My take on this awesome shot of Ingenuity in front of two distant dust devils, and in animation.



Click to view attachment

A very beautiful dust devil was recorded by Mastcam-Z Right in the sequence of the 5th flight of Ingenuity.
Substracting MCZ frames by a frame without the dust devil makes it more visible.

Another day, another devil in dust alley:

Click to view attachment

(preview does not seem to play, click to animate)

https://mars.nasa.gov/mars2020-raw-images/p...0_01_295J01.png

a little late with this ... I just saw Andreas post this as well - post #54
Here is a gif of the best frame of the dust devil on Sol 84
Click to view attachment

Dust devil alert! (yes, another one). Sol 92, just about due west, right side of this image (NLF_0092_0675115622_792EBY_N0040136NCAM00698_00_0LLJ01_800).

Phil

Click to view attachment

Sol 92 LMastcam-Z, with a dust devil in the upper left corner:
Click to view attachment

Since this place looks like Dust Devil Alley... I wrote a little workflow program to look for them in the NAV sequences

This is SOL 84 ... (we've seen this --- eyes left) ... the gain is cranked in the program
Click to view attachment


This is Sol 78 ... The program found a near-invisible DD
Click to view attachment


(The blues are darker areas that change color before they are gained... it's a work in progress

The nice dust devil of sol 92 captured by MCZ-Right, contrast enhanced:



And an animation:

Click to view attachment

Another set of images captured by Hazcam and Navcam of one or possibly two dust devils, a good two minutes apart, on sol 102, in the favored spot, in front of the darker crater wall:

Click to view attachment

Here is a GIF of 21 images (NAVCAM LEFT) for 5 minutes around local noon on Sol 101
It is from a program that looks for dust devils
Short Story: there are none close by (but there is something going on on top of the deltas in the center)
With the gain turned all the way up -- I suspect I am seeing compression artifacts, exposure differences, sensor sensitivities

Click to view attachment

Sol 100 Dust Devil Program
21 image GIF - NAVCAM-LEFT - 5min starting at 2:15 pm local

(1) left of center on top of a delta
(2) right of center in front of the deltas (speedy)

Click to view attachment

A more stationary dust swirl on the center of the delta, on sol 107:

Click to view attachment

Click to animate. The frame spacing is about 14s but played at 0.5s for dramatic effect.

quickest workflow:

- filter raw images to select frames as specifically as possible.
- use bookmarklet to extract list of raw image urls
- copy list and paste as wget arguments on console to download, remove urls if necessary
- optionally enhance images with gimp or imagemagick
- upload images to ezgif.com
- use ezgif to crop, enhance, label and optimize animated gif
- download from ezgif.

Sol 107 Dust Devil Watch
Finer details still elude the detector so far ... but that delta is far away!
5 minute GIF starting 1:15pm local time
Scintillating bits are compression/stretching (?) or real (?)
Click to view attachment

Sol 111 -- Dust Devil Watch
Probable: One very far DD -- far left on the top of the crater rim - persists over several frames
Otherwise not much else
Gain all the way up - and then some - (very scintillating!)
GIF
Click to view attachment

Sol 114: a big dust devil was passing close to the rover
Click to view attachment

Hey, all. At the suggestion of long-time member PDP8E this thread has been created for everybody to post their images and animations of dust devils observed by Perseverance. Please be sure to include the sols when taken and times as well if they're available to make this data (and your great work) as useful as possible for any researchers who may take an interest. Thanks!

EDIT: I'll be moving several posts from the Perseverance threads over here to kick it off.

SOL 35 -- 1:20PM local
3 frame segment of a 360 degree quick peek (no DDs in the other frames)
Facing SW, the landing site is in the center, Kodiak is to the right.

Click to view attachment

A larger field of view including Kodiak hill showing the dust devil mentioned by Olivier:



Contrast increased:


And another dust devil, again on sol 114:

SOL 117
WOW -- No fancy detector required!
It is more of a GUST than a DD...

Click to view attachment

Two dust devils on sol 120 in a difference image.
The right one (black here because negative) was photographed at 13:40:26 local mean solar time,
the left one (white because positive) at 13:49:45. Its continuation into the sky is faintly visible.
Click to view attachment

Hmm, time to start naming the dust devils after the members of the band Kansas, perhaps after cities in Kansas, and crowd-sourcing a competition to see which Kansas grade school can count the most?

Dust In The Wind.
https://youtu.be/tH2w6Oxx0kQ

SOL 121 Dust Devil Watch
360 degree quick peek (3 frames each in 5 stances)
GIF
Click to view attachment

SOL 123 Dust Devil Watch
21 Frames -- stationary stance -- 1PM local -- 14 secs between images -- 4 minute span
GIF
Click to view attachment

... the detector still has issues with the DD parts in the Sky ...

SOL 127 Dust Devil Watch (not 128)
360 degree quick peek (3 frames each in 5 stances) -- 3:20 PM local
GIF
Click to view attachment

edit 06-30-2021 -- fixed SOL # from 128 tp 127 -- oops -- reloaded GIF

SOL 128 Dust Devil Watch
21 Frames -- stationary stance -- 11 AM local -- all quiet (only compression and sensor noise)
GIF
Click to view attachment

(I posted this initially to the EDL thread but can't seem to find it now, so here it is for posterity)

Perseverance recorded this DD during descent. Simeon Schmauss spotted this in LCAM footage and produced this stabilized and reprojected view, shown here at 20x real time. The DD is seen moving westward at roughly 8.5 m/s (20 mph) and was estimated to be about 25 m (75 ft) wide.

https://twitter.com/stim3on/status/1379575241970683908?s=20

Click to view attachment

Note the dark streak seen in the crater floor between the prominent foreground peninsula and the background peninsula. That got me and Simeon looking at HiRISE views of this area to see if that might be a transient feature, maybe a DD track. Sure enough, that track was visible in the first HiRISE view after Perseverance's arrival but was not present just six months earlier. That track is ~34m wide.

Click to view attachment

Across this broader HiRISE swath you can see dozens or hundreds of similar transient streaks, some estimated to be ~80m wide, predominantly oriented east/west.

Click to view attachment

This place is a popular playground for these devils!

Hi Ryan,
Thanks for those DD maps!
And a personal thank you for the RoverPics website over on github. Great Job!
Karle

SOL 132 - Dust Devil Watch
360 degree quick peek (3 frames each in 5 stances) -- 2:30 PM local
Improved noise reduction
Only one DD found - pretty quiet
GIF
Click to view attachment

SOL 134 Dust Devil Watch
360 degree quick peek (3 frames each in 5 stances) -- 8:45 AM local (early...)
Improved noise reduction
All quiet ... The mid-morning sun angle and lighting is interesting
GIF
Click to view attachment

SOL 137 Dust Devil Watch
21 Frames -- stationary stance -- 5:00 PM local
A new version of the detector - noise reduction / refined structure

A nice DD in the distance ----<oops see post #33>----
GIF
Click to view attachment

SOL 140 Dust Devil Watch
360 degree quick peek (3 frames each in 5 stances) -- 2:30 PM local

Nice DD in the SW stance (4th set)
GIF
Click to view attachment

on SOL 137 my dust devil detector was TRICKED!
I was checking the location of the SOL 137 'dust devil' and it wasn't there!

The sequence of 21 images was 5 minutes and 23 seconds starting at 17:05:27 PM (late in the day... long shadows)
Here is the first and last image in a GIF
Click to view attachment

The detector dilates the area around initial image differences and made it larger....
< see post #31>

I don't think this one was mentioned before: on sol 129 the Navcams caught a DD against the crater wall to the SW:

https://mars.nasa.gov/mars2020-raw-images/p..._195J01_800.jpg



Phil

SOL 142 Dust Devil Watch
360 degree quick peek (3 frames each in 5 stances) -- 3:45 PM local
Looks like the same place as SOL 140
Very Quiet (I turned down the denoise knob just to see if it is that quiet... it is)
GIF
Click to view attachment

SOL 144 Dust Devil Watch
360 degree quick peek (3 frames each in 5 stances) -- 10:30 AM local
All Quiet
GIF
Click to view attachment

SOL 145 Dust Devil Watch
360 degree quick peek (3 frames each in 5 stances) -- 4:30 PM local
All Quiet -- Late in the day. -- NO DDs
Click to view attachment

abbreviated GIF

SOL 146 Dust Devil Watch
21 Frames -- stationary stance -- 3:00 PM local
ACTIVE -- a brief surface dust-gust in the center -- a strong gust on the right that blows into a type of dd?
GIF
Click to view attachment

SOL 147 Dust Devil Watch
360 degree quick peek (3 frames each in 5 stances) --3:45 PM local
Quiet -- NO DDs -- Abbreviated GIF
Click to view attachment

SOL 148 Dust Devil Watch
21 Frames -- stationary stance -- Noon local
VERY ACTIVE -- EIGHT DDs
GIF
Click to view attachment

... I hope Ingenuity survived...

REVISED SOL 148
I revved the program to move it into HSV color space (RGB is not intuitive for saturation and luminance)
The dust devils are now the correct Hue (with the brightness turned up just 33%)
It also kills the annoying scintillation from the noise and compression artifacts.
Click to view attachment

Detecting Dust Devils with multiple images is fairly straightforward: average all the frames, and compare each image to the average -- use subtract, ratio, and other techniques.
Detecting faint DDs in one image requires a different approach: increase the contrast while retaining the image quality-- somehow.
This is a test image from SOL 15 (!). The technique is Stochastic Equalization / Adaptive Contrast (SEAC)
The image has a pronounced screendoor effect from the Bayer (in the sky) and has an overall 'old-timey postcard quality', but the DD is visible and the contrast is increased
Another image 30 minutes before had no artifact at this position.
--GIF
Click to view attachment

A storyboard for the dust devils from sol 148 (because the animated gif is too big for upload).

The difference between each image and the median of all 21 images is enlarged and then added to the median.
Since compression artifacts give large color blobs in the sky, I desaturated the sky.
The time markers start with the first image which was aquired at the local mean solar time of 12:01:50.

The large dust devil to the right of center in the first images seems to be quite close, because it casts a large shadow on the dune field in the foreground.
Later, the swirling dust around the brighter central column darkens the sky a bit.

Click to view attachment

That is an excellent presentation of the complex events on that sol. Thanks for posting it!

Phil

Tau, awesome work! (I hear you on the size issue!)

Some very faint dust devils are even better visible in the difference images.
The images are converted to gray scale to remove amplified color noise.
From 0 min 43 sec to 1 min 26 sec I can count seven dust devils active at the same time.

Click to view attachment

Another view of the SOL 148 Event
Annotated -- 9 dust devils and a few gusts.
The event is 5 minutes long and we have 21 seconds of images.
Click to view attachment

SOL 151 Dust Devil Watch
360 degree quick peek (3 frames each in 5 stances) --1:45 PM local
Several DD's
Click to view attachment

This is amazing. Anybody want to build a wind farm on mars? Poor insight lander.

Really nice dust devils on sol 151.

Click to view attachment

Click to view attachment