The Editor's Blog
About Us
Fire Photography
Fire Reports and Photography
Fire Information Links
Media Information Links
Weather Information Links
Advertise With Us
News from JEMS
Download Issue 2
Latest News
Visit Our Sponsors
LAFD Fire Blog

Lost Password?
No account yet? Register
Please Note: This Site Is Currently Being Upgraded - If it looks a little sparse, it is because content is still being transferred from the old site.  Thank you for understanding

Dry Lightning PDF Print E-mail
Written by Janice Coen   
_9en2007low.jpgOn June 20, 2008, an unusual early season coastal thunder and lightning storm moved in from the Pacific Ocean across northern California. According to various sources, in less than a day over 6000 lightning strikes were recorded in northern California resulting in at least 842 distinct fires that were fought by over 25,000 firefighting personnel from local, state, federal, and international sources.  These strikes were identified as ‘dry lightning’, a term that can be misleading. In fact, dry lightning occurs throughout the western U.S. including southern California and has played a role in numerous large wildfires often causing multiple ignitions in remote locations.

“Dry lightning” refers to lightning strikes that occur without precipitation such as rain or hail reaching the ground.  The term is misleading because the clouds and thunderstorms that cause it do produce precipitation (collisions between particles in the cloud is what builds up the electric charge) but the precipitation evaporates before it reaches the ground.  These "dry thunderstorms" are most common in the interior western United States during the summer and are notable for two reasons - this evaporation below the cloud can produce strong gusty winds at the surface, even microbursts (small, severe outflows of wind beneath clouds), and they are a common igniter of wildland fires.

Why in the west? Basics of a dry thunderstorm

As bubbles of air heated by the warm ground rise, they expand and cool. Cloud base occurs at the height where air has cooled to its dew point temperature. The air temperature decreases rapidly upward while the dew point does not decrease as quickly, and where the two become the same indicates where the cloud base will be. This can be hundreds of feet above the ground in humid areas or 2-3 miles above the ground in dry climates.

High-based thunderstorms form at times when the weather is very warm but relatively dry near the ground (i.e. the “relative humidity” is low). Because it is so dry, air must be lifted to higher levels to cool enough for water to condense to a liquid and thus for clouds to form.  Cloud base can be so high that when the rain falls it evaporates before reaching the ground -- but the lightning generated in the storm can reach the dry land below, even if the base of storm itself is miles high. When lightning does strike in a dry environment on fuel that has not been wetted by rain, a fire is ignited in fuel that is dry enough to spread the flames.

How does lightning form?  

Clouds become electrified by the vertical separation of positive and negative electrical charge between the top and bottom of clouds. Clouds vary greatly in their ability to become electrified and produce lightning -- much depends on the sizes and types of precipitation particles inside.

Charge separation appears to require the presence of water in two different solid ice forms and a liquid form.   These two different ice forms are (1) ice crystals that are growing by drawing water vapor from the air and small enough to be floating upward in the wind and (2) graupel, which are ice pellets under 0.5 inches in diameter that are growing by gobbling up cloud droplets with which they collide and falling down through the cloud. (You can think of graupel as small hail.)  The liquid form is “supercooled” cloud droplets -- i.e. cooled below 32 degrees F but still liquid.  This is important because clouds become electrified only where significant amounts of ice particles and supercooled water have formed at heights above the freezing level, evidence that supports our current theory about how this works (Latham et al., 2007).

We believe that positive and negative charges are separated during rebounding collisions between small ice crystals and larger, heavier graupel particles in the presence of water droplets.  During these collisions, negative charge is selectively transferred to the falling graupel particles and positive charge is transferred to the rising ice crystals, so that the net effect is for graupel particles to bring negative charge downwards in the cloud and for ice particles to take charge towards the top of the cloud. In general, this leads to the thunderstorm ‘dipole’ or ‘tripole’ with a negative charge region near the middle of the cloud and a positive charge region near the top of the cloud, and, in the case of the tripole, usually a small positive charge region below. Recent evidence suggests that ‘inverted polarity’ storms also exist with a midlevel region of positive charge with negative charge above and below.  This process still puzzles scientists as negative cloud charge tends to develop in cells or blobs, rather than the uniform layers previously theorized.

The lightning flash itself is an electrical discharge caused by the charge separation in clouds. Concentrated electric fields accelerate charged particles called “ions”.  Oppositely charged particles are drawn towards high points on the ground below. The attraction between positive and negative charges ultimately overcomes the air's natural resistance to electrical flow -- rushing together, they complete the electrical circuit. The path of a typical flash from cloud to ground is made by a sequence of “stepped leaders” that move some charge for 1/1000th of a second, pausing between steps for a longer time of about 50/1000th of a second. Within each step, the flash can shift direction toward a stronger electric field – this creates its crooked appearance. As a flash approaches several regions of opposite charge on the ground, it can branch into several paths. Approximately half of flashes contain subsequent strokes that terminate at more than one location. The mean separation of these strike points averages 2 km, with a maximum separation of approximately 7 km.  A careful look shows that just before reaching the ground, the stepped leader draws up surges of positive charge from sharp objects near the ground. The impulses meet tens of yards above ground and a connection is established. A “ return stroke” dashes upward much faster than the stepped leader's descent producing a bright flash as it heats air to 54,000 degrees F and creates the shock wave known as thunder. Some flashes end after a single return stroke, but more often other strokes follow (typically 4 about 1/20th sec apart, but as many as 20 have been detected continuing for on the order of a second) as continuous non-branching pulses take advantage of the existing path.

“Cloud to Ground” (abbreviated CG) is the term for this type of lightning strike that connects the cloud to the ground. Most lightning strikes do not reach the ground -- 80-90% of lightning is “Cloud to Cloud," (CC) which means the strike travels within the cloud (for example, from the negatively charged base to the positively charged top) or between clouds. “Cloud to Air” lightning is a bolt from a cloud into the air around it.

So, how does this apply to dry lightning? Dry lightning occurs in two situations. The first is the very high-based dry thunderstorm in which any rain generated in the storm evaporates before reaching the ground.  The second is lightning that comes down from a thunderstorm anvil (the high-level trail of clouds that flow out of the top of a thunderstorm) into an area where it is not raining. These anvil-to-ground lightning strikes can appear to strike out of an apparently cloudless sky after a storm has passed when in fact they originate from the anvil of a storm as far as thirty miles away.

So, is there anything different within clouds about how dry lightning forms, compared to regular lightning? Probably but scientists do not have enough evidence to say how.  For example, dry thunderstorms are often generated when enough moisture from the Gulfs of Mexico and California, known as the “North American monsoon” is drawn over an area to form stormy clouds at high levels of the atmosphere. One would expect the processes in-cloud are different because, at these higher altitudes, the temperatures - a very important factor – are cooler.  However, these are not always the source of dry lightning – occasionally it occurs in normal storms that are just not producing lots of precipitation.

Polarity of strikes versus ignitions?

_9en2037x1low.jpgWhile overall 80-90% of CG strikes carry negative charge to earth, the remainder are positive flashes. These powerful positive bolts carry as much as ten times the current of negative CGs and often last longer. They frequently emerge from high cirrus anvil clouds rather than from a storm's core.

Some storms produce many more positive flashes than usual. Numerous studies suggest that the presence of smoke, pollution or dust in the atmosphere encourages the development of anomalous proportions of positive flashes. Looking closely at the overwhelmingly negatively charged particles coming from prescribed fires in a variety of wildland fuels, Vonnegut et al. (1995) concluded that thunderstorms formed as a result of forest fires have probably grown from air that contains negative charges, rather than the usual positive ‘good weather’ space charge.  Electrification would proceed in the negatively charged cloud that grows from a fire in almost the same way that it does in an ordinary cloud, except that the charge structure is reversed, becoming one of the ‘inverted polarity’ storms mentioned above, with positive charge at cloud base, and it is to be expected the CG lightning will bring positive rather than negative charge to earth. The charge on these particles they measured could be reversed by reversing the atmospheric charge surrounding the fire.  The effect of fires on cloud electrification could also occur because of the smoke particles' chemical effect on the number and sizes of ice crystals within storms. Lyons et al. (1998) examined Southern Plains thunderstorms into which smoke from Mexican forest fires was flowing and found up to three times the usual number of positive flashes were observed in these smoke-altered storms.  In turn, positive strikes are thought to be more likely to ignite a wildland fire because, by transmitting a higher current for a longer time, they transfer a corresponding higher amount of energy to the strike point, making it more likely to reach the ignition point.  This poses the possibility of a feedback loop, in which smoke-altered storms are more likely to produce the polarity of strikes that ignite wildfires that in turn produce smoke that modifies clouds so as to produce more positive strikes.

Lightning Detection Networks

The U.S. National Lightning Detection Network (NLDN) consists of over 100 remote, ground-based sensing stations made by Vaisala Inc. that detect the electromagnetic signals given off when lightning strikes the earth's surface (i.e. Cloud to Ground lightning only). (Details about the sensors can be found at http://www.vaisala.com/weather/products/networks.html). These remote sensors send the raw data via a satellite-based communications network to the Network Control Center operated by Vaisala Inc. in Tucson, Arizona. Within seconds of a CG lightning strike, the NCC’s central analyzers process information on the location, time, polarity, and peak current amplitude of the first stroke in each detected CG flash. The lightning information is then communicated to users across the country. Lightning data collected by the NLDN from 1989 till the present is available for use in two different categories, real time and archive. Real time data subscribers receive live, second-by-second data on lightning activity within their own designated area of interest.  For more information on the development of NLDN, see Cummins et al. (1998) and Orville (2008). A sample of NLDN data displaying recent lightning activity is at http://thunderstorm.vaisala.com/explorer.html. Although real-time NLDN data is not available to public for free, some products from government agencies use it (see below).

Another device is National Weather Service’s Lightning Sensor, part of the Automated Surface Observation System (ASOS).  These sensors complement the NLDN data by detecting both cloud-to-cloud (CC) and cloud-to-ground (CG) strikes. The ASOS lightning sensor requires simultaneous radio (sound) and optical (visual) pulses. Range estimates are generated with the CG strikes by analyzing the radio signal strength. No directional capability is present, and no range estimates are available for CC strikes.

Dry Lightning Forecasting

First, the big picture. Only about 0.1 % of all CG strikes are associated with a detected natural wildfire, although these fires may consume many acres. A study by Hall (2007) shows that, climatologically, although the season of peak precipitation coincides with peak lightning production season, the number of natural fires produced decreases towards this peak, either because the fuel moisture increased during the previous month of abundant precipitation or because the number of number of wet lightning events increased or both. For example, in Arizona and New Mexico, peak fire season begins several weeks before peak lightning and precipitation season.  Thus, one might expect to find fires ignited by dry lightning on the leading edge of precipitation season.

What are the factors that result in dry lightning and how can we predict it? There are no complete answers, but guiding principles would include the following. First, there must be very dry air below the cloud to evaporate rain before it reaches the ground. This most commonly occurs in the sinking air accompanying particular weather events -- dry air gradually descends from aloft over an entire region resulting in a very dry and often very warm air mass. Although simple warm air masses may produce very warm conditions, these alone do not produce the extraordinarily dry conditions associated with dry lightning.  Secondly, it appears that the cloud bases during a dry lightning outbreak are several thousand feet higher than during a "normal" lightning event.  Thirdly, there must be a steep drop in temperature in the middle part of the atmosphere (the rate at which temperature drops with height above ground is called the ‘lapse rate’). Lastly, dry lightning events require significant upper level forcing, that is, a disturbance that kicks the cloud into growing. If all that is present is surface heating, the storms will be wet. The low level moisture that is required to produce late day thunderstorms by heating alone is sufficient to ensure that the rain produced will be only partially evaporated as it descends below the base of the storm.

How can you find prediction of dry lightning in the Fire Weather Forecast?

The National Weather Service (NWS) Fire Weather Forecast routinely includes the field Lightning Activity Level (LAL), which is a number from 1 to 6 that reflects the forecaster’s expectation about the coverage of thunderstorms and the frequency and character of cloud-to-ground lightning. While LALs ranging between 1 (no expectation of thunderstorms) and 5 reflect increasing cloud depth and sky coverage with thunderstorms, more intense rain, and more frequent and intense lightning, a LAL of 6 specifically forecasts that dry lightning is likely to occur in the fire weather zone.

What experimental lightning products are available?

Several products outside the official Fire Weather forecast can indicate whether wildland fires will be ignited by dry lightning in the near future. The first two focus on whether meteorological conditions favor dry lightning, a third focuses on effectiveness of lightning igniting a particular fuel type in its current condition, and the last integrates this assessment of ignition effectiveness with current lightning observations.

The NOAA’s NWS Office in Las Vegas, Nevada offers a weather forecasting product for their forecasting area (http://www.wrh.noaa.gov/vef/dlpi.php) called the Dry Lightning Potential Index.  It provides a quick visual image of the potential for and spatial coverage of meteorological conditions that might signal the development of dry lightning based on forecasts of afternoon low level humidity and atmospheric temperature lapse rate (the resistance to vertical motion). Qualifying factors like fuel moisture, the type of fuel, the presence of the needed upper level cloud-triggering mechanisms, etc., are not included. The product is refreshed twice per day by 4 AM and 4 PM PDT.

NOAA’s Storm Prediction Center produces short-term national fire weather forecasts and outlooks (http://w1.spc.woc.noaa.gov/products/fire_wx/overview.html).  Among the experimental products is a western U.S. forecast product for lightning and dry thunderstorms (http://www.spc.ncep.noaa.gov/exper/fire_wx/). This product is more complicated in its decision process in picking out locations of favorable atmospheric temperature lapse rates and low relative humidity in the lowest 2000 ft above ground. It also adds a valuable second graphic that highlights any upper level triggering mechanism.

The U.S.D.A. Forest Service Wildland Fire Assessment System (WFAS) produces a daily product called the Lightning Ignition Efficiency (http://www.wfas.net/content/view/18/33/) based on the fuel type (from 1-km current cover type maps), and each type’s susceptibility to ignition based on fuel depth (such as the duff layer in short-needled species) and/or 100-hour dead fuel moisture before the storm (more important for long and medium length needles). The ignition efficiency on a 1 km pixel is given on a per discharge basis - if the efficiency is high, then about 9 discharges will result in one ignition; if the efficiency is extreme, about 5 or fewer discharges will result in an ignition, according to this product’s web page. The algorithm built the ratio of positive and negative discharges into the calculation. Documentation of the Lightning Ignition Efficiency algorithm is in Latham and Schlieter (1989).

Also in WFAS, the Potential Lightning Ignition map (http://www.wfas.net/content/view/65/95/) integrates the Lightning Ignition Efficiency map with daily Cloud to Ground lightning strike data from the NLDN network discussed earlier. (See Sopko et al., 2007.) The Potential Lightning Ignition map splits the NLDN lightning strike data and efficiency calculations into separate datasets for positive and negative discharges. Then, the Potential Lightning Ignition values are calculated as the Lightning Ignition Efficiency of each 1 km cell multiplied by the number of strikes within that cell. Separate polarity based potential ignition values are calculated for positive and negative strikes and then recombined for the total potential ignitions per 1 km pixel. (Positive discharges yield higher efficiency values and thus increase the anticipated likelihood of an ignition.)  Under extreme conditions, 2 positive discharges (per pixel) would result in one ignition.

Currently the rainfall that occurs with the lightning-producing cell is not part of the efficiency calculation. The Lightning Ignition Efficiency map is based on the 100-hr fuel moisture value before the lightning storm. WFAS produces an experimental product called the Dry Lightning Map by merging daily estimated rainfall and NLDN CG lightning strike data (http://www.wfas.net/content/view/67/97/).  Although higher rainfall amounts correspond with higher CG strike rates, isolated pixels that have higher potential ignition values are where there is a greater likelihood of a dry lightning strike having occurred.


The author gratefully recognizes helpful comments from John Latham (NCAR and the University of Manchester).


Cummins, K. L., M. J. Murphy, E. A. Bardo, W. L. Hiscox, R. B. Pyle, A. E. Pifer, 1998: A combined TOA/MDF technology upgrade of the U.S. National Lightning Detection Network. Journal of Geophysical Research. 103,9035-9044.

Hall, B. L., 2007: Precipitation associated with lightning-ignited wildfires in Arizona and New Mexico.  International Journal of Wildland Fire. 16,242-254.

Latham, D. J., J. A. Schlieter, 1989: Ignition probabilities of wildland fuels based on simulated lightning discharges. Res. Pap. INT-411. Ogden, UT: U.S. Department of Agriculture, Forest Service, Intermountain Research Station. 16 pages.

Latham, J., W. A. Petersen, W. Deierling, and H. J. Christian, 2007: Field identification of a unique globally dominant mechanism of thunderstorm electrification.  Quarterly Journal of the Royal Meteorological Society.

Lyons, W. A., T. E. Nelson, E. R. Williams, J. A. Cramer,  T. R. Turner, 1998: Enhanced positive cloud-to-ground lightning in thunderstorms ingesting smoke from fires. Science. 282,77-80.

Orville, R. E., 2008: Development of the National Lightning Detection Network. Bulletin of the American Meteorological Society. 89,180-190. (Feb 2008)

Sopko, P. D. Latham, I. Grenfell, 2007: Verification of the WFAS Lightning Efficiency Map In: Butler, Bret W.; Cook, Wayne, comps. The fire environment--innovations, management, and policy; conference proceedings. 26-30 March 2007; Destin, FL. Proceedings RMRS-P-46CD. Fort Collins, CO: U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station. CD-ROM. p. 539-551.

Vonnegut, B., D. J. Latham, C. B. Moore, and S. J. Hunyady, 1995: An explanation for anomalous lightning from forest fire clouds.  Journal of Geophysical Research. 100(D3):5037-5050.

For more information:

Vaisala, Inc.’s National Lightning Detection Network, http://thunderstorm.vaisala.com
and http://www.vaisala.com/weather/products/networks.html, last referenced 4/3/09.

Lightning detection:   http://www.srh.noaa.gov/ohx/educate/lightning.html

Rakov, V. A., and M. A. Uman, 2003: Lightning: physics and effects.
3rd Edition.  Cambridge University Press. ISBN 0521583276, 9780521583275
687 p.

UCAR lightning basics and FAQ: http://www.ucar.edu/communications/infopack/lightning/faq.html