The knowledge of where thunderstorms are occurring can be of great value to public safety due to the hazards which such storms can present to various human activities. There are, of course, the obvious dangers associated with the lightning strokes themselves. However, there are also other factors associated with thunderstorms which can present their own particular hazards, such as sudden intense rainfall, large hail and tornado activity. The Birmingham tornado in 2005 and the Boscastle flood in 2004 were both associated with strong thunderstorm activity. Other factors also associated with thunderstorms, such as wind shear, turbulence and severe icing, can also be of particular concern to the aviation industry.
The Met Office designed and built "ATDnet" long range lightning location system is capable of locating thunderstorms at distances in excess of 10,000km and is operational 24 hours a day, 7 days a week.
How can lightning be detected?
When a stroke of lightning occurs, a pulse of electromagnetic energy is produced which propagates, at nearly the speed of light, in every direction away from the stroke location. Some of the energy in this electromagnetic pulse is manifested as a bright flash of light in the visible electromagnetic spectrum - this is the lightning "flash" which we can see. However, the electromagnetic pulse associated with the lightning stroke also produces electromagnetic emissions in a wide band of radio frequencies, many of which are outside of our visual range. The ATDnet system detects the electromagnetic pulse in the very low frequency, or VLF, waveband - at much lower frequencies than normal radio waves. These electromagnetic pulses created by lightning activity are known as 'sferics', from the word 'atmospherics or "atmosferics".
These "sferics", or "sferic waveforms," are capable of travelling great distances because they propagate by "bouncing" between the surface of the Earth and particular layers of radio-reflective charged particles in the upper atmosphere of the earth, in a region known as the ionosphere. This surface/ionosphere duct is known as the "waveguide" and sferics can travel in the waveguide in a similar way to light travelling within a fibre optic cable.
Sferic waveforms can therefore travel vast distances as they can bounce several times in the waveguide with relatively little signal attenuation. The different sensors in ATDnet system network pick up these sferic waveforms in the VLF band and since each sferic waveform has its own characteristic shape when digitised (due to the different ways electric charge breaks down in each particular lightning stroke event), this unique waveform shape can be used to identify particular lightning strokes arriving at the different sensors in the network.
A single sensor will not be able to determine the location a sferic, but a sferic is detected at a minimum of four sensors in the ATDNet system network, the differences in the sferic arrival times at each of the sensors, whose arrival times are timed to better than millionths of a second, can be used to determine the origin of the lightning stroke unambiguously, using an Arrival Time Difference (ATD) technique.
The ATDnet system currently consists of a network of 11 operational sensors positioned in various locations around the world, mainly in Europe, but some further afield. These sensors detect lightning strokes at great distances from their actual locations and feed data pertaining to each lightning event back to a central computer in the Met Office Headquarters which then correlates these data and uses it to determine the origin of the lightning stroke.
Arrival Time Difference
If you have a number of lightning sensors placed around the world, the sferic waveforms they sense will arrive at those different sensors at slightly different times, depending on where the lightning strokes originate in relation to the geographical location of the sensors in the network. The differences in arrival times will be very small as the sferic waveforms propagate very close to the speed of light. The difference in the time taken for the sferic to reach one sensor relative to another is called the ATD (Arrival Time Difference).
If one sensor is chosen as a reference, each of the other sensors will have an ATD relative to this. These ATDs are very specific values, meaning that there are only certain locations where a stroke could occur which could cause that specific arrival time difference between the reference sensor and the sensor in question When all the possible values (points) of equal arrival time difference are plotted on a representation of the earth's surface, they form a hyperbola . Then, if this process is repeated for the same sferic waveform but using the reference station and another sensor in the network which sensed the particular sferic waveform, another hyperbola of arrival time difference can be calculated which will intersect with points of the first hyperbola- and these points of intersection show where the sferic could originate. But, as there could be more than one point where the two hyperbolae intersect, this means that there is potential ambiguity in the fix location- but this ambiguity can be resolved by repeating the whole process again so that a third hyperbola can be drawn, intersecting the first two hyperbolae and resolving any ambiguity. The whole process is repeated again using all the sensors in the network which sensed the particular sferic waveform, utilising the same reference station. The place where all of the hyperbolae intersect determines the location of the lightning stroke and this location is known as a "fix".
How accurate is the system?
The system is designed to automatically quality control the output to try and prevent incorrectly located strike locations from being displayed in products sent out to our customers. The false strike location rate must be kept low for our customers to maintain confidence in the system.
It can be difficult to measure the actual location accuracy and detection efficiency of the system (detection being a measure of the the number of lightning stokes which actually occur measured against the number of strokes which the system actually senses) as there are relatively few ways to know accurately what the lightning situation is in particular areas at particular times. However, there are other countries with lightning detection systems in place measuring more localised storms (usually within a range of approximately 200km of the sensors) and the Met Office has compared ATDnet output to some of these systems in the past giving confidence that ATDnet is working well.
The location accuracy of each lightning stroke is also determined by the system as part of the fix location calculation and is given as a theoretical value (in km). These values show ATDnet location accuracy to usually be within the following limits:
United Kingdom = 1.0-3.0 km
Western Europe = 2.0-5.0 km
Eastern Europe = 2.0-10 km
Eastern Atlantic = 10-15 km
Central Africa = approximately 20 km
South America = approximately 30-50 km
Examples of ATDnet detection:
The following images give an idea of the extent of the coverage of ATDnet. These data are
useful when used in conjunction with satellite imagery and radar precipitation data to show particularly active areas, and is especially useful in data sparse regions such as the Atlantic Ocean.
The Infra Red (IR) overlaid hemispheric images show the correlation of ATDnet data with cold cloud tops. The lightning strokes are displayed as hourly groups of different colours, purple being the oldest and red being the most recent (at the time of the IR image). This shows how storm systems are moving (continuous colour sequence tracks), developing (red but no previous colours) and dying out (previous colours but no red).