WASHINGTON — The U.S. National Oceanic and Atmospheric Administration (NOAA) and Taiwan have developed initial requirements for a collaborative weather satellite program, and will spend the rest of the year putting together an acquisition strategy with an eye toward launching the satellites starting in 2014, U.S. government officials said.

The satellites will use a relatively new method for obtaining atmospheric data called GPS radio occultation, which has been used in operational weather forecasting since a demonstration constellation was launched in 2006. That system, called the Constellation Observing System for Meteorology, Ionosphere and Climate (COSMIC), also was a joint U.S-Taiwanese program.

In addition to building the follow-on COSMIC-2 constellation, NOAA also might purchase additional GPS radio occultation data from commercial firms if companies have them available, David Ector, NOAA’s COSMIC-2 program manager, said in an April 8 interview.

GPS radio occultation was conceived as a way to obtain atmospheric density, pressure, moisture and temperature data from space. A low-orbiting satellite with a GPS receiver can observe the signal that is transmitted by a particular GPS navigation satellite operating in a much higher orbit. As the GPS satellite approaches the horizon relative to the observing satellite, its signals pass through the atmosphere, which can cause frequency variations depending on environmental conditions. Software is then used to infer atmospheric characteristics based on the frequency changes.

The first COSMIC system consists of six satellites designed to collect more than 2,000 atmospheric soundings each day. The constellation was designed to operate through 2011, but is expected to retain some residual capability until around 2013.

The COSMIC-2 system will feature 12 satellites designed to collect more than 8,000 soundings per day. In addition to operating with the U.S. GPS satellites, COSMIC-2 will be able to use signals from the planned European Galileo satellite navigation constellation and possibly the Russian Glonass constellation, Ector said.

Whereas Taiwan paid about 80 percent of the demonstration system’s cost, COSMIC-2 is notionally an even funding split between the two nations, Ector said. NOAA has requested $3.7 million for COSMIC-2 in 2011, and Taiwan has approved the program but not yet provided funding for it, Ector said.

NOAA spokesman John Leslie was not able to provide an estimated total cost to build and launch the system.

With a set of initial requirements in place, NOAA and Taiwan are putting together an acquisition plan. NASA’s Jet Propulsion Laboratory in Pasadena, Calif., built the receivers for the first constellation, and will develop a more sophisticated receiver for the follow-on constellation, Ector said.

Taiwan will be responsible for procuring the satellite platforms, or buses, which are expected to weigh less than 100 kilograms each, Ector said. Taiwan chose Orbital Sciences Corp. of Dulles, Va., to build the demonstration COSMIC platforms, which are based on the bus Orbital designed for the original Orbcomm data communications system.

The governments are considering a variety of launch options, Ector said. The satellites may be launched on the Minotaur rockets built by Orbital, the Falcon 1e rocket built by Space Exploration Technologies Corp. of Hawthorne, Calif., or as secondary payloads on an Evolved Expendable Launch Vehicle built by United Launch Alliance of Denver.

The number of COSMIC-2 ground stations to be built and their locations has not been determined. With only a few ground stations around the world for the demonstration system, the data take as long as three hours to be received, processed and disseminated. The latency requirement will be far shorter for COSMIC-2, and that can be achieved with either more ground stations or satellite-to-satellite data links, Ector said.

The University Corporation for Atmospheric Research, a nonprofit academic organization in Boulder, Colo., developed the software now used to exploit GPS radio occultation data. Exploitation of COSMIC-2 data will still rely on this software, but it has not been determined whether the organization will still be responsible for processing of the data.

“We would hope to be the lead climate data processing center because we’ve had so much experience in that, and we are certainly one of the top three or four processing capabilities in the world for radio occultation data,” Rick Anthes, president of the University Corporation for Atmospheric Research, said in an April 6 interview.

At least one commercial firm has announced plans to build and launch its own constellation of GPS radio occultation satellites. GeoOptics of Pasadena, Calif., plans to launch six or 12 satellites around June 2012. NOAA has expressed a strong interest in buying data from the satellites, as have a host of other governments around the world, GeoOptics President Tom Yunck said in an April 9 interview.

Broad Reach Engineering of Golden, Colo., has been developing the GPS receiver for GeoOptics’ planned satellites, and the firm was recently put under contract to begin work on the satellite platforms, Yunck said. GeoOptics soon plans to sign a contract to launch the satellites on a Falcon 1e rocket, he said.

GeoOptics and Iridium Communications of Bethesda, Md., completed studies for NOAA last year on the cost and feasibility of providing commercial GPS radio occultation data. GeoOptics has said it will launch its constellation with or without a firm commitment from the U.S. government to buy the data, but would prefer to have some up-front investment.