Mitigating Gaps in Weather Satellite Data

Overview
Since the 1960s, the United States has used both polar-orbiting and geostationary satellites to observe the Earth and its land, oceans, atmosphere, and space environments. Polar-orbiting satellites constantly circle the Earth in an almost north-south orbit providing global coverage of environmental conditions that affect the weather and climate. As the Earth rotates beneath it, each polar-orbiting satellite views the entire Earth’s surface twice a day. In contrast, geostationary satellites maintain a fixed position relative to the Earth from a high-level orbit of about 22,300 miles in space. Used in combination with ground, sea, and airborne observing systems, both types of satellites have become an indispensable part of monitoring and forecasting weather and climate. For example, polar-orbiting satellites provide the data that go into numerical weather prediction models, which are a primary tool for forecasting weather days in advance, including forecasting the path and intensity of hurricanes and tropical storms.

Geostationary satellites provide frequently-updated graphical images that are used to identify current weather patterns and provide short-term warnings. For more than 40 years, the United States has operated two separate operational polar-orbiting meteorological satellites systems: the Polar-orbiting Operational Environmental Satellite series, which is managed by National Oceanic and Atmospheric Administration (NOAA) — a component of the Department of Commerce; and the Defense Meteorological Satellite Program (DMSP), which is managed by the Air Force. The government also relies on data from a European satellite program, called the Meteorological Operational (MetOp) satellite series. These satellites are positioned so that they cross the Equator in the early morning, midmorning, and early afternoon in order to obtain regular updates throughout the day.

With the expectation that combining the two separate U.S. polar satellite programs would result in sizable cost savings, a May 1994 Presidential Decision Directive required NOAA and DOD to converge the two programs into a single new satellite acquisition, which became the National Polar-orbiting Operational Environmental Satellite System (NPOESS). However, in the years that followed, NPOESS encountered significant technical challenges in sensor development and experienced program cost growth and schedule delays, in part due to problems in the program’s management structure. After several restructurings and recurring challenges, in February 2010, the Executive Office of the President’s Office of Science and Technology Policy announced that NOAA and DOD would no longer jointly procure NPOESS; instead, each agency would plan and acquire its own satellite system. Specifically, NOAA, with support from the National Aeronautics and Space Administration (NASA), would be responsible for the afternoon orbit, and DOD would be responsible for the early morning orbit. The U.S. partnership with the European satellite agency for data from the midmorning orbit would continue as planned.

Subsequently, NOAA initiated its replacement program, the Joint Polar Satellite System (JPSS). JPSS consists of a demonstration satellite—called the Suomi National Polar-orbiting Partnership (NPP)—launched in October 2011; two satellites, with at least five instruments planned for each, to be launched by March 2017 and December 2022, respectively; two stand-alone satellites to accommodate three additional instruments; and ground systems for the entire program. In June 2012, the GAO reported that NOAA and NASA had made progress in establishing the JPSS program and in launching and operating the demonstration satellite, but noted that program officials expect there to be a gap in satellite observations before the first JPSS satellite is launched.

Specifically, NOAA officials anticipate a gap in the afternoon orbit from 18 to 24 months between the time that NPP reaches the end of its lifespan and when the first JPSS satellite is fully ready for operational use. The GAO identified other scenarios where the gap could last from 17 to 53 months. For example, the gap would be 17 months if NPP lasts 5 years until October 2016 and JPSS is launched as planned in March 2017 and undergoes a 12-month on-orbit checkout before it is fully operational. Alternatively, if NPP lasts only 3 years—which NASA engineers consider possible due to poor workmanship in the fabrication of the instruments—and JPSS launches 1 year later than currently planned, the gap in satellite observations could reach 53 months.

After NPOESS was disbanded, DOD also began planning its own follow-on polar satellite program. However, it halted work in early 2012 because it still has two legacy DMSP satellites in storage that will be launched as needed to maintain observations in the early morning orbit. The agency currently plans to launch its two remaining satellites in 2014 and 2020. Moreover, DOD is working to identify alternatives to meet its future environmental satellite requirements. However, in June 2012, the GAO reported that there is a possibility of satellite data gaps in DOD's early morning orbit. The two remaining DMSP satellites may not work as intended because they were built in the late 1990s and will be quite old by the time they are launched. If the satellites do not perform as expected, a data gap in the early morning orbit could occur as early as 2014.

Satellite data gaps in the morning or afternoon polar orbits would lead to less accurate and timely weather forecasting; as a result, advanced warning of extreme events would be affected. Such extreme events could include hurricanes, storm surges, and floods. For example, the National Weather Service performed case studies to demonstrate how its forecasts would have been affected if there were no polar satellite data in the afternoon orbit, and noted that its forecasts for the "Snowmaggedon" winter storm that hit the Mid-Atlantic coast in February 2010 would have predicted a less intense storm further east, with about half of the precipitation at 3, 4, and 5 days before the event. Specifically, the models would have under-forecasted the amount of snow by at least 10 inches. Similarly, a European weather organization recently reported that NOAA’s forecasts of Superstorm Sandy's track could have been hundreds of miles off without polar-orbiting satellites—rather than identifying the New Jersey landfall within 30 miles 4 days before landfall, the models would have shown the storm remaining at sea.

In June 2012, the GAO reported that while NOAA officials communicated publicly and often about the risk of a polar satellite data gap, the agency had not established plans to mitigate the gap. At the time, NOAA officials stated that the agency would continue to use existing satellites as long as they provide data and that there were no viable alternatives to the JPSS program. However, the GAO report noted that a more comprehensive mitigation plan was essential since it is possible that other governmental, commercial, or foreign satellites could supplement the polar satellite data.

For example, other nations continue to launch polar-orbiting weather satellites to acquire data such as sea surface temperatures, sea surface winds, and water vapor. Also, over the next few years, NASA plans to launch satellites that will collect information on precipitation and soil moisture. Because it could take time to adapt ground systems to receive, process, and disseminate an alternative satellite’s data, the GAO noted that any delays in establishing mitigation plans could leave the agency little time to leverage its alternatives. The GAO recommended that NOAA establish mitigation plans for pending satellite gaps in the afternoon orbit as well as potential gaps in the early morning orbit.

In September 2012, the Under Secretary of Commerce for Oceans and Atmosphere (who is also the NOAA Administrator) reported that NOAA had several actions under way to address polar satellite data gaps, including (1) an investigation on how to maximize the life of the demonstration satellite, (2) an investigation on how to accelerate the development of the second JPSS satellite, and (3) the development of a mitigation plan to address potential data gaps until the first JPSS satellite becomes operational. The Under Secretary also directed NOAA’s Assistant Secretary to, by mid-October 2012, establish a contract to conduct an enterprise-wide examination of contingency options and to develop a written, descriptive, end-to-end plan that considers the entire flow of data from possible alternative sensors through data assimilation and on to forecast model performance.

In October 2012, NOAA issued a mitigation plan for a potential 14 to 18 month gap in the afternoon orbit, between the current polar satellite and the first JPSS satellite. The plan identifies and prioritizes options for obtaining critical observations, including alternative satellite data sources and improvements to data assimilation in models. It also lists technical, programmatic, and management steps needed to implement these options.

However, these plans are only the beginning. The agency must make difficult decisions on which steps it will implement to ensure that its mitigation plans are viable when needed. For example, NOAA must make decisions about (1) whether and how to extend support for legacy satellite systems so that their data might be available if needed, (2) how much time and resources to invest in improving satellite models so that they assimilate data from alternative sources, (3) whether to pursue international agreements for access to additional satellite systems and how best to resolve any security issues with the foreign data, (4) when and how to test the value and integration of alternative data sources, and (5) how these preliminary mitigation plans will be integrated with the agency’s broader end-to-end plans for sustaining weather forecasting capabilities. NOAA must also identify time frames for when these decisions will be made. We have ongoing work assessing NOAA’s efforts to limit and mitigate potential polar satellite data gaps.

Geostationary environmental satellites transmit frequently updated images of the weather currently affecting the United States to every national weather forecast office in the country. These are the satellite images that the public often sees on television news programs. NOAA plans to have its $10.9 billion. Geostationary Operational Environmental Satellite-R (GOES-R) series replace the current fleet of geostationary satellites, which will begin to reach the end of their useful lives in 2015. The GOES-R program has undergone a series of changes since 2006 and now consists of four geostationary satellites and a ground system. However, problems with instrument and ground system development caused a 19-month delay in completing the program’s preliminary design review, which occurred in February 2012. In June 2012, we reported that GOES-R schedules were not fully reliable and that they could contribute to delays in satellite launch dates. Program officials acknowledged that the likelihood of meeting the October 2015 launch date was 48 percent.

While NOAA's policy is to have two operational satellites and one backup satellite in orbit at all times, continued delays in the launch of the first GOES-R satellite could lead to a gap in satellite coverage. This policy proved useful in December 2008 and again in September 2012 when the agency experienced problems with one of its operational satellites, but was able to move its backup satellite into place until the problems were resolved. However, beginning in April 2015, NOAA expects to have only two operational satellites and no backup satellite in orbit until GOES-R is launched and completes an estimated 6-month post-launch test period. As a result, there could be a year or more gap during which time a backup satellite would not be available. If NOAA were to experience a problem with either of its operational satellites before GOES-R is in orbit and operational, it would need to rely on older satellites that are beyond their expected operational lives and may not be fully functional. Any further delays in the launch of the first satellite in the GOES-R program would likely increase the risk of a gap in satellite coverage.

In September 2010, we reported that NOAA had not established adequate continuity plans for its geostationary satellites. Specifically, in the event of a satellite failure, with no backup available, NOAA planned to reduce its operations to a single satellite and if available, rely on a satellite from a foreign nation. However, the agency did not have plans that included processes, procedures, and resources needed to transition to a single or foreign satellite. Without such plans, there would be an increased risk that users would lose access to critical data. We recommended that NOAA develop and document continuity plans for the operation of geostationary satellites that included implementation procedures, resources, staff roles, and timetables needed to transition to a single satellite, foreign satellite, or other solution. In September 2011, NOAA developed an initial continuity plan that generally includes these elements. Specifically, NOAA’s plan identified steps it would take in transitioning to a single or foreign satellite; the amount of time this transition would take; roles of product area leads; and resources such as imaging product schedules, disk imagery frequency, and staff to execute the changes. In December 2012, NOAA issued an updated plan that provides additional contingency scenarios.

However, it is not evident that critical steps have been implemented, including simulating continuity situations and working with the user community to account for differences in various continuity scenarios. These steps are critical for NOAA to move forward in documenting the processes it will take to implement its contingency plans. Once these activities are completed, NOAA should update its contingency plan to provide more details on its contingency scenarios, associated time frames, and any preventative actions it is taking to minimize the possibility of a gap. We have ongoing work assessing NOAA’s actions to ensure that its plans are viable and that continuity procedures are in place and have been tested.