Lenticular Images
The increased use of smartphones and tablets has made it much easier to show videos in person-to-person settings such as conference poster sessions. This increased use has also led to an increase in damaged phones and tablets due to unfortunate drops. However, there is a more sturdy method of showing short animations of cloud, NWP, or student activity time lapses: lenticular cards.
Blank lenticular sheets can be purchased online, but it is important to check the lenses per inch (LPI) against a printer's resolution in dots per inch (DPI). If the former is not a multiple of the latter, the lines will not stay aligned with the lenses. For example, a 600 dpi printer cannot use cards with 70 LPI, but 50 or 75 LPI cards will work.
A simple and free program for producing lenticulars is StereoPhoto Maker. The pre-compiled binaries for Windows or MacOS contain a lenticular option under the Edit menu. Then, select the time-lapse stills you wish to use, enter the resolution of the sheets and printer, and multiple the number of inches width of your desired output by 100.
Most blank lencticular sheets come with an adhesive backing. When your image is printed, you can remove the plastic over the adhesive and lay it over the image. Lining up the lenses with the printed lines may not be easy, but with practice each person can develop their own technique that works. Now the card can be rotated from side to side to view the short animation of the bit of meteorology you choose to show.
Stereo Photogrammetry
Parallax, the relative change in an object's position due to a change in the viewing location, is used to determine a wide range of distances, such as between a barn owl and its prey or between the earth and its nearer stars. Handheld viewing of stereo pairs of photographs for a 3D effect dates back to 1851 (Jerald 2016). Collecting these pairs of cloud formations is particularly easy when traveling in an aircraft, and many such pairs are in the excellent cloud atlas of Scorer (1972).
Since the parallax is related to the z-distance away from the camera, this can be combined with a pixel's x and y location within a photo to give a geometric displacement vector away from the camera. If the camera's exact geographic location is known, the location of each pixel in the photograph can be determined. While this information about the camera location can be determined by scouting shooting locations in advance, Rasmussen et al. (2003) describe a method for calculating the values based on landmarks in the photos that can be determined from high-resolution satellite imagery.
The above assumes an undistorted image, but this is impossible due to distortions in perspective caused by the camera lenses. Photogrammetry software can include this camera calibration data in its calculations, but the properties need to be determined using a series of photos of a standard object. One of the most popular is a chessboard, where the corners of each square are skewed to parallelagrams as the view angle changes.
Stereo Photos - Terrain-induced Rotor Experiment (T-REX)
In the early and middle 1950's, the US air force funded atmospheric scientists at UCLA to conduct a pair of field programs studying mountain wave and rotor activity in Owens Valley, CA to the lee of the southern Sierra Nevada (Grubišić and Lewis 2004). The backbone of the instrumentation were a pair of sailplanes with cameras mounted viewing their instrument panels. This data and ground tracking of the planes using phototheodolites allowed researches to map the wave and rotor structure.
Fifty years later, scientists returned to Owens Valley with the Sierra Rotors Project in 2004 followed by the larger Terrain-induced Rotor Experiment (T-REX) in 2006 (Grubišić et al. 2008). The latter used up to three aircraft operating at different altitude ranges (Wyoming's King Air, UK Met's BAe146, and NCAR's Gulfstream V), in addition to an array of ground based Lidars, wind profilers, surface stations, and radiosonde sites. The AMS special collection on T-REX contains forty articles, and there are many more conference presentations on events during the experiment or using its datasets.
The Terrain-induced Rotor Experiment (T-REX) home page
T-REX also provided an opportunity for several researchers to test more experimental instrument platforms. Austria-Innsbruck operated an instrumented car, Yale used instrumented kites and a converted airplane K-band radar, and the Desert Research Institute collected time-lapse stereo photographs for photogrammetry (Grubišić and Grubišić 2007). Most of the sessions took place at a GPS-logged site in the Alabama Hills looking northward over the instrumentation area, though photos were also taken from a site nearer to the mountains and looking westward from the ridges of the downstream White Mountains.
IOP 6 - 25 March 2006
While only a limited amount of photogrammetric analysis has been performed on this dataset, the images are still available for the project's most active cases. The strongest by most metrics was Intensive Observing Period (IOP) 6 which resulted in the strongest (72 mph) wind gust measured during T-REX (though an 81 mph gust occurred during IOP 16 of the Sierra Rotors Project) and -2 g accelerations on the King Air which led to a temporary grounding. Stereo time lapse from that morning and afternoon show a dense foehnfall over the Sierra and a roll cloud displaced well to the east side of the valley. While the stereo session ended in the early evening due to fading light, a time lapse series was taken with a single camera during the peak of the event.
IOP 11 - 9 April 2006
However, the best events meteorologically may not be best suited for particular instruments. The Wyoming King Air used a dual-Doppler cloud radar during its T-REX flights, but the more broken roll clouds of IOP 6 did not produce the best returns. However, during IOP 11, a short and fiesty event preceding a cold front passage, a more extensive line of clouds formed over the center of the valley. Overlaying the Doppler radar data with a three-dimensional model of the clouds based on the stereo imagery is an ideal application of photogrammetry.
Most of the T-REX IOP's were characterized by these cumuliform roll clouds, but very few produced the photogenic altocumulus lenticularis often seen in the Sierra Nevada. One exeption was the first day of IOP 13 when the stereo imagery begins with a smooth, high-altitude wave cloud that dissipates over the next few hours to be replaced by a mostly southerly flow over the valley. If this cloud is as cirriform as it seems, it may be too far from the cameras to result in a usable parallax but the analysis is still worth attempting.
IOP 13 - 15 April 2006
Stereo Photos - Deep Propagating Gravity Wave Experiment (DEEPWAVE)
Mountain ranges in the Northern Hemisphere interrupt airflow with strong vertical shear that traps most gravity wave energy at lower levels. However, Southern Hemisphere ranges produce more vertically propagating energy that reaches into the middle atmosphere. To further study these events, the Deep Propagating Gravity Wave Experiment (DEEPWAVE) was conducted on the South Island of New Zealand during the summer of 2014.
The upper troposphere through the mesosphere was investigated by NCAR's Gulstream V aircraft, while the DLR's Falcon also flew at lower altitudes. There were also ground based profiling and radiosonde sites, including one operated by EOL upstream of the Alps at Hokitika. For a two week period, this was also the home base for a stereo photogrammetry project run by Saint Cloud State University.
The primary goal of the photography was to document the upstream cloud structure prior to a corresponding evening flight of the Gulfstream. This occurred during the first wave event over the island, identified as Intensive Observing Period (IOP) 3. During the afternoon session, a variety of cloud types could be observed, including a marine boundary layer exhibiting undulatus effects and a line of cumulus fractus upstream of the Alps (Billings et al. 2016). Only limited photogrammetric analysis was completed of the event, but estimates of the width of the nearer clouds did show that the required horizontal resoultion for idealized numerical simulations were higher than the resources available at the time.
13 June 2014 - Serpentine Road
In between research flights, the cameras were still deployed approximately every other day and documented a variety of weather events at locations up and down the western coastline. The day before IOP 3, the team set up halfway between Hokitika and Greymouth and faced the camera looking westward over the Tasman Sea. The rapidly evolving cumulus clouds carried by the low-level northerly flow were photographed at five second intervals. Osborne and Billings (2015) described the initial steps in the photogrammetric analysis. A specific future focus could be evaluating the relationship between the vertical wind profile as measured by the cloud motion to the wave heights visible in the foreground via Charnock's Law.
On a less synoptically active day with low cloud and drizzle, the team traveled south to the town of Ross. Along Woolhouse Road, the cameras were this time directed west toward the Alps. While the higher peaks were completely obscured, the lower ridges outlined lines of fog which had formed in the valleys. Over the course of the photo session the depth and extent of this fog varied and is another intriguing focus for a photogrammetric analysis.
7 June 2014 - Woolhouse Road