Satellite constellations can be optimized to better predict flooding around the world, researchers from the University, Cornell University and The Aerospace Corporation said in a study published on Feb. 11.
The study examined the current global constellation of satellites used to predict floods to determine the risk posed if any of these satellites were to fail and explored how one could optimize the constellation to more efficiently predict floods, said Eric Wood, co-author of the study and civil and environmental engineering professor at the University.
“We’re interested in combining our global modeling of the hydrology of the Earth’s surface,” Wood said. “[The Cornell researchers] are very interested in optimizing how you would put up constellations of satellites, and Aerospace — which is the third party — they simulate orbits.”
The National Aeronautics and Space Administrationis currently using 10 such satellites, Wood said, which provides global coverage approximately every three hours.
“Some of them have been up there way beyond their lifetime,” Wood said. “The issue is: What is the risk to monitoring precipitation, especially over areas of the world where there aren’t a lot of gauges on the ground?”
Africa and South America in particular have fewer ground gauges, putting these areas at greater risk of unpredicted floods, Woods said.
Areas with strong and short duration thunderstorms are at the highest risk of unpredicted flooding activities, explained Nathaniel Chaney GS, another co-author, adding that with better coordination from different countries and agencies, this problem could largely be addressed.
Examples of past floods that occurred without much warning in the past decade include recent floods in South America, flooding around 2012 or 2013 in West Africa and several bad floods in Mozambique a few years back, Wood said.
“There’s probably a major flood once a week someplace in the world,” he explained. “If they’re unprepared because they don’t know it’s coming, the loss of life is significant.”
The existing system of satellites is still insufficient, Chaney said.
“The main idea that we get from that paper is even with all the satellites we have right now, we’re still far away from being able to capture the sub-three-hour timescale rainfall events,” he said.
The research team used the Blue Waters supercomputer in Illinois, Chaney noted.
“We are using 2.5 million hours of computation to answer this problem in terms of hydrologic modeling,” he said. “What we did couldn’t have been done five or six years ago.”
The optimization solution that the group proposed was better coordination of satellites within a constellation.
“The major finding is that putting up constellations of satellites one by one and not looking at the total problem results in performance that is less than desirable and really results in a overspending or underperformance given the amount of money they spend,” Wood said.
Even if results like these might be discussed at meetings, NASA is not currently planning on putting any additional satellitesinto orbit in part due to monetary constraints, Wood said. The cost of satellites can run from a few million dollars to over a billion, he added.
However, this research could help them best prepare for the loss of satellites in the future and fill in the holes in an efficient way, Wood said.
The same techniques employed in this study can also now be employed to look at similar issues on the Earth’s surface, such as evaporation, weather information and humidity in the atmosphere, he added.
Hydrology is an unusual subject of a satellite model, notedJon Herman, a co-author of the study and a Cornell graduate student in civil and environmental engineering.
“Usually the world of hydrology and satellite constellation modeling don’t really talk to each other,” Herman said. “Satellite missions might not be planned with hydrology in mind, and in hydrologic modeling, it is just sort of taken for granted that the rainfall data is there, so by combining them I think we’re on to something.”