CBS-CIMO Remote Sensing/Doc. 4.1(3), p. 1
WORLD METEOROLOGICAL ORGANIZATION______
Joint Meeting of
CBS Expert Team on Surface-based
Remotely-Sensed Observations
(First Session)
and
CIMO Expert Team on Remote Sensing
Upper-air Technology and Techniques
(Second Session)
Geneva, Switzerland, 2327 November 2009 / CBS-CIMO Remote Sensing/
Doc. 4.1(3)
(16.XI.2009)
______
ITEM: 4.1
Original: ENGLISH ONLY
REMOTE SENSING SYSTEMS (SUCH AS GPS, MICROWAVE, AND INFRARED RADIOMETERS, LIDARS, ETC.) FOR THEIR USE IN WIGOS
Update of the CIMO Guide – Locating the source of lightning events
(Submitted by the Secretariat)
Summary and Purpose of DocumentThe document presents a fully revised version of the CIMO Guide Part II, Chapter 7 that was kindly provided by Mr H.-D. Betz. The original chapter was called “Locating the sources of atmospherics” while the revised chapter is called “Locating the sources of lighting events”.
ACTION PROPOSED
The meeting is invited to review the information provided in this document and advise on its suitability to be published in the CIMO Guide. The meeting will further be invited to provide any amendments and corrections to this chapter which would be needed.
1
CBS-CIMO Remote Sensing/Doc. 4.1(3), p. 1
CHAPTER 7
Locating the Sources of Lightning Events
Contents:
7.1General
7.1.1Lightning Discharges and Definitions
7.1.1.1Lightning Phenomena
7.1.1.2Definitions
7.1.1.3Stroke Strength
7.1.1.4Lightning Rates
7.1.2Meteorological Requirements
7.1.3Observation Methods
7.1.3.1LF and VLF/LF-Techniques
7.1.3.2VHF-Techniques
7.1.3.3Time-Of-Arrival Methods (TOA)
7.1.3.4Direction Finding (DF)
7.1.3.5Stand-Alone Detectors
7.2Lightning Detection Networks (LLS)
7.2.1VLF/LF-Networks for the Detection of Cloud-to-Ground Strokes
7.2.2VHF-Networks for the Detection of Cloud Lightning
7.2.3VLF/LF Technique for the Detection of Total Lightning
7.2.4Global Networks
7.2.4.1Ground-Based Global Networks
7.2.4.2Space-Born Observation
7.3Detection Efficiency (DE)
7.4Location Accuracy
7.5Network Comparisons and Practical Implications for Operational Systems
7.6Utilization of Lightning Location Systems by Meteorological Services
7.6.1Storm Recognition and Alarm for Severe Weather
7.6.2Derived Products, Nowcasting and Forecasting
7.6.3Lightning and Climate
7.6.4Verification of Lightning-Induced Ground Damage
7.7Supra-Regional Networks and Unification of Lightning Data
7.8Research in Atmospheric Electricity and Open Questions
References
7.1General
7.1.1Lightning Discharges and Definitions
7.1.1.1Lightning Phenomena
Lightning phenomena can be observed in many different ways. In addition to visible light and audible thunder, lightning discharges change the static electric field and emit electromagnetic radiation in a very broad spectrum, ranging from Hz (Schuman resonance) to GHz. This chapter focuses mainly on the detection and location of EM signals in the two frequency bands around 10 kHz (VLF) and 100 MHz (VHF), which are exploited by most operational Lightning Location Systems (LLS).
The global electric circuit regulates the movement of electrical charges in the atmosphere. Under fair weather conditions the vertical static field amounts to approximately 100 V/m near the Earth’s surface. When strong convection occurs, collisions produce charged particles and updrafts separate charged regions so that electric fields between ground and clouds may rise to the order of some 50 MV. Lightning discharges are the consequence, though due to the km-distances involved fields of this magnitude cannot directly induce lightning but give rise to certain initiation processes that are still under investigation (see 7.2.3 and 7.8). Separation of large amounts of charges causes a variety of electrical recombination phenomena that represent intricate effects of high complexity.
Electrical discharge activity generally starts with the formation of leader channels, creating a conductive path and depositing charge along the channel; typical leaders, especially the negative ones, proceed in steps thereby producing VHF emission. Some of the leader channels contact regions of opposite charge and give rise to a sudden charge neutralization, producing EM radiation dominantly in the VLF/LF band. When the contact is with ground, a return-stroke (CG, cloud-to-ground) occurs, and when two charge regions within the cloud connect a cloud-events takes place (IC, in-cloud), also named K-change, recoil streamer or IC-stroke. In both cases reasonably long channels (km) carry a short (s) and strong (kA) current. A complete lightning flash exhibits many additional neutralization features, which are described in the literature (Volland 1982; Rakov and Uman, 2003); in particular, subsequent leaders and strokes may be produced, susceptible to detection by LLS. Furthermore, continuing currents of smaller strength but longer endurance occur that are more difficult to detect.
A certain fraction of EM lightning pulses excite modes in the Earth-ionosphere wave-guide and can be detected in distances of up to ~104 km from the stroke. These signals, often referred to as sferics, can be used by global networks (see 7.2.4). Furthermore, leader formation and neutralization currents heat up the channels so that optical radiation is emitted, which can also be viewed from space. Special satellites are equipped with Lightning Imagers that can detect and locate the sources (see 7.2.4.2).
7.1.1.2Definitions
The following definitions relate to lightning and its detection:
Cloud flash: flash that does not contact to ground.
Cloud-to-ground flash: flash that contains at least one return stroke.
Cloud stroke (IC): short (s) and intensive (kA) current in a long channel (km) within a cloud or between two close clouds; part of a cloud flash; many IC events such as K-changes and a variety of other recombination processes can occur during a cloud flash.
Cloud-to-ground stroke, or return stroke (CG), or strike : main pulse (s) of strong (kA) electric current in a lightning channel connecting to ground; many CG strokes can occur during a ground flash.
Direction finding (DF): locating procedure that uses the direction of arrival of a lightning signal.
Discharge: often used synonymous with flash.
Event: specific part of a flash, typically any isolated signal measured during a flash, needs specification.
Fix: estimated location of a stroke as deduced from a network (mainly used for global networks).
Flash, lightning flash: complete neutralization process that involves many electric events (leaders, strokes, K-processes, continuing currents etc.) within a time interval of up to ~1 s; refers to a cloud flash or a ground flash.
Global network: LLS that employs sensor baselines of the order of 1000 km and detects signals in a narrow range of the VLF band.
Leader: charged conducting channel formed after the initiation process.
Lightning Location System (LLS): network for detection and locating of lightning sources; most modern LLS detect mainly leader sources or strokes.
Multiplicity: number of return strokes during a CG flash.
Polarity: CG strokes are either negative or positive, according to the sign of the electric charge lowered from the cloud to the ground.
Return stroke (CG): stroke that connects to ground (see: cloud-to-ground stroke).
Sferic, or atmospheric: signal from a lightning stroke that travels over long distances.
Source point, or leader source point: place of origin of a leader step.
Stroke: see CG-stroke and IC-stroke.
Time-of-arrival method (TOA): locating procedure that uses the time of arrival of the stroke signal measured at an array of receivers.
Total Lightning: comprises all lightning flashes; a network is said to have total-lightning detection capability when it reports all flashes independent of whether they are cloud flashes or contain CG strokes. A cloud flash is reported when at least one of its components is detected and located.
7.1.1.3Stroke Strength
Lightning strokes are characterized by their strength, with amplitudes of negative or positive polarity. The commonly applied definition of the strength is the peak value of the electric current in the stroke channel, measured in units of amperes (A). Typical magnitudes of stroke strengths are tens of kilo amperes (kA).
Direct measurements of stroke currents have been performed at instrumented towers hit by lightning. It must be emphasized that this type of lightning, termed upward lightning, is induced by elevated objects and represents quite rare strikes; it is different from the usual downward lightning that hits ground. Lightning location networks operating in the VLF/LF regime measure the electric or magnetic far-field radiation (in V/m or Tesla), requiring a transformation to kA. Available evidence supports the approximate validity of the established scaling procedures, at least within an accuracy of some 20-30%. In a network, the strength of the return strokes is usually obtained as an average of the values obtained from the contributing sensors, partly smoothing substantial differences among individual measurements, and range-normalization according to 1/D, where D signifies the distance from the lightning. In the near field, i.e. below ~30 km distance from the lightning source, irregularities must be expected.
Recent literature often suggests that the first return stroke of a flash has a typical strength of ~30 kA and, in a statistical sense, exceeds the strength of subsequent return strokes within a CG flash. By contrast, LLS all over the world do not find a significant difference between the strengths of first and subsequent return strokes, and report a typical strength of ~10 kA (somewhat less in more efficient networks) independent of the stroke order. Part of the apparent discrepancy may be due to different rise times of current surges in newly created as opposed to existing channels. Consequently, first and subsequent return strokes may have to be scaled somewhat differently. A final clarification for more refined practical scaling in networks is not yet available.
As regards cloud strokes, no reliable scaling procedure is at hand. However, since signal shapes of IC strokes often – though not generally – resemble the ones from CG strokes, it is common practice to scale IC-strengths in the same way as CG-strengths. Stroke-type dependent scaling or other modifications, including more rigorous justifications, remain to be found.
7.1.1.4Lightning Rates
The intensity of a thunderstorm may be described in terms of the electrical activity involved, appropriately summed over given time intervals and observed as a function of time. In a second step, the derived rates can be related to the observation area so that rate densities are obtained. When one disregards quantities such as sound or optical light and considers VLF/LF and VHF emission, three quantities can be measured:
a) Number of stepped leaders: VHF techniques allow mainly the determination of negative leaders.
b) Number of strokes: both CG and specific IC strokes can be counted with VLF/LF techniques.
c) Number of flashes: leader channels and strokes can be grouped to flashes, based on experimental evidence that a flash last less than ~1.3 s (mostly a value of 1 s is used) and flash components are usually located within an area of ~10 km. It must be noted, though, that flash numbers do not precisely reflect the entire electric activity, because details are lost; for example, differing multiplicities do not enter into flash counts.
As regards leaders and strokes it must be kept in mind that observation and locating are possible only when i) the emitted intensity is high enough and ii) the distances to the involved sensor positions are small enough to allow detection above threshold (see 7.3).
A meteorologically important measure is the IC/CG ratio during a storm. Comparable quantities are CG and IC stroke rates, both determined from VLF/LF networks. Alternatively, IC/CG ratios can be derived from flash rates, when either leader channels from VHF systems or strokes from VLF/LF systems have been grouped to flashes.
7.1.2Meteorological Requirements
Locating regions of thundery activity by means of real-time lightning data provides the meteorologist with valuable information. Especially useful are data for very large areas so that displacements of storm cells become visible and, thus, developing threats can be estimated. It can also provide clues as to the instability of air masses and the location and movement of fronts, squall lines, tropical storms and tornados. The related data quality is detailed in chapter 7.7.
Meteorological services to aviation have become important almost everywhere in the world. Thunderstorms represent a major hazard to flying because of vigorous air motions, hail, and lightning strikes. Although weather radar and steadily improving satellite images display storm cells, the thundery nature is readily and uniquely identified by lightning data. This enables, for example, the determination of storm-free corridors. Similar considerations apply to the launching of spacecraft. Of course, all commercial and public outdoor activities can benefit from sophisticated lightning observations and associated nowcasting.
Storms producing severe weather conditions are known to exhibit in most cases greatly enhanced IC rates. For this reason, meteorological services can issue storm warning on a better basis when lightning data comprises efficient reporting of cloud lightning (IC strokes, K-changes or stepped leaders).
Other sectors where lightning data is helpful by timely storm warning concern vulnerable installations; examples are airports, transmission lines or wind turbines. Adequate nowcasting allows definition of hazardous time spans for closure of airports or interruption of outdoor work, or initiation of preventive measures. Many current research projects are devoted to the improvement of nowcasting that includes lightning data.
7.1.3Observation Methods
7.1.3.1LF and VLF/LF-Techniques
Most LLS utilize the VLF/LF band from ~1 kHz to several 100 KHz, whereby dominant emission comes from ~10 KHz (wave length ~30 km). As long as ranges shorter than ~600 km are relevant, the detected signals represent ground waves that propagate with distance, D, according to ~1/D. They are quite indifferent to obstacles, and follow the curvature of the Earth. Receivers use simple rods or loops as antenna for the electric or magnetic field of the lightning pulse. The electric field delivers the stroke amplitude and its polarity; detection of only the magnetic flux yields the polarity only after the stroke location has been determined in the network. Depending on the noise level and system thresholds relatively small signals can be detected, but stroke locating succeeds only when the signal arrives at a sufficient number of sensors. Thus, the network efficiency depends also on the baseline of the receivers (see 7.3). Stroke locating is achieved with DF, TOA, or a combination of both.
When sensor baselines are of the order of 150 – 300 km strokes can be located with adequate efficiency in the network area and – depending on the techniques applied – a certain fraction of IC strokes becomes detectable. In addition to event time and location, stroke type, strength and polarity can be reported. In the global networks with much larger baselines of the order of 1000 km the spatial coverage is enormous, time and position can be reported using TOA, but all other stroke details are lost.
7.1.3.2VHF-Techniques
Lightning discharges emit EM radiation in a very broad spectrum. The radio source band around 100 MHz has been exploited for the detection and ranging by interferometry (DF) and, more recently, by TOA techniques, which are sensitive to bursts and pulses, respectively, emitted mainly by negative stepped leaders(Lojou 2006; Lojou et al., 2008). Depending on the signal strength and the detection efficiency, a certain number of radio source points can be located, ranging from zero to ~1000 per leader channel. This kind of channel mapping can provide fine details of complex channel formation during a lightning discharge. Given the locations, grouped source points define leader channels that can be counted, and grouped channels describe a flash. Though highly qualified for lightning research, VHF networks are much more complex than LF/VLF systems and produce more detailed data than what is needed for meteorological services.
High-speed video observations are suited for photographing leaders and reveal even more details than VHF mapping; a typical flash shows a wealth of leaders, but not all of them find a contact to charge regions that allows flow of substantial currents and a final stroke. Thus, VHF data is a measure for electric cloud activity, but the actual lightning strokes are not measured. When a leader channel is detected it is not necessarily clear whether a stroke will occur later; in particular, when a CG stroke is produced, VHF observation does not identify the stroke or its type. Therefore, an additional VLF/LF system is required for CG reporting; this additional LLS must also be able to discriminate the stroke type because it is also sensitive to IC strokes with amplitudes above the instrumental threshold. Some limitations of VHF networks arise from the relatively short range, because the VLF signals propagate only along the line of sight and attenuate according to ~1/D².
7.1.3.3Time-Of-Arrival Methods (TOA)
TOA techniques measure the arrival time of a lightning pulse at the detector as accurate as possible. Since the travel times from the lightning source to the employed detectors depend on the respective distance, the measured arrival times will differ characteristically. Least-squares procedures, standard in high-precision data analysis, allow determining the lightning source point such that the differences between measured and recalculated travel times are minimized. The small differences involved require adequate synchronization of the sensor clocks; nowadays GPS receivers readily grant this, being accurate to ~100 ns. Nevertheless, final timing of lightning pulses may represent a difficult task especially when pulse shapes are complex and – due to propagation effects – differ at different sensor sites. Finally, differing propagation conditions influence timing procedures. Typically, an overall time precision of 1s or less can be achieved, allowing location accuracies of the order of 300 m. Of course, mismatches can occur that give rise to entirely false locations. For this reason, advanced LLS employ both a sufficient number of sensors and sophisticated analysis software. In principle, four non-redundant reports suffice for a 2D-location. Determination of 3D-locations and redundancy of solutions suggest the use of more than four sensors.
7.1.3.4Direction Finding (DF)
Since the EM field is a vector field, employment of orthogonal loops allows measurement of the horizontal components of the magnetic flux and, thus, of the direction of incidence. In principle, bearing angles from only two sensors allow calculation of a lightning source point, but some redundancy is recommended in order to improve the location accuracy. Unfortunately, so-called site-error effects introduce substantial uncertainties that depend on natural and technical factors of the local environment, not necessarily constant in time, and it becomes difficult to keep inaccuracies below ~1°. Experience with LLS based on the use DF show that many mismatches occur, which are annoying for forecasters. For this obvious reason, use of DF in VLF/LF networks, and corresponding interferometry in VHF systems are subject to much larger errors as compared with modern TOA techniques and are not recommended any more for modern high-precision systems.