Even limited-government conservatives, like me, would concede that the federal government has a role in weather prediction, at the very least for military operations and national security. In fact, it was the U.S. Army and Navy that launched TIROS-1, America’s first low Earth orbit weather satellite, in 1960. As a Navy pilot who has flown combat missions in Afghanistan and Iraq as well as counterdrug missions in Central and South America, I have relied heavily on weather forecasting provided by the U.S. government. When forecasting is incorrect, the result can be dangerous and even fatal situations.

Beyond military applications, the American economy has come to depend on accurate forecasts, including transportation, utility companies, agriculture, financial services, tourism and more. As population centers have grown, the devastating economic and human impact of severe weather events has also grown. The May 20, 2013, tornado that struck Moore, Oklahoma (my home state) took 24 lives and destroyed $2 billion worth of property.

Unfortunately, the United States ranks just fourth in accurate and timely weather forecasting despite spending much more than the rest of the world combined.

The National Oceanic and Atmospheric Administration is doing some good things to correct this situation. First, NOAA is investing in high-performance computing, which is necessary for the numerical weather prediction models that will enable us to improve weather forecasting. Second, NOAA is exploring options to utilize commercial satellite companies.

The United States can dramatically improve weather forecasting, save taxpayer dollars and reduce risk by empowering the commercial weather and satellite industries. This is why I introduced the Weather Forecasting Improvement Act. I am so grateful for the leadership of Chairman Chris Stewart (R-Utah) and Ranking Member Suzanne Bonamici (D-Ore.) who guided this important legislation through the House Science environment subcommittee. Because of their bipartisan leadership, the bill passed through the Science Committee and the House of Representatives without opposition. If it does not pass the Senate this session, I look forward to quickly reintroducing it next year.

This bill will help weather data follow the classic transition from the government to the private sector. As in other space ventures, the government should be involved in the initial phases of developing unproven technology, but once the risks and rewards are quantified, private businesses can step in and provide better and cheaper solutions. NOAA and Congress can help facilitate this transition, but it will happen either way.

The historic government monopoly of weather satellites and associated data is now at the point of creating unnecessary costs, delays and risks that could dramatically degrade U.S. weather forecasting.

As the Suomi NPP satellite comes to the end of its life and the Joint Polar Satellite System (JPSS)-1 is delayed, NOAA has indicated there will likely be a gap in afternoon weather data from our polar-orbiting sun-synchronous satellites. According to NOAA, if the gap is not mitigated, it will “erode everyday weather forecasts and expose the nation to a 25% chance of missing extreme event forecasts that matter most.” That includes tornadic thunderstorms, floods and winter storms. Additionally, hurricane path forecasts will be degraded, resulting in false-alarm evacuations that could cost the country billions of dollars annually. For people from tornado-prone states like Oklahoma or hurricane-prone states on the East Coast or Gulf of Mexico, this risk must be mitigated.

We should not rely on data from China, Russia or even Europe for our forecasts. Doing so would introduce a very real vulnerability to our security and result in the export of American tax dollars, private capital and jobs.

A JPSS gap mitigation analysis of alternatives was called for by NOAA, the Government Accountability Office, the Department of Commerce Office of the Inspector General, and the $4 billion-per-year U.S. commercial weather industry. The report made many recommendations to mitigate the gap but didn’t address its underlying cause: shrinking congressional budgets and a slow-to-react government monopoly of weather satellites.

The taxpayer is underserved by the government monopoly in three ways: inflexibility, increased risks and higher costs.

First, flexibility is necessary to benefit from rapid technological advancement and shifting priorities. When a government monopoly purchases a single multibillion-dollar satellite at the limit of its budget, it lacks the flexibility to be responsive when the budget shrinks or the costs increase. Programs are canceled or other programs are cannibalized to cover the costs. A recent example is the cancellation of the National Polar-orbiting Operational Environmental Satellite System even after massive taxpayer investment. Taxpayers are tired of investing billions of dollars into programs that never materialize.

Second, the government monopoly on satellites is incentivized to exacerbate risks, not alleviate them. NOAA has reduced costs by packing many sensors onto a single satellite to be launched by a single rocket. This is how the $11 billion JPSS program has been designed. If the already delayed JPSS-1 is completely removed as a result of a launch failure, a power system problem or space debris, the United States would be left with no polar-orbiting satellites until JPSS-2 is ready to launch many years later. Eighty percent of the data that go into our numerical weather prediction models come from the polar-orbiting satellites. If we lose the only satellite in the afternoon sun-synchronous orbit, weather forecasts would severely degrade. Given the human and economic stakes in today’s population centers, the risks are too high for this all-or-nothing approach.

Finally, in a government monopoly all costs are spread across the single customer: the U.S. taxpayer.

To be clear, I am not suggesting that JPSS-1 or JPSS-2 be canceled. The federal government has a role in weather prediction and the JPSS program includes sensors and capabilities that the private sector is not prepared to replace. But going forward, NOAA should adopt the Department of Defense’s principles of disaggregation as a method of risk mitigation. Smaller satellites riding on more rockets result in greater resiliency and less risk. Sensors can also ride on other satellites as hosted payloads to reduce both cost and risk. The DoD has already implemented this model with satellite communications and satellite imaging. NOAA should adopt this model for weather data.

NOAA can accomplish this by purchasing data from a robust and diverse weather data collection industry. In this model, the government would buy data rather than hardware and unleash the creative and innovative power of competition and entrepreneurial capitalism. If the private sector can provide the data, then the government should not interfere. The DoD adopted this model for satellite imagery a decade ago, and the private sector developed applications previously never imagined. Now our economy uses this technology for Google Maps, energy development, real estate development, agriculture and much, much more.

The data purchase model for weather would mitigate the data gap, leverage the nimbleness and innovation of the private sector, spread the risk and costs among private equity, private companies and foreign governments, and relieve the pressure on NOAA’s budget — of which half is spent on government-owned and -operated satellites. With numerous companies launching multiple smaller satellites, we can expand the amount of data available, collect data more frequently, deliver data more quickly, spread the cost and risk, and allow the private sector to innovate rapidly. We can achieve better weather prediction at a lesser cost and risk.

The private sector is ready. PlanetiQ and GeoOptics are raising private capital and could launch numerous GPS radio occultation satellites to mitigate the data gap in about two years. Tempus Global Data and GeoMetWatch are in a similar position planning to launch hyperspectral sensors.

As a cost comparison, JPSS is an $11 billion program that will serve from mid-2017 until 2025 at an annualized cost of over $1.4 billion. PlanetiQ plans to sell GPS radio occultation data for $40 million per year. This is merely an example, but that is only 3 percent of the cost with no money required up front. The government only has to pay for the data it receives.

To be fair, private-sector data will be used differently in weather prediction models. Data assimilation systems and numerical weather models will need to be adjusted to account for the new and additional data from private companies. NOAA can test the effectiveness of the various types of weather data using Observation System Experiments and Observation System Simulation Experiments.

To unleash the commercial weather satellite industry, we need NOAA to send a signal that it will purchase data from the private sector. This could be as simple as a letter of intent. Better yet, following the success of the DoD, NOAA could spend a small amount to be an anchor customer of data for small, emerging companies.

As the commercial sector for weather data grows and better forecasting and cost savings are achieved, NOAA can use its internal resources to advance cutting-edge technologies, conduct experimentation and testing, improve the data assimilation systems and invest in high-powered computing and numerical weather models. In order to maximize its impact, NOAA should unleash the creativity, innovation and nimbleness of the private sector where the technologies have already been proved, and allocate its resources where the private sector is not yet willing to go.

U.S. Rep. Jim Bridenstine (R-Okla.) serves on the House Science and Armed Services committees.