What are 'sferics', and how are they obtained?
When a lightning discharge occurs, radio waves are emitted over a broad spectrum of frequencies. For the vast majority of people, such 'atmospherics' (or 'sferics') are simply a nuisance, leading to the familiar 'crackle' that can be detected on a home radio set, particularly in the 'AM' medium or long wavebands.(@see note 1 below)
However, in the 1920's and 1930's, Robert Watson-Watt, a British scientist (and sometime employee of the UK Meteorological Office), developed a method of displaying the discharge information on a crude cathode ray tube, and by taking simultaneous observations on the same flash, the source could be located with reasonable accuracy. (@see note 2 below) This triangulation method continued in use in the United Kingdom until 1988.
The system now employed by the UK Met Office is the Arrival Time Difference (ATD) system. The origin of the lightning flash is computed from the time difference of an atmospheric arriving at several widely-spaced 'listening' stations - located across the UK, the Mediterranean & northern Europe. Each lightning stroke has an individual signal, or wave-form (in the Very Low Frequency part of the electro-magnetic spectrum), and by using accurate (atomic) clocks, and synchronisation between the detector stations, and the control station (Exeter, Devon (UK), originally at Beaufort Park, near Bracknell), an accuracy of at least 5 km (often down to 2 km) can be achieved, although in GTS SFUK bulletins (known within the Met Office as SFLOCs - SFeric+LOCation), the accuracy is limited by the code form to 0.5 deg lat/long.
One of the listening locations is automatically designated as a 'reference' (by the control station) for each discharge, and reports the time at which it detects a 'sferic' event to the control. The control station then 'asks' the remainder of the outstations for detailed wave forms of sferic events close to this time and calculates the time differences and so computes a set of possible locations. Provided more than three outstations are active (out of the 7 available) in acquiring that event, a unique location can be determined for that particular return.
The system is highly effective, though on a few occasions, 'spurious' returns can be seen (usually easily eliminated by reference to IR imagery); also when very active clusters are detected, isolated events elsewhere may not be adequately resolved - though these are now rare events. In addition, be wary of occasions when lightning returns 'appear' to all stop at once; this is usually due to a failure of the central computer, which needs re-setting before the system will function properly.
The system is fully automatic, and theoretically can detect lightning over a large portion of the globe. The system is in a state of continuous development - a major upgrade having been put in place in Autumn, 2001.
For more on the system see the Met Office Education page at this link:- http://www.metoffice.gov.uk/education/secondary/students/lightning.html
(@1:This means that an ordinary home radio set can be used as a crude lightning detector, by tuning to a portion of the waveband -- try the LW section -- that is not used by a broadcast station. During lightning activity, irregular crackles will be heard, and with a little experience it will soon be possible to pick out 'close' from 'distant' discharges by this method - a good reason to hang on to your old portable radios after the 'digital revolution'! )
(@2:This use of triangulation of signals was later adapted in his method of aircraft detection used during the early part of the second World War.)