- Rain signalled (Image: Per Foreby/Flickr/Getty)
- dn23136-1_300[1].jpg (16.58 KiB) Viewed 2307 times
Network operators use microwaves to link cellular towers together. When it rains, the signal received at the next tower down the line can be weakened by reflection off raindrops.
A team in the Netherlands has exploited this effect to build a rainfall map for the country. The researchers say their technique could be useful for improving flood forecasts and providing real-time rainfall monitoring in poorer countries where conventional meteorological equipment is rare, but mobile phones are commonplace.
The technique relies on each cellular tower recording the signal strength of the microwave transmissions it receives from its neighbours, which are usually spaced about 3 kilometres from one another. The signals are transmitted at a constant strength, and should fall off in a predictable way with distance – when it isn't raining.
Using data provided by T-Mobile, Aart Overeem of the Royal Netherlands Meteorological Institute in De Bilt and his team came up with an algorithm that measures the difference between the expected signal strength and weaker signals received when raindrops interfered.
Rainfall snapshot
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
There are roughly 8000 microwave links in the Netherlands, and the team was given access to data on about 2400 of those, with signal strength recorded every 15 minutes – enough to get a snapshot of rainfall across the whole country. By contrast, the meteorological institute has just 32 rain gauges that take a reading every 10 minutes.
The team mapped rainfall between June and September 2011 and found that readings derived from cellular data tended to agree with those from the traditional combination of rain gauges and radar.
Mary Lynn Baeck, a hydrometeorologist at Princeton University, says that alternative approaches to taking weather data are needed in poorer countries where urban populations are expanding. Overeem's technique "has the potential to give good quantitative rainfall estimates for real-time hazards forecasting, as well as regional and global climate model analysis in regions of the world where the impact could be great," Baeck says.