We’re so used to seeing those swirling patterns cross our television screens that we never stop to think: Once upon a time, weather satellite imagery was a new technology, and meteorologists and newscasters alike managed to do their jobs without it.

Today, though, that technology is an indispensable part of how government agencies and aid groups understand and respond to severe weather events. Watching a storm gather, and understanding the factors that contribute to it, tells responders where it will land and how quickly, helping them support vulnerable communities and distribute resources where they think they’ll be most needed.

But they’re not always right – in part because weather modeling isn’t perfect, but also because they’ve had to leave out at least half the story. What about the people in the storm’s path? Where are they going (if they leave at all), and how quickly will they get there? Who gets left behind? Until now, it hasn’t been possible to visualize that information in the way meteorologists can visualize an oncoming hurricane.

Thanks to large-scale population movement data – anonymized and aggregated by Facebook – that’s beginning to change.

Imagine a newscaster guiding an audience through the track of a hurricane, then using population data to show how many are at risk. As the storm comes closer to land, they also show the movement of people fleeing the coast. Hour after hour, they’ll track the people leaving their homes. When do they leave, and how far do they go? How long are they in transit? And how many people remain as the storm gains ground?

Now, government agencies and NGOs alike will have answers to these questions. And disaster response will change significantly, for the better, as a result. Direct Relief has already begun to partner with officials in Harris County – which includes Houston and was severely impacted by Hurricane Harvey in 2017 – to help them use Facebook population mapping to plan future disaster response efforts.

“In the aftermath of Hurricane Harvey, we had to rely on physical damage assessments to get a sense of the damage,” said Francisco Sánchez, Jr., a Harris County deputy emergency management coordinator who participated in the Harvey response. “Mapping population data and matching it with other data networks will help us identify hard-hit communities and their specific needs more quickly.”

Here’s how:

1) In the moment, emergency responders will be able to allocate resources more precisely. Under normal circumstances, responders know where people tend to gather (like shopping centers and schools) and can make an educated guess about where best to position resources – whether they’re distributing face masks, providing medicine and personal care items, or placing a medical mobile unit in a vulnerable community.

But people don’t always move to expected locations or at anticipated speeds. And now, organizations will be able to account for that. If they’re watching the data as people evacuate, they’ll incorporate that knowledge in when they’re deciding where to place resources, and change plans in response to demand.

“Disasters move at a quick pace, so if data partnerships can give us real-time information about what happened to who, when, and where, we can make better decisions about how to respond more effectively,” Sánchez said.

2) As things unfold, communication will be more effective. When people connect to the Internet via apps, they’ll transmit information about their network connectivity – how soon they’re able to come back online, whether they have a fast 4G connection or are struggling with a 2G signal.

With that information, aid organizations will know where to provide emergency bandwidth, so people can receive important messages and communicate with loved ones. They’ll also know when people are completely in the dark and, when necessary, get reliable communication to them in person.

3) Once people evacuate, it will be easier to help the vulnerable individuals left behind. Data about specific communities is already available: average income, prevalence of disease, whether there’s a high percentage of children, if there’s a nursing home or retirement community nearby. And when people don’t own a car or aren’t mobile enough to move quickly, they’re more likely to stay put during an emergency.

When that knowledge is combined with a stream of constantly updating population data, it’s easier to figure out who’s probably left behind and how to assist most effectively, whether that’s providing evacuation support or helping them shelter in place.

4) Over the long term, aid groups can provide better support for displaced communities. When a big storm like Harvey wreaks havoc, people can’t go back to their homes – sometimes for years. And displaced people are particularly vulnerable: They’re more likely to use up their financial resources and more likely to have a medical emergency or end up in the hospital.

When organizations track population movement data, they know where displaced communities have gone, how long they’re staying away, and how far they’re dispersed. That means they can coordinate with nearby organizations to provide more continuous, reliable care while people slowly recover from a disaster.

5) For the future, accumulated data from a lot of individual disasters will teach responders more about how people behave as they’re unfolding. Until recently, no one was recording this behavior on a large scale. Responders went into individual disasters with the weight of collective experience behind them, but without much hard data.

If aid groups know more about how things unfolded during a number of past disasters – how quickly people moved away from the coastlines, how far they traveled, whether the roads were open, and who couldn’t make the journey – they can be pretty confident of how events will unfold in the future.

That’s a never-before-available level of advanced knowledge, and it means that emergency responders can place resources wherever they’ll be needed most urgently, before a disaster even hits. When people arrive at a shelter or facility, they’ll be much more likely to have medicine, food, and clothing – even a reliable Internet connection – waiting there for them.