Weather forecasting has made steady progress during the past several decades, yet the financial costs of extreme weather are staggering and getting worse. Part of the problem is that forecast improvements for the most impactful types of weather, including hurricanes, tornadoes, flooding and winter storms, have been slower to materialize, often resulting in fluctuating forecasts and large uncertainties even at short lead times.
To reverse this troubling trend, traditional approaches to advancing the space-based and in situ observations, models, and decision tools that drive weather forecasts and warnings must be re-imagined.
Last year the world suffered $268 billion in economic losses from weather disasters, according to insurance broker Aon. In the U.S. alone, the National Oceanic and Atmospheric Administration (NOAA) estimates severe weather and climate events such as hurricanes, tornadoes and wildfires cost $98.9 billion in 2020, and a total of $243.3 billion from 2018 to 2020. Already in 2021, a Texas economic research firm projects the February winter storm that collapsed the state’s electric grid could cost more than $200 billion.
Behind these astronomical costs are weather forecasts that have gradually improved over time through an international collaboration of the public, private and academic sectors that has built a global forecasting infrastructure, made up primarily of systems owned and operated by governments. But many forecasts still lack the accuracy, detail, lead time and context required to enable proactive decisions.
For example, Atlantic Basin hurricane track errors have decreased from 250 miles (402 kilometers) three days before landfall 20 years ago to 100 miles today, according to NOAA’s National Hurricane Center. But such errors often still create unacceptable uncertainty for emergency managers to properly plan and evacuate. Meanwhile, hurricane intensity forecasts have shown barely any improvement in 30 years, with some storms unexpectedly rapidly intensifying just before landfall and catching communities off-guard.
Tornado warning times increased from three minutes 40 years ago to 14 minutes in 2010, but have actually decreased by about five minutes in recent years. And while general weather forecasts have improved by about one day per decade (for instance, today’s five-day forecast is as good as a four-day forecast was 10 years ago), that already slow rate seems to be getting slower.
Meanwhile, large gaps in observational data outside the United States and Europe have resulted in uneven access to reliable forecasts, leaving billions around the world blind to weather. Such data gaps also diminish the accuracy of forecasts in the United States, because local forecasts depend on global data. For example, to accurately predict the formation and path of a hurricane that ultimately makes landfall in Florida, we need adequate observations off the coast of Africa where the storm originates, and across the Atlantic Ocean as it develops into a tropical system.
There is no one magic bullet to improving forecasts and reining in the costs of extreme weather and climate.
Governments should continue to lead multi-sector efforts to advance forecasting, which in some cases includes the continued development of government-owned and operated systems. But to truly accelerate both the science and operational advances needed, governments should fully embrace the maturing and expanding capabilities of the commercial sector across the entire weather value chain.
Thanks to innovations in sensor technologies, miniaturization and new business models, a number of private companies are now deploying networks of instruments in space, on the ground, and across the ocean to fill the large data gaps that have prevented more meaningful and faster forecast improvements. Even if each of these data sources would cost governments tens of millions of dollars per year, that is still pennies on the dollar compared to building, owning, and operating their own systems.
A handful of pilot programs at NOAA, NASA, and the Department of Defense during the past several years have validated the viability of commercial weather satellites to support operations and research. But the potential of private-sector data to accelerate forecast improvement, at a time when the costs of extreme weather are rising dramatically, warrants a bigger role for the commercial sector. It is time to move beyond pilot programs to incorporating commercial data into programs of record, especially as NOAA, NASA, and the Department of Defense plan their future constellations to replace aging assets.
To truly tap into the powerful innovation of the private sector, though, we must think beyond a few targeted, yet piecemeal commercial data buys.
A recent report on U.S. space-based environmental monitoring (SBEM) by The Aerospace Corporation highlights the potential contribution of industry: “The alignment of future SBEM planning activities across the U.S. government combined with the emergence of the commercial sector presents a rare opportunity to conduct a national dialogue to explore a whole-of-nation approach to address strategic SBEM challenges.”
Industry is well positioned to play a more significant role in such a “whole-of-nation approach,” not only by providing targeted observations to augment government data, but also by unleashing powerful innovation on grand forecast challenges. For example, our conversations with numerous customers impacted by weather revealed global, near-real-time precipitation data as a major gap. So we developed a miniaturized precipitation radar and plan to launch a constellation of them to improve operational weather forecasts around the world.
How can governments more effectively take advantage of emerging commercial weather capabilities?
The answer is to follow the lead of other industries that were once exclusively the domain of governments, but have been revolutionized by greater participation of the private sector, such as satellite imagery, satellite communications and space launch.
When it comes to weather, governments can and should expect much more of the private sector. Instead of only asking industry to provide specific data or sensors, governments should issue and fund more open-ended challenges to improve forecasts of the most extreme and costly weather phenomena. Let industry innovate the path to a solution—which may involve not only new and improved observations, but also artificial intelligence, specialized models and smart software platforms—rather than constraining it to predetermined data or sensor types.
The Environmental Defense Fund estimates that under a warming climate the costs of extreme weather will increase by more than $8 billion every year, reaching $12 trillion by 2050. If there ever was a challenge that required all hands on deck, this is it. Industry is poised to tackle this challenge with creative ideas, innovative solutions and a sense of urgency — governments just need to ask.
Rei Goffer is the co-founder and chief strategy officer for Tomorrow.io, a weather intelligence and climate security company.