In terms of precipitation measurement techniques, rain gauges, weather radar and satellite precipitation estimation are commonly regarded as the most widely used systems. However, very recently I have learnt about a new measurement system through an e-mail from a job recruiter; the company retrieves high-resolution precipitation data from microwave links of cellular networks. It sounds pretty interesting, so I did a bit of research on it, which might be also interesting for you.

blog_dec_o_01-copy
from https://pixabay.com/

The very initial study was conducted by Messer et al. (2006). The principle is simple: the 10-100 GHz microwave bands are weakened by rain so we can covert the “signal attenuation” into “rain intensity” using a function like the one used in weather radar. Several studies (e.g. Leijnse et al. 2007; Zinevich et al. 2009) have examined that the precipitation data from the mobile network system shows reasonably good agreement with measurement by gauges or radars, although the validation was conducted on only a few rainfall events. More recently, in 2015, Ericsson and partners have conducted a pilot study, the Microweather project, over 8-months. The results show that microwave links can be used as accurate, high resolution rainfall measurement tools.

erricson-copy
from the June 2016 issue of Ericsson Mobility Report

According to the June 2016 issue of Ericsson Mobility Report, the microwave links can provide higher spatiotemporal resolution (<1min, <1kmthan average gauge or radar measurements (~5min, ~1km2), with a closer observation (~the height of cellular towers) from surface compared to radar (> 500 m); the closer we are to the ground for observations, the more accurate measurement we can make on water at ground level – rainfall. Moreover, the extra cost for the microwave link system is small, since the required infrastructure, i.e mobile network, is already in place. Hence, this technique has high potential for areas where there is a poor weather station system, but a dense cellular phone networks (Gosset et al. 2016).

Advertisements