. This chapter examines the launch and in-orbit performance of POES and GOES satellites and uses this information to predict future performance. These predictions are based on Monte Carlo studies, with appropriate assumptions about launch successes, availability of satellites on the ground, delay in the launch and check-out of a replacement for a satellite that fails in orbit, and the expected life of consumable supplies on the satellites.1 These studies are compared with the schedule of POES and GOES spacecraft in orbit, under contract, and planned to determine whether the NOAA satellite program is on a sound footing regarding the continuity of observations for the NWS. Finally, POES and GOES ground systems are examined for possible vulnerability as well as the possibility of other spacecraft providing some backup to POES and GOES in an emergency.

Past Performance of Geostationary Operational Environmental Satellites and Polar-Orbiting Operational Environmental Satellites

The operational experience for GOES over about two decades is summarized in Table 4-1. Useful life of the U.S.-launched geostationary satellites ranges from

1  

 The Monte Carlo model operates using random numbers to select the probability of events occurring in a large number of simulated scenarios. The resulting ensemble of scenarios is presumed to represent the probability distribution (PD) of the true events. It is analyzed to estimate the PD of key results, such as number of operating satellites versus time. Sensitivity to assumptions about various parameters (launch success, satellite mean-time-to-failure, etc.) can be determined by varying the appropriate parameter, rerunning as a new case, and comparing the results.



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 27
--> 4 Ensuring Continuity The meteorological requirements for polar and geostationary satellite data were discussed in Chapter 3. Based on these requirements, the committee assessed the number of each type of satellite that is required to be operating in orbit. This chapter examines the launch and in-orbit performance of POES and GOES satellites and uses this information to predict future performance. These predictions are based on Monte Carlo studies, with appropriate assumptions about launch successes, availability of satellites on the ground, delay in the launch and check-out of a replacement for a satellite that fails in orbit, and the expected life of consumable supplies on the satellites.1 These studies are compared with the schedule of POES and GOES spacecraft in orbit, under contract, and planned to determine whether the NOAA satellite program is on a sound footing regarding the continuity of observations for the NWS. Finally, POES and GOES ground systems are examined for possible vulnerability as well as the possibility of other spacecraft providing some backup to POES and GOES in an emergency. Past Performance of Geostationary Operational Environmental Satellites and Polar-Orbiting Operational Environmental Satellites The operational experience for GOES over about two decades is summarized in Table 4-1. Useful life of the U.S.-launched geostationary satellites ranges from 1    The Monte Carlo model operates using random numbers to select the probability of events occurring in a large number of simulated scenarios. The resulting ensemble of scenarios is presumed to represent the probability distribution (PD) of the true events. It is analyzed to estimate the PD of key results, such as number of operating satellites versus time. Sensitivity to assumptions about various parameters (launch success, satellite mean-time-to-failure, etc.) can be determined by varying the appropriate parameter, rerunning as a new case, and comparing the results.

OCR for page 27
--> TABLE 4-1 SMS/GOES Operational Experience West Operational Spacecraft (135 degrees west longitude) Central Operational Spacecraft (98 to 112 degrees west longitude) East Operational Spacecraft (75 degrees west longitude) SMS-2 03/10/75 to 04/04/78   SMS-1 11/15/74 to 01/08/76 GOES-1 04/04/78 to 07/13/78   GOES-1 01/08/76 to 08/15/77 GOES-3 07/13/78 to 03/05/81 (fail)   GOES-2 08/15/77 to 01/26/79 (fail) GOES-4 03/05/81 to 11/26/82 (fail)   SMS-1 01/26/79 to 04/19/79 GOES-1 11/29/82 to 06/01/83; no IRa   SMS-2 04/19/79 to 08/05/81 (fail) GOES-6 06/01/83 to 07/30/84 (Break in west service) GOES-6 07/30/84 to 03/25/87 GOES-5 08/05/81 to 07/30/84 (fail) GOES-1 08/27/84 to 02/03/85; no IRa (fail) (Break in west service)   (break in east service) GOES-6 03/25/87 to 01/21/89 to 01/21/89 (fail) (Break in west service) GOES-7 01/21/89 to 01/20/95 GOES-7 03/25/87 to 01/21/89 (Break in east service) GOES-7 01/20/95 to 01/11/96   METEOSAT-3 02/28/93 to 01/23/95b GOES-9 01/11/96 to present   GOES-8 01/20/95 to present a Infrared (IR) channel failed 3/24/79, eliminating all night-time images from GOES-1. b A spare METEOSAT was loaned by the European consortium, EUMETSAT. After installation of the necessary ground equipment and communication links of the NESDIS CDA station in Wallops Island, Virginia, METEOSAT-3 was moved to 75 degrees west longitude where it operated until 1/23/95, at which time it was moved to 70 degrees west longitude. Notes: One SMS/GOES spacecraft, GOES-G, suffered a launch failure in June 1986. When a spacecraft was moved from one location to another, the departure date shown above is also assumed to be the subsequent arrival date at the new location. The actual movement rate is about .5 to 1 degree/day. Low-level satellite winds (picture pair or triplet) have been operational with interruptions as indicated above. They are produced from IR data, with exceptions as noted (that is, when only visible data were available). Regarding the data collection system (DCS), the Satellite Services Division of NESDIS does not believe there has been a major failure since the new DCS automated processing system (DAPS) was established in 1989. Data sets were periodically lost because of problems with ground equipment, especially hard disk failures, starting in August 1993. New disks installed in September 1994 resolved the problem.

OCR for page 27
--> 1.55 years to 9.30 years (Hernandez, 1994).2 The breaks in service for a west operational GOES and an east operational GOES are also listed in Table 4-1 and discussed below; the failures are described in the accompanying notes. From July 1984 to March 1987, and again from January 1989 to February 1993, just one geostationary spacecraft that provided forecast support was in operation.3 Three factors were responsible: The lifetimes of four GOES spacecraft (GOES-2 through GOES-5) were much shorter than expected because of a component problem. The launch vehicle for GOES-G failed. The development of the new GOES-Next series of geostationary satellites experienced delays. Generic problems, such as the problems that arose on GOES-2 through GOES-5; launch failures (e.g., GOES-G); and delays in procuring and introducing new technology (GOES-Next) justify the conservative planning of launch schedules. Fortunately, making backup arrangements with a consortium of European nations was possible. One of their geostationary weather satellites, METEOSAT-3, was borrowed from 1993 until the first satellite in the GOES-Next series could be completed, launched, and used operationally in 1995. The METEOSAT-3 was used to provide coverage of the eastern United States and the Atlantic Ocean. Table 4-2 is a summary of the operational experience for POES, including outages (and degradations) of sensors on the morning POES and afternoon POES. The problems experienced by the POES are described in the accompanying notes. The useful life of the POESs launched since 1978 has ranged from 12 days to more than eight years. The more recent satellites have lasted longer than expected, with the exception of NOAA-13. Through December 1995, the mean life, defined as the total operating time divided by the number of failures, has been five years (Broadhurst, 1996). Because of the limited number of failures in this series of spacecraft, Broadhurst also calculated an estimated average mission life of 4.3 years with a 70 percent confidence factor. These computations do not include the ITOS series of six satellites, which bear little resemblance to the current satellite or instrument designs. Predictions of Future Performance. NASA commissioned engineering studies to evaluate the probabilities of GOES and POES continuity as the POES program makes the transition to the 2    Useful life is the length of time a satellite actually performed its mission. 3    METEOSAT-3 started operation as the east satellite in February 1993.

OCR for page 27
--> TABLE 4-2 POES Operational Experiencea Morning Operational Spacecraft—about 0730 local sun time     Atmospheric Sounding Spacecraft Imaging: AVHRR HIRS SSU MSU NOAA-6b 7/79 to 8/81, 9/81 to 2/82 no AM images 8 to 9/81 and 2/82 to 6/83 7/79 to 6/83 7/79 to 6/83 7/79 to 6/83 NOAA-8c 6/83 to 6/84 and 7 to 10/85 6/83 to 6/84 6/83 to 6/84 6/83 to 6/84 NOAA-6d 10/85 to 1/87 no AM i images 6/84 to 7/85 failed 6/84 to 1/87 6/84 to 11/86 NOAA-10 11/86 to 9/91 11/86 to 9/91 none 11/86 to 9/91 NOAA-12e 9/91 to present 9/91 to present 9/91 to present 9/91 to present Afternoon Operational Spacecraft—about 1330–1430 local sun time     Atmospheric Sounding Spacecraft Imaging: AVHRR HIRS SSU MSU TIROS-Nf 1/79 to 1/20/80 1/79 to 1/20/80 1/79 to 1/20/80 1/79 to 1/20/80 TIROS-N 1/30/80 to 11/30/80 1/80 to 2/81 1/80 to 2/81 1/80 to 2/81 NOAA-B launch failure, 1980; no PM data 2/81 to 8/81       NOAA-7 8/81 to 2/85 8/81 to 2/85 8/81 to 2/85 8/81 to 2/85 NOAA-9g 2/85 to 11/88 2/85 to 11/88 2/85 to 11/88 2/85 to 11/88 NOAA-11h 11/88 to 9/94 11/88 to 1/95 11/88 to 1/95 11/88 to 1/95 NOAA-13i spacecraft failed after 12 days       NOAA-14 1/95 to present 1/95 to present 1/95 to present 1/95 to present a Only the imaging and sounding instruments, of prime importance to the NWS modernization, are covered in this table. Dates shown are for the provision of operational data, not necessarily component life. b NOAA-6 had no AVHRR 8/81–9/81 due to excessive jitter. Also MSU may have degraded beginning 11/79. c NOAA-8 attitude control system failed 6/84 and recovered later, affecting all sensors. d After NOAA-8 failed, NOAA-6 again provided coverage on a limited basis form June 1984 to January 1987. e NOAA-12 AVHRR degraded periodically by line jitter (11/91, 1/92, 12/92, 3/94). f NASA prototype spacecraft. g NOAA-9 HIRS long-wave channels became noisy, significantly degrading soundings beginning 12/84. MSU lost one channel (of three) 2/87; a second failed 5/87; these failures significantly degraded soundings. h NOAA-11 AVHRR and HIRS degraded from 10/89 due to orbit drift that caused sunlight to periodically enter instrument apertures. No aerosol optical thickness or global vegetation index products 9/14/94 to 3/20/95. The NOAA-11 AVHRR failure on 9/14/94 resulted in loss of aerosol products until NOAA-14 was operational on 3/20/95. Sea surface temperatures were switched to NOAA-12. During this period, radiation budget and coast watch products were produced only from NOAA-12, instead of from both morning and afternoon spacecraft. i The electrical problem that caused a failure of NOAA-13 has been identified. Design modifications were made in subsequent spacecraft to preclude a recurrence.

OCR for page 27
--> NPOESS era. Based on stated assumptions, these studies included predictions (based on Monte Carlo techniques) of satellite in-orbit availability for the next decade or so. The results of these studies are not exact, however, because of the uncertainty and dispersion in the reliability data and because of the assumptions used. However, the committee found that the results do provide order of magnitude estimates and indicate their sensitivity to variations of parameters and assumptions. Four important assumptions were used in the GOES study (Hernandez, 1994): Satellite reliability mean-time-to-failure was based on earlier experience.4. Engineering studies of propellant consumption (provided by NASA) were used. Failed satellites were assumed to be replaced as needed, with a delay to account for preparation of the replacement satellite for launch, availability of a launch slot, and check-out of the new satellite in orbit. Launch vehicle reliability was assumed to be 90 percent, the industry average at that time. This operating scenario projected ten GOES launches from 1994 through 2009, with six failures during that time, ending with two operational satellites and two backup satellites in orbit (see related discussion on pp. 33–34). Hernandez (1994) includes a figure (presented here, with scale adaptations, as Figure 4-1) showing that, under the assumptions listed above, the probability of two GOES spacecraft operating in orbit from now until 2012 is approximately 95 percent. Allowances have been made for some of the more pessimistic startup conditions used by Hernandez. For example, the committee notes that industry launch reliability as of December 1995 is better than the 90 percent value assumed in the Hernandez study. The Monte Carlo model of the original POES study (Hernandez, 1995) has been updated and is now called the NASA “Mission Planning Model” (Mazur, 1996). The new set of assumptions reflect an additional year of experience with POES: Satellite life is based on TIROS/POES experience through December 1995. The availability of POES satellites is based on the current production schedule. More precise launch vehicle reliability data are used (e.g., Titan II. 94 percent; Delta, 97 percent; Ariane-5,5 90 percent). The response time to replace a failed satellite has been updated. 4    In this regard, it should be noted that GOES-8 and GOES-9, launched in 1994 and 1995, are of a major new design compared to previous GOES. Thus experience with the new design is very limited. 5    Ariane-5 reliability was introduced to account for launches of the METOP satellites.

OCR for page 27
--> Figure 4-1 Availability of GOES service in a two-satellite constellation. Adapted from Hernandez (1994). These factors were used to predict the probability of POES satellites operating in orbit. Figure 4-2, derived from the data in Mazur (1996), shows that the probability at any given time of at least one POES operating in orbit from 1996 to 2002 (when METOP-1 is scheduled to be available) is better than 95 percent. If METOP-1 and METOP-2 are included as a part of the operational scenario, the probability of at least one polar satellite operating remains better than 95 percent to at least 2010. Launch of the first NPOESS is currently planned for about 2008, and this time frame was included in the operational scenario. These models may be used to indicate behavior and approximate levels of availability of the POES and GOES satellites. Sensitivity studies in the referenced reports showed the system to be fairly insensitive to variations in parameters, such as satellite reliability and lifetime (due to expendables). The strongest Figure 4-2 Availability of POES service by at least one satellite in a two-satellite (AM and PM) constellation. Derived from data in Mazur (1996).

OCR for page 27
--> influences were the programmed delivery dates of satellites and launch vehicles and the long delays between failure of a vehicle in orbit and the launch of a replacement. The models are useful as tools to judge the effect of alternate choices in the replenishment launch-decision process discussed in the section on Satellite Launches later in this chapter. Schedule of Future Geostationary Operational Environmental Satellites and Polar-Orbiting Operational Environmental Satellites The planned launch schedule for GOES is shown in Table 4-3. The actual schedule will depend primarily on the need for replacement satellites, of course, but also on spacecraft and launch availability. NOAA has planned, but has not yet procured, a block of four additional satellites, essentially of the same design, to provide coverage through about 2012. This plan assumes three-year lifetimes for GOES-8 and GOES-9, with one launch failure in the series; a four-year lifetime for GOES-K; and five-year lifetimes for the remaining five spacecraft, following TABLE 4-3 GOES Schedule Satellitea Need Dateb Satellite Availability Date Planned Launch Datec GOES-8 Launched Apr 1994     GOES-9 Launched May 1995     GOES-K Apr 1997 Apr 1997 Apr 1997 GOES-M Apr 1998 Aug 1999d Apr 2000 GOES-L Apr 2001 Apr 2001 Apr 2002 (failure) GOES-N Apr 2001e   Apr 2002 GOES-O Apr 2004   Apr 2005 GOES-P Apr 2006   Apr 2007 GOES-Q Apr 2008   Apr 2010 GOES-R to be determined     a This planning schedule assumes one launch failure (GOES-L), a lifetime of only three years for GOES-8 and GOES-9 (the first of a new design), a four-year lifetime for GOES-K, and five-year lifetimes for subsequent satellites of the same design (through GOES-Q). GOES-R is proposed as an new design with competitive procurement (requiring longer lead time). Spacecraft through GOES-M are under contract. The plan assumes that approval to contract for GOES-N et seq. and necessary funding will be received on schedule. b The earliest date a satellite and launch vehicle are required to ensure mission continuity. c Estimated launch date if no unexpected problems occur. d In some cases, cuts in funding can interrupt the manufacturing schedule and delay the availability of satellites and launch vehicles. e GOES-N through GOES-Q are not yet on contract.

OCR for page 27
--> successful launches. The Hernandez (1994) prediction of GOES continuity discussed earlier projected that ten launches in the period 1994 through 2009 would give a probability of approximately 95 percent of having two GOES operating in orbit until 2012. The current schedule (Table 4-3) calls for nine spacecraft to be launched in the same period. The planning launch schedule for POES is shown in Table 4-4. This table, adapted from Winokur (1996), covers the period in which the POES program will undergo the transition to the NPOESS program. The launch of NOAA-K, scheduled for 1997, will provide some sensor improvements: the AVHRR/2 will have an additional channel (for a total of six) and some adjustments to spectral intervals; HIRS-3 will be introduced; and the MSU and SSU will be replaced by an advanced MSU to improve soundings in cloudy regions. Four additional spacecraft, NOAA-L,-M,-N, and-N', are under contract. Launch dates extend to 2007 (see Table 4-4). Negotiations with the European Organization for the Exploitation of Meteorological Satellites (EUMETSAT) are under way. If agreement is reached, Europe will launch and operate the METOP mid-morning satellites in lieu of the morning POES now provided by NOAA. The launch schedule for NOAA-M is based on successful conclusion of the European agreement. In May 1994, President Clinton signed a presidential decision directive to merge the POES system and the DMSP into a single system to “reduce the cost of acquiring and operating polar-orbiting environmental satellite systems, while continuing to satisfy U.S. operational requirements for data from these systems.” In response to this directive, an Integrated Program Office was established within NOAA on October 3, 1994. This office is staffed with DOD, NASA, and NOAA personnel. TABLE 4-4 Polar Satellite Planning Launch Schedule Spacecraft Orbita Launch Yearb NOAA-K AM 1997 NOAA-L AM OR PM 2000 NOAA-M AM OR PM 2001 METOP-1 AMc 2002 NOAA-N PM 2003 METOP-2 AMc 2006 NOAA-N' PM 2007 1st NPOESS PM 2008d a See footnote 2 on page 10. b Planning schedule (calendar years) as of July 1, 1996. c The European METOP satellite are to assume the AM orbit: NOAA and NPOESS satellites will cover the PM orbit. d Subsequently, NPOESS also will replace DMSP in the early morning orbit (about 5:30 a.m. local standard time).

OCR for page 27
--> The present schedule calls for the launch by the United States of the first new, merged-system satellite, NPOESS in 2008, with subsequent launches as required to maintain continuity of coverage for the following decade. Studies are under way to develop a detailed definition of the capabilities of the sounders, imagers, and other sensors that will be flown on these new satellites. Present plans call for three NPOESS spacecraft to operate simultaneously, with two U.S. satellites (at the 0530 and 1330 equator-crossing times) and one European satellite, the METOP, at the 0930 equator-crossing time. The first METOP launch is planned for about 2002. It is anticipated that the NOAA POES spacecraft (through-N and-N') and METOP will provide coverage for NOAA until the new NPOESS is operational (see Table 4-4). The need and timing for “gap-fillers” in case of a delay in NPOESS (including METOP) depend on the availability factors discussed here and on the launch-decision process. It should be noted that the DMSP spacecraft can provide only a limited level of backup to the POES (see p. 36). Finding 3. The polar program, as presently planned, depends on the availability of European METOP satellites in polar orbit from 2002 to 2010 and the availability of NPOESS beginning about 2008. Recommendation 3. NOAA should closely coordinate the POES program with the progress of the NPOESS and METOP satellite programs so that “gap-filler” satellites are not needed. Finding 4. In the longer term, the replenishment strategy depends on the new NPOESS for polar-orbiting satellites and on the procurement of additional GOES satellites. (NOAA plans to procure a new design beginning with GOES-R.) Longer lead times than normal are required when new designs, which are planned for GOES-R, METOP, and NPOESS, are introduced. Recommendation 4. When considering ways to develop new spacecraft and incorporate major new improvements in technology, NOAA should carefully consider the lead times dictated by the required launch schedules and the very long procurement cycle. NOAA should develop schedules for the transition from current designs to new ones, such as NPOESS and GOES-R, that adequately account for the necessary lead times for funding approval, procurement, design and development, fabrication, and verification. Polar-Orbiting Operational Environmental Satellite Ground System Spacecraft programming and commands originate at the Satellite Operations Control Center (SOCC) in Suitland, Maryland. Commands, spacecraft telemetry, and environmental data are relayed between the SOCC and CDA stations by commercial communication satellites.

OCR for page 27
--> The current POES ground system consists of two major subsystems: the polar acquisition and control subsystem (PACS) and the central environmental satellite computer system (CEMSCS). The PACS components are located at two CDA stations near Wallops Island, Virginia, and Fairbanks, Alaska; at the SOCC; and at a western European station near Lannion, France. PACS includes all components necessary to command and control the spacecraft, to monitor the health of the spacecraft using housekeeping telemetry, and to retrieve and transmit environmental data to the CEMSCS. The CEMSCS ingests, preprocesses, and stores the raw satellite data along with other information, such as data related to Earth location and quality control. This ground system for POES appears to be adequate for the foreseeable future. Polar-Orbiting Operational Environmental Satellite Backup Reliable and continuous service from POES is a high priority design criterion. As noted in Chapter 3, at least one operational POES is needed in orbit at all times to meet data requirements for NCEP's global numerical prediction models. NOAA meets this requirement by having two operational POES in orbit at all times and launching another one whenever a primary spacecraft system or sensor fails. In the event of a catastrophic POES failure, significant launch delay, or launch failure, NOAA and DOD have shared processing arrangements to provide a modest level of backup using sounder data from DMSP satellites. Because the current DMSP satellites are designed to meet military requirements, their sounding capabilities and orbit can provide only limited backup, which does not satisfy NOAA's primary requirements. The planned NPOESS, the follow-on to the current POES and DMSP, will have three operating spacecraft in orbit. One of these will be the European METOP. NPOESS plans reflect both U.S. civil (not only NWS) and military requirements, as well as European needs via their contribution of METOP. This NRC study of the high priority NWS needs for continuity finds that at least one operational POES should be in orbit at all times (p. 21). A backup POES should also be in orbit to prevent unacceptable degradation of service when the operational POES fails. If these plans for NPOESS materialize, there should be adequate spacecraft to meet NWS's backup requirements for polar orbiting satellites. However, it is important to maintain the planned schedules of the METOP and the NPOESS. (See findings and recommendations 3 and 4.) Geostationary Operational Environmental Satellite Ground Facilities and Backup Systems The operations ground equipment for GOES consists of components located at the CDA station near Wallops Island, Virginia, and the SOCC at Suitland,

OCR for page 27
--> Maryland. Operational management and planning are performed at the SOCC, where all elements of the system are monitored, evaluated, scheduled, and commanded. Communications links, ground support equipment, and data transmission paths complete the interfaces among the GOES-specific components and other equipment. This network routes broadcast and mission data. The operations ground equipment at the CDA station receives streams of raw images, sounder information, and other data from the GOES spacecraft. The data are formatted in real time, and this data stream is transmitted from the CDA station via its corresponding GOES spacecraft to primary system users. The data stream relayed by the GOES is also received at the CDA station and at the SOCC for use in other internal operations, such as diagnosing instruments and spacecraft and monitoring the quality of processed sensor data. The Wallops CDA station is a potential single point of failure for the GOES system. Although the Wallops CDA station is internally redundant and can sustain many subsystem failures without significant effect on system performance, any phenomenon that could shut down the station, such as a hurricane, flooding, major fire, or explosion, would result in a complete cutoff of data from the GOES satellites. Minimum prudent backup would require an antenna subsystem and receiving, transmitting, data formatting, and processing subsystems to command a GOES satellite, receive telemetry, and acquire and distribute data from the sounding and imaging instruments. These backup systems would need to be located in an area geographically remote from the Wallops facility to ensure the continuity of operations. Finding 5. The Wallops CDA station is a potential single point of failure for the GOES system. Shutdown of the station would result in a complete cutoff of data from the GOES satellites. Recommendation 5. NOAA should implement an adequate backup system to the Wallops CDA station to ensure the uninterrupted operation of GOES satellites and the acquisition of sufficient data to generate basic image and sounding products to meet NWS mission requirements. Geostationary Operational Environmental Satellite Backup From 1993 to 1995, when only one GOES spacecraft was operational, EUMETSAT loaned NOAA a geostationary weather satellite, METEOSAT-3. The satellite was repositioned to 75 degrees west longitude in order to cover the eastern United States and Atlantic Ocean (see Table 4-1). This permitted restoration of U.S. geostationary satellite east-west coverage (imaging only) until the new GOES-8 satellite became operational. Subsequently, EUMETSAT permanently deactivated METEOSAT-3, the necessary ground equipment at NESDIS

OCR for page 27
--> was removed, and communication satellite channels between NESDIS and EUMETSAT were discontinued. A formal agreement, “Backup of Operational Geostationary Meteorological Satellite Systems,” was consummated between NOAA and EUMETSAT in 1993 (NOAA and EUMETSAT, 1993). The agreement describes the backup EUMETSAT/NOAA will provide and the conditions under which backup will be provided in case of the launch or in-orbit failure of a GOES or METEOSAT. The agreement discusses gridded image data and wind vectors derived from the image data. Geostationary satellite sounding data are not included because METEOSAT does not have a sounding capability. Under the agreement, a satellite will be deemed to have failed if the expected products (images and winds) cannot be provided. The agreement also specifies conditions under which a backup METEOSAT or GOES satellite would be moved to benefit the other party. Because of communication range, a METEOSAT would not be moved farther west than 50 degrees west longitude (rather than the 75 degrees used with METEOSAT-3). Under the agreement, if NOAA requests the move of an operable METEOSAT westward to a position that ensures U.S. coverage, it would be moved if an operational GOES fails, if no other operable GOES is in orbit, if a GOES launch is not possible within four months, and if at least two operable METEOSATs are in orbit. If EUMETSAT makes the same request for a GOES to be moved to ensure European coverage, but not farther east than 5 degrees west longitude, it would be moved only if the operational METEOSAT fails, if there is no other operable METEOSAT in orbit, if a METEOSAT launch is not possible within four months, and if at least two operable GOES are in orbit. NESDIS now receives the full suite of processed METEOSAT data (satellite subpoint at zero degrees longitude) covering Europe, Africa, and adjacent oceans. These data are processed by EUMETSAT in Darmstadt, Germany, and rebroadcast through the METEOSAT to users, including NESDIS, via the Wallops CDA station. This data processing and relay system would be used if a METEOSAT were moved to 50 degrees west longitude to take the place of a GOES. (A METEOSAT can only be moved and operated west of 50 degrees west longitude if the communication satellite links are reestablished and the necessary ground equipment is installed at Wallops for command and control of the satellite from the EUMETSAT ground station at Fucino, Italy.) The METEOSAT-6 backup satellite in orbit today has limited imaging and no sounding capability. Thus, under the current agreement and satellite availability, EUMETSAT can now provide only limited backup to GOES, even at 50 degrees west longitude. If the current METEOSATs fail and only two GOES remain in operational condition in orbit, the agreement could force the U.S. system down to only one GOES. The other GOES would be moved to a position to ensure European coverage (a longitude not specified in the agreement) to replace the METEOSAT. Thus, the backup appears to be of limited value for meeting NOAA's requirements.

OCR for page 27
--> Finding 6. Although arrangements have been made for a European geostationary meteorological satellite, METEOSAT, to provide backup to GOES under certain conditions, the coverage by METEOSAT or its successor is very limited compared to the coverage provided by GOES. Recommendation 6. NOAA should ensure that a replacement GOES satellite can be launched and operated as soon as an operational GOES fails to continue the full level of coverage expected of this series of satellites. Satellite Launches As discussed earlier in this chapter and in Chapter 3, the NOAA mission requires at least two operational geostationary satellites in orbit at all times. At least one operational polar-orbiting satellite and an additional backup satellite in orbit are needed at all times to ensure continuous service. The launch-decision process is schedule driven, from call up until a satellite is in orbit and operational. Another satellite must be available to replace one that fails. It now takes nearly two years to call up, launch, and check out a GOES, if one is available on the ground. More than a year of this time is required for scheduling and launching a GOES using current commercial launch services. Thus, it appears that only a standby GOES satellite in orbit can ensure that there is no interruption of coverage. Since the availability of geostationary satellites from EUMETSAT appears to offer backup of limited value, NOAA is examining the merits of storing a GOES in orbit (in a non-operating mode) rather than on the ground. Finding 7. One additional GOES is needed in orbit as a ready spare to protect against a dangerous, protracted loss of full, two-satellite coverage if one operational GOES fails. Even more severe, although far less likely, would be an outage of continual, real-time coverage if both satellites in orbit should fail before a replacement could be made operational. Dependence on commercial launches for GOES can lead to delays of well over a year, even if a spacecraft is available for launch. Recommendation 7. To ensure the continuity of two-GOES coverage, NOAA should store a standby GOES in orbit, rather than on the ground, if this is technically and operationally feasible and cost effective. The time to prepare, launch, and check out a POES is on the order of one year because government launch services are used. Other challenges to the continuity of both POES and GOES service include launch failures, satellite system and sensor failures, delays due to the introduction of new technology into the observing system, budgets, and a ten-year procurement cycle. The determining factors in continuity of service are the time between a

OCR for page 27
--> failure in orbit and an operating replacement and the recovery time after a launch failure. Consideration of these factors has led to establishing a launch-decision process in which a group of senior managers of the NWS and NESDIS regularly review the status of satellites in orbit, satellite production, the availability of launch times, the readiness of the next satellite in the short range, and the replenishment strategy in the long range. They also review satellite availability prediction studies and update them regularly. This review process recently led to a decision to delay the launch of GOES-K from May 1996 to April 1997. The review process also led to a decision to change satellite orientation and other steps to extend in-orbit lifetimes, thereby maintaining continuity of coverage. Finding 8a. The launch-decision process used by NWS and NESDIS is appropriate. Finding 8b. The planned GOES, POES, METOP, and NPOESS procurements are adequate to provide continuity of the NOAA geostationary and polar-orbiting satellites for at least the next 15 years, if they are funded and carried out on the current schedule. Recommendation 8. To ensure continuity, NOAA should fund and procure the planned block of four GOES-Next spacecraft in a timely fashion and should avoid further delays in the METOP and NPOESS programs.