Assumed Risk and Safety
I want to start out with a discussion on lighting safety. For those of us that chase storms and photograph lightning, we operate under assumed risk. We know that, at some level, we are exposing ourselves to hazards that accompany severe weather, but we have determined that we are willing to risk the exposure because of our passion and desire to witness and document the incredible phenomenon that is weather. Educating ourselves on the hazards helps us to make informed decisions about our level of exposure and therefore better manage our risk. Unfortunately, lightning is a hazard that can literally strike very quickly with no ability to see it coming and get out of its way. We can, however, reduce the amount of time we are exposed and the level of exposure.
The following video was graciously provided by seasoned Australian storm chasers Clyve Herbert and Jane O’Neill with the help of Mike O’Neill (no relation).
I would like to point out a couple of things. First of all, it appears Clyve waits until after a flash occurs before he goes out to snap some pictures. Knowing that flashes tend to occur with gaps of time in between, he waited until there was a flash and then went out to take pictures, thus reducing the likelihood of another flash occurring while exposed. I do this as well. Although I did not include the longer clip, repeatedly he expeditiously takes images and then quickly returns to his vehicle to minimize the time he is in the open and to utilize the “faraday cage” protection that the metal of the vehicle provides.
Also, he clearly reacted to something before the close lightning flash occurred. He stated to Mike that he heard a clicking sound and sensed something was about to happen. The clicking was likely corona discharge (also known as point discharge) which is the development of ionization and streamers from pointed objects in the presence of a very strong electric field. Streamers are not hot (yet) and are the precursor to the thermalized leader. I have heard corona discharge from the antenna on my vehicle and nearby tall objects and even from antennae onboard a ship at sea. It is a clear signal that the electric field is very high and that a lightning flash can occur at any moment.
So what caused the chain to spark? The still image below with a wider field of view shows that there was a ground flash not far away from the vehicle.
It appears to terminate in some bushes and trees. However, it did not appear to strike the fence. However, the fence to which the gate is likely attached could extend in that direction and possibly have experience a direct connection. If so, the fence would have conducted some of the current along the metal fence. The gap in the chain at its attachment points, along with a possible resistance change due to the different metal type of the chain, increased the resistance resulting in an arc developing between the gap. If the channel did not connect with a fence, then current flowed through the metal gate due to ground current (also known as step potential). The difference in distance between the individual gate metal posts and the grounded channel resulted in a voltage difference in each metal post which caused current to flow through the gate.
Either way, if Clyve was closer to the gate or even worse touching the gate when it sparked, he could have been subject to substantial current. His awareness and reaction proved very fortuitous. He did state that he did still feel a shock as he did not get to the vehicle before the flash. This shock was likely current flowing through his body due to the step potential of his separated legs.
Again I thank Clyve, Jane and Mike for allowing me to share their experience. If you haven’t had the chance to read through the Lightning Safety section of my blog, I would encourage you to do so before continuing as I discuss the five different ways lightning can injure or kill you.
Slowing Down Lightning
So what as storm chasers and photographers can we see when it comes to lightning? With the knowledge gained from data collected and analyzed from high-speed cameras and correlated electric field sensing instrumentation, we actually can see and interpret quite a bit. It has gotten to the point for me, and those that have worked with me on lightning research, where we can frequently classify the polarity of leaders and ground flashes in realtime. It is as if lightning has been slowed down for us, slow enough to identify its behavior and personality. Sounds strange but it is true, and it is my hope that others will be able to experience this after reading this blog and watching for the distinct behavior differences.
If you have completed all the sections preceding this final section, you now understand that lightning is simply the development of hot, ionized, conductive leaders of plasma that develop and grow due to the high electric field generated between charge regions in clouds. If they initiate in the storm cloud, they develop bidirectionally with a positive and negative end. You can think of the positive end as having a deficit of electrons and the negative end a surplus. We have also shown that there are clear differences between the two polarities of leaders, and that these differences lead to differences in ground flash behavior.
Storms frequently exhibit a personality due to the conditions present during their development. For example, some times storms produce numerous cloud flashes and very few ground flashes. Jokingly, I call these non-cooperative storms. Other times the opposite is true, and a storm will infrequently produce ground flashes with little to no cloud flashes. Some times I see negative leaders repeatedly propagating along cloud base and have noticed that this tends to occur early or late in the season when the temperatures are colder. When an MCS passes over Rapid City, I can pretty much tell you when upward lightning is about to occur. When the towers are “in the zone” for upward lightning, there are time spans of 3-10 minutes between flashes, and when the flashes occur, they are horizontally extensive with brightness racing across a large portion of the sky.
The easiest way to identify positive leaders in real time is due to their flickering progression due to the recoil leader development along decayed branches. Anytime you see repeated strobing in the clouds, or flickering in the visible leaders, you are likely witnessing recoil leaders. If a ground flash exhibits multiple return strokes, chances are it was a negative ground flash, and that recoil leader development on the upper portion of the positive leader network following the return stroke led to multiple connections with the ground along the same channel path. Positive ground flashes tend to be single strokes, but frequently have branches connect with ground as well resulting in multiple termination points. However, negative ground flashes can do this as well. If you see the downward progression of a branched initial leader preceding the first return stroke, you probably just witnessed a branched, recoil leader producing, positive leader of a positive ground flash. In fact, before I started using high speed cameras, I called these “slow positive ground flashes” because I could see them propagating to ground. This was before I filmed recoil leaders for the first time, and therefore, had no idea that numerous recoil leaders were reionizing decayed branches as the leaders travelled downward. A colleague even referred to these as Medusa flashes with the many bright branches representing her hair of snakes. In fact, below is the first “slow positive” ground flash I captured on video back in 2000.
A single 60 image per second (ips) video frame last 17 milliseconds. In that time, multiple recoil leaders can initiate on a decayed positive leader branch. Because many of these recoil leaders fail to reconnect with the main channel, they illuminate only a segment of the decayed branch. Integration of all the recoil leaders produced in a 17 ms frame frequently appear as a detached bright leader segment that precedes the leader’s forward development. Our eyes integrate this as well.
The video segment below shows the development of an upward positive leader recorded at 7,207 ips with a high-speed camera as well as with a standard-speed video camera (60 ips). The high-speed recording resulted in 135 µs exposures (139 µs image intervals) and 17 ms exposures for the standard-speed recording. A total of 122 high-speed images were recorded during each standard-speed video exposure. The standard-speed video image is, therefore, an integration of the activity recorded by the high-speed camera during the 17 ms exposure. Annotations on the standard-speed video show the features that identify the leader as positive due to the recoil leader production.
The following is a time-integrated high-speed video segment that corresponds in time to a single standard-speed video image from the previously shown upward flash. The detached recoil leaders are clearly visible in both images.
Here are more standard-speed video images showing recoil leader development during upward positive leader propagation.
The decreased sensitivity of digital still camera sensors compared to video sensors and the longer exposure times used at night (i.e., 20 s) results in recoil leaders recording as faint leader segments. Below is a video showing positive leader development captured at 1,000 ips. Three different positive leaders of differing intensity show the spectrum of behavior modes exhibited by positive leaders. The weak positive leader (top) was weakly luminous, highly branched and produced numerous recoil leaders. The middle positive leader was brighter and only branched a few times near the end of the recording and produced fewer recoil leaders. The bright positive leader branch at the bottom did not branch and did not produce any recoil leaders.
The image of this event below shows how the spectrum of positive leader development appears when captured by a digital still camera. The image was captured using ISO 100, f/6.3 and a 20 s long exposure. Although the recoil leaders where intensely bright in the high-speed video, their short duration and the decreased ISO sensitivity of the digital still camera results in them appearing faint in the upper portion of the image. The non-branched lower leader channel remained brightly luminous during its entire development and this recorded as a brightly luminous leader on the still image.
Below are additional examples of positive leader development associated with +CG flashes as captured by digital still camera. The recoil leader producing positive leader branches are the primary indicator of leader positive polarity.
If you compare these still images with those of negative leaders shown below, you will notice that there are distinct differences.
My favorite type of lightning is that associated with the backside or trailing stratiform portion of a Mesoscale Convective Systems (MCSs). The challenge is that this area tends to have widespread rain so filming the flashes can be difficult. However, if you can get to the edge of the precipitation area of an active system, you will frequently be treated to spectacular horizontally extensive flashes that travel large distances and last more than one second. The reason for these types of flashes is that the charge regions in a well developed system tend to spread out over large distances in layers and the leaders propagate through these layers. +CG flashes tend to occur in this region as well and the resulting negative leader extension that frequently follows the return stroke tends to spread out over wide areas. Usually the negative leaders are incloud and seen as fast moving brightness that travels across the sky. However, positive leaders which encompass the other end of the bidirectional leader network will frequently chase behind the negative leaders in trail, below cloud base and clearly visible. These are the spectacular anvil crawlers or spider flashes that induce a collective “woah” by those that witness them. The positive leaders tend to be highly branched and therefore decay and produce prolific recoil leader activity, which displays a flickering forward progression that can be easily followed by eye due to their overall slower speed and immense travel distances. These are truly the displays that Zeus uses when at his angriest.
Below is a standard-speed video that shows a spectacular crawler flash that filled the sky with recoil leader producing positive leaders. The incloud brightness that precedes the visible leaders is due to preceding negative leader growth associated with the negative end of the bidirectional leader development.
A small portion of this huge flash was captured by a high-speed camera recording ant 7,207 images per second (ips), and in this video you can clearly seeing the positive leader meandering behavior, decay and subsequent recoil leader development on the decayed branches. I suspect that after you watch the video below, it will seem unbearably long compared to the standard speed version above.
So as a storm chaser and/or photographer, next time you are out filming lightning, see if you can determine what polarity and types of lightning flashes you are observing. I suspect that the more you do this, the slower the lightning will appear. Good hunting.
Equipment and Techniques for Photographing Lightning
I thought I would add a summary of the equipment and techniques I use for photographing lightning. I suspect that many that read this section are already capturing lightning on both digital still and video.However, given that I am asked frequently about my methods, I wanted to make them available.
Due to the different strengths and weaknesses of each camera brand, I use Sony, Canon and Nikon DSLR cameras for photographing lightning.
During the night, I shoot in manual mode with the aperture set at f/8 and ISO 100. I use an intervalometer and set the exposure time based on the ambient light of the scene. For shooting lightning in a city, the maximum exposure time is around 10 seconds. For darker scenes, I will extend the time to 30 seconds. If the lightning is very close, I will close down the aperture to f/11 or even f/16 if I think there is a chance for a return stroke to occur within 100 meters. If the lightning is beyond 5 km, I will open the aperture to f/5.6 or f/4 depending on the visibility.
During the day, I shoot in Aperture Priority Mode with the aperture set to f/8 and ISO 100. The exposure time is determined by the camera based on the ambient light, and the shutter is triggered by a Lightning Trigger which senses infrared light produced by lightning. Most lightning triggers can initiate the camera within 100 milliseconds depending on the brand of camera. Be sure to place the camera in multiple image drive mode so that multiple images can be taken during a flash. Also, for both day and night, be sure and focus the camera and then set the focus to off or manual so that the camera does not try to focus before closing the shutter. Also, disable the image stabilization and set a predetermined fixed white balance such as daylight or cloudy so the camera again does not have to “think” before taking the picture. The focal length of the lens depends on the field of view desired and relative closeness of the lightning that is occurring. I typically use 14-24 mm lenses for close lightning and 24-70 mm for more distant lightning.
I always shoot RAW so I can obtain the maximum amount of digital information and have the greatest latitude for editing and adjustment given that lightning brightness can vary significantly with each flash. I use Adobe Lightroom CC for my digital image editing workflow.
Rolling shutter is the nemesis of lightning videography. The scanline recording leads to awful artifacts as the rapidly changing light is recorded line by line while the scanline moves down the sensor. Cheaper and higher light sensitive CMOS sensors utilize rolling shutter recording and have become the standard in most video cameras. There are very few models out there that still utilize global shutter recording which is essential for recording lightning without artifacts. Global shutter recording exposes the entire sensor to light utilizing either mechanical or electronic shuttering and then at the end of the exposure time, the sensor recording is terminated across the entire sensor via the same method. While the next exposure takes place, the data recorded from the previous exposure is shifted to adjacent non-recording photosites (pixels) and is read out line by line.
I use a Panasonic HPX-170 video recorder that utilizes global shutter recording at 1280x720p60 and for 4K video, I use a Blackmagic Design Production Cinema 4K which can record 3840x2160p30 via a global shutter. I utilize Adobe Premiere Pro CC for my video editing along with Adobe Photoshop CC for animations and annotations for both my standard-speed video and high-speed video.
I use Vision Research Phantom high-speed video cameras for my research and personal high-speed videography. Models I have worked with include models v7.1, v12.1, v310, v711, Miro 4 and Miro LC321S. These cameras utilize CMOS sensors that incorporate global shutter recording. For lightning research, I use monochrome sensors and operate between 1,000 and 100,000 images per second. I find that 10,000 ips is the optimum recording speed for observing lightning processes, however, 50,000 ips or greater is required for resolving faster processes such as recoil leaders and return strokes. I use Phantom Camera Control software to work with the proprietary cine files, and then export Apple QuickTime mov files for post processing.
Other essential items for lightning photography and videography include a stable tripod and rain proof cover. The hazard with utilizing a tripod, is that you are touch the tripod when a return stroke occurs nearby, ground current (step potential) due to the distance separation between your legs and the tripod legs will result in current flow through the tripod and your arm and your legs. Therefore, I minimize the time I am touching the tripod.