Data Collection Recommendations

Recommendations for Optimizing Storm Chase Data Collection to Serve Scientific Research

Past scientific efforts to research tornadoes have benefited significantly from data contributions by storm chasers. The El Reno Survey project team has developed the following set of guidelines for storm chasers on data collection and documentation practices, based on what has been learned from analysis of imagery and data collected by chasers during the 31 May 2013 El Reno tornado. The purpose of these recommendations is to offer advice to motivated storm chasers on how to optimize data collection to support scientific studies. Chasers can contribute to academic research on future storms by incorporating as many of these measures as possible into their field activities. Doing so may also yield a more complete and organized archive of materials from each of your chases, and it doesn’t require much extra time. The basis for these recommendations comes from the El Reno team’s collective experience as chaser-researchers, and many lessons learned during our analysis of materials provided to our project by dozens of El Reno chasers. We have also polled several expert chaser-researchers for their inputs and suggestions.

BEFORE the chase:

  1. Set accurate time (to the nearest second, if possible) on clocks and all electronics with time-tagged recording. This includes dashboard clocks, video and still cameras, weather instrumentation on your vehicles and laptop computers.  Check that cell phones and tablets are properly synchronized as well. Bookmark this page to ensure you can get accurate time whenever you are online.
  2. Clean both interior and exterior sides of all windows on your vehicle, and check that windshield wipers wipe cleanly without leaving streaks. Apply Rain-X to minimize the need for wipers.
  3. Free up memory on computers, cards and camera hard drives.
  4. For chasers using video: We found that the best visual documentation came from chasers running mounted cameras continuously (rooftop and dash-mounted GoPro cameras give excellent results) supplemented by footage from a second camera used for zoomed sequences and panning.
  5. If possible, turn off the autofocus function on your video camera and set to infinity as your default.
  6. Pack a tripod and/or other mounts for your camera(s), and use them as much as possible for steadiness and sharpness of imagery.
  7. Ensure you have a means of recording your driving route. If using GPS, turn on the logging function. Another method is to make voice recordings, documenting road navigation while running video. Notes written on paper maps, post-it notes or a scrap pad can also suffice – the aim is to record location and time together.
  8. If running weather instruments, try to have a means of logging observations electronically at a frequency of once per second, if possible, or at least once per minute.
  9. Create a folder on your computer, named by the date and chase location, for consolidating all of your materials; many chasers already have their own systems for archiving materials.

DURING the chase:

  1. While filming a storm, re-check that focus is set on infinity and autofocus is OFF.
  2. Use tripods and camera mounts whenever possible, especially for video. If no mount is available, make efforts to minimize camera motion. Resting the camera on a fencepost or roof of your vehicle can be fairly stable; however, keep aware of lightning hazards. (i.e., don’t be the highest point around, or near a tall lightning attractor; and don’t place that aluminum-legged tripod on the wet ground with cloud-to-ground flashes nearby.)
  3. If possible, geo-tag all images and video with GPS coordinates.
  4. If possible, set your camera on a moderate wide-angle or normal focal length, rather than zoomed-in, unless the storm/mesocyclone is at some distance, or if there is a feature of particular interest.
    • When filming mesocyclones, moderate wide-angle views from mounted cameras generally give the best results, and are most useful for creating time-lapses.
    • Ultra-wide angle views (a popular option with GoPro cameras) have reduced utility for research purposes, due to their severe distortion creating a fish-bowl perspective.
    • For reference, on the GoPro Hero 3 camera, the focal length equivalent for the field of view (FOV) options are as follows: Wide FOV – 14mm; Medium FOV – 21mm; Narrow FOV – 28mm.
  5. While running sequences of zoomed-in video, pan out to wide angle occasionally to capture the broader storm structure.
  6. Incline your camera upward, so that only the lower 20% of the field of view is at ground level. Keeping some ground features in the foreground is important for fixing filming locations.
  7. Whenever possible, make oral notes at the start of each video sequence, giving time, estimated location and view direction. Example: “6:12 pm, we are parked a quarter mile south of the intersection of Smith Street and Road 15, looking southwest at a developing mesocyclone.” Do this for all video cameras used during your chase.
  8. If running mounted video while driving, or if a passenger is performing hand-held filming, make notes on the audio track to document driving route changes. Example: “It’s 6:23, we are turning right from Route 62 onto Osborn Rd.
  9. If you have sufficient camera memory available, when in a storm situation run video as continuously as possible, even if you place it on the car seat while repositioning to a new location. Setting the camera in a dash-cam mount is preferable. Even with a limited sky view, a running camera may record your in-car conversations, the time of lightning flashes and thunder, and provide inference on driving speed and stops at intersections – all of which could be valuable in post-analysis.
  10. Remember to use windshield wipers only when necessary, and to lower the volume on car radios, etc., to improve audio quality recorded on video.
  11. Whenever you observe a cloud-to-ground lightning flash, call out “CG” for the audio track on video and give the estimated direction relative to your location. We have found that CG flashes are an optimal means of synchronizing videos that lack time calibration. Usually, only a fraction of the CG flashes you note visually will be captured within your camera’s field of view, so the oral notations offer additional confirmation.
  12. Make similar oral notes for hail occurrences, noting the hailstone size. When possible, take photographs of large hailstones alongside objects of known size; carry coins, a golf ball, baseball, softball and a short ruler in your chase kit for this purpose.
  13. GoPro and similar fixed-zoom sports/action cameras give results superior to many conventional video cameras and are excellent sources of scientific-quality imagery.  As stated above, however, the widest field of view diminishes scientific utility due to high distortion.

AFTER the chase:

  1. Document as much as possible, as soon as possible. Memories fade fast!
  2. Same day:
  • If you have not done so already: create a folder on your computer, named by the date and chase location, for consolidating all of your materials.
  • Copy all video files, still images, GPS logs, screen grabs of radar, meteorological observations recorded, etc. to the folder.

Example of a logical file name for convenient indexing: “20130531_El Reno_Oklahoma”

  • Back up your chase folder by burning a DVD, copying to an external drive and/or uploading to your preferred cloud data storage resource. Doing so will allow you to clear memory cards for your next chase.
  1. As soon as possible, post-chase:
  • Type up notes to document the chase route, including rural roads and locations of stops, storm evolution and events witnessed, persons encountered, and so on.
  • There are many fine examples of what can be considered to be “good” chase documentation on personal blogs.  See here (Bill Reid), here (John Allen) and here (Dave Lewison) for examples of thorough chase documentation from some of our El Reno Survey participants.

Our team would like to thank Roger Edwards, Robin Tanamachi, Rich Thompson, and Jon Davies  for their reviews and inputs.

We plan to update these recommendations periodically based upon feedback received and new lessons learned. We will also prepare similar recommendations for researchers interested in working with crowd-sourced data.

Your chase data may turn out to be valuable to the scientific community!  We appreciate your help!