Predicting floods through mobile telephony
Flood warnings? There's an app for that
Researchers in Tel Aviv have established that variations in microwave transmissions, specifically those used to connect up cell towers, can be used to measure humidity and thus predict flooding.
The research, which is published in Atmospheric Chemistry and Physics (pdf) used historical data to demonstrate that variations in microwave connections can be used to monitor water vapour levels and thus should be able to predict when a flood is likely.
"Our method provides reliable measurement of moisture fields near the flood zone for the first time," study co-author Pinhas Alpert told the American Friends of Tel Aviv University. "Accurate predictions of flooding were difficult before because there haven't been enough reliable measurements of moisture fields in remote locations," he added.
The idea is to be able to use existing infrastructure to measure the risk of flooding: as long as it isn't raining, or foggy, or with low cloud cover, and as long as the wind isn't too strong. The study explains that those things all interfere with the measurements, but on a clear, dry, day, without too much wind, you can add remotely-monitored humidity to the weather forecaster's arsenal, as long as the network operators don't mind sharing the efficiency of their microwave backhaul.
Network operators do collect information about their microwave connections, but it's hard to imagine they'd be happy to share every detail. They might, if the technology could really predict the next Katrina-style flooding as the American Friends propose, but it seems to us that it would have been a little windy, and perhaps raining, in the critical days before Hurricane Katrina flooded New Orleans. ®
Sponsored: Hyper-scale data management