Better Weather Forecasting at a Lower Cost


Climate change is increasing the need for accurate atmospheric and storm forecasting. While large satellites, daily use of radiosondes, special flights into storms and real-time linkage of ground-based weather stations have greatly improved weather and climate modeling, they still lack the required precision in forecasting, as we have seen with so many major storms and weather events, from powerful hurricanes to unexpectedly heavy rainfall and flooding. Yet there is a new methodology available, at much lower cost, that can dramatically strengthen climate modeling and weather forecasting. It is called GPS radio occultation (GPS-RO), and with it comes a new acquisition strategy based on procurement of satellite information as a service.

This new technology is based on the principles of radio propagation through the atmosphere. Scientists have long known that radio signals at a particular frequency are distorted in a predictable manner, and by measuring the distortion, specific atmospheric parameters — such as air density, pressure, temperature and moisture — may be determined. If thousands of such exact measurements could be collected simultaneously or near-simultaneously, over large areas of the Earth, and at varying altitudes, from the surface high into space, far more accurate weather forecast models could be constructed.

Fortunately, there is a way to collect such measurements, using the signals emitted by GPS radios. All that is required is a constellation of small, inexpensive, GPS-receiving satellites in orbit to sample the GPS satellite radio signals as they are distorted by the atmosphere. Properly deployed and programmed, these small GPS receivers in low Earth orbit would continuously track the signals of the numerous GPS (and soon Galileo and Glonass) satellites as they rise and set through the atmosphere, yielding thousands of data points of atmospheric variables at varying locations and altitudes, including deep within storms and over areas of the Earth difficult to sample but critical in weather formation.

GPS-RO offers temperature accuracy and vertical resolution 20 times better than any other spaceborne sensor operating or planned. As to cost, the instrument is not very different from the GPS receiver in an automobile, though of a scientific grade. A typical space-qualified RO instrument costs about $500,000, while the average atmospheric sensor on the National Polar-orbiting Operational Environmental Satellite System (NPOESS) is over $300 million. An RO instrument can be flown on a microsatellite costing less than $4 million to build. Eighteen can be launched together on a $10 million Falcon 1e. The total cost for an 18-satellite constellation is less than $100 million.

This GPS-RO technology has been successfully demonstrated in a small research constellation called the Constellation Observing System for Meteorology, Ionosphere and Climate (COSMIC). In 2006, scientists at the National Center for Atmospheric Research used a handful of RO profiles to achieve an unprecedented forecast of the development of a weak tropical depression into a full-blown hurricane (Ernesto) four days in advance. Scientists at the U.S. National Oceanic and Atmospheric Administration (NOAA) have shown impacts from RO on forecasts in the Southern Hemisphere surpassing those of any other space instrument, though COSMIC provides only 5 percent of the data volume of proposed operational systems. Weather services on three continents are using the experimental COSMIC RO data operationally, and eagerly looking forward to more. But, unfortunately, this research system will die shortly.

Yes, this is disruptive technology. It means taking advantage of the revolution in microelectronics to launch instruments in space that are small, simple and relatively inexpensive, rather than large, complex and costly. It means using small rockets to reach orbit rather than behemoth boosters. And all this makes it possible for the government to procure weather data from commercially sound private companies. Cost studies have been developed that indicate a private company can deploy an equivalent system for less than a third of the government-procured system. Moreover, the system will be in operation at least three years sooner. The cost of that system can be spread over a dozen or more international government subscribers. The cost to any one subscriber will be only a fraction of the cost of a traditionally acquired government system. The crushing burden to U.S. taxpayers could be relieved and the recognized international need for sustained, low-cost information  could be met to support the U.S. commitment to the Global Earth Observation Systems of Systems (GEOSS.)

This is an innovative alternative acquisition method in line with the policies of the administration of U.S. President Barack Obama to reduce costs and improve efficiency. It answers NOAA’s recognized need for new ways of doing business and responds to a call the agency formally made to industry two years ago.

Clearly, we cannot continue on the path that brought us NPOESS. As a consequence of the response to this initiative, private investors in the United States have offered to fund development of commercial RO systems. They are prepared to deploy the first operational constellation two years from now. A private operation will offer data on a trial basis, at no charge, for customers to evaluate. If they like it, they can subscribe with payment on delivery. Upfront financial risk to the government is eliminated. With a traditional government-funded procurement, most of the system cost is incurred building and launching the satellites, before any data begin to flow. As we know, long delays may ensue, further driving up costs. And then the launch might fail, handing the taxpayer a total loss. With the commercial approach, that risk is absorbed in proven ways.

Moreover, in a traditional mission program, the question of successor systems must inevitably arise. When a mission approaches its end of life — or rather, long before — the government must consider if and how to fund a follow-on. The painful and costly process begins again. In the commercial approach, continuing coverage is factored into the subscription price. Keep paying the subscription fee and the data will continue to flow, in perpetuity. In maintaining and replenishing their constellations, system proprietors will introduce technical advances as they become available. As we see with every variety of consumer device, suppliers will compete on both cost and performance, and the customer — in this case, the U.S. taxpayer — will be the winner.

There is a tradition of technological revolutions first propelled by government needs and investment and then ultimately establishing a commercial foothold. In space, we saw it first with telecommunications, then with remote Earth imaging. The NextView model for commercial Earth imaging, which was pioneered by the U.S. National Geospatial-Intelligence Agency and spawned the successes of DigitalGlobe and GeoEye, has set a standard for the rest of the government to follow. The savings to the government and the stimulus to American technology and private enterprise are beyond dispute. This is surely the way of the future.

The value and impact of GPS-RO for weather and climate applications (and space weather too, though that’s another story) are unquestionable. The price-performance combination is unbeatable. It is time to support the future of GPS-RO as a sustained, operational asset in the U.S. environmental observing system, and procure the resulting information as a service. Buy the data, don’t buy the system.


Retired U.S. Army Gen. Wesley K. Clark, a former NATO supreme allied commander and U.S. presidential candidate, is chairman of the investment bank Rodman & Renshaw LLC and CEO of Wesley K. Clark & Associates. Retired U.S. Navy Vice Adm. Conrad C. Lautenbacher Jr., a former NOAA administrator, is vice president of polar programs for CSC Corp. Both men serve on the board of advisers for GeoOptics LLC, which plans to collect and sell GPS -RO data.