Wednesday, August 28, 2013

I am in Alaska

Fairbanks, specifically, and it's sooo pleasant here this time of year. I'm getting back into the groove of coming in to the GI each day to read and write. And of course I'm already planning weekend adventures. The drive up was uneventful, I took the 'direct' route, I70 -> I15 -> Alaska Highway, in order to save time. Definitely take the Cassiar Highway over the Alaska Highway if it's wilderness and adventure you want. I did timelapse the whole thing. Here are a couple of pictures from along the way:

Montana mountains
Mountain near Gates of the Mountains in Montana.

A somewhat boring image of the Kiskatinaw curved wooden bridge on the old Alaska Highway.

Canadian traffic jam
Canadian traffic jam.

Signpost forest in Watson Lake.

Monday, August 19, 2013

I am in Arkansas

Milky pond

I've been here since last Wednesday. Came to visit some people once more before I disappear to Alaska for four months. It's been a good time, though I must admit there are a few I'd like a little more time with, but c'est la vie. I plan to leave for AK on Tuesday or Wednesday and be in Fairbanks next Monday or Tuesday. We had a nice going away party last night, or as my friend Colby liked to call it, a 'go away' party, since we just had a going away party for me before the sprites campaign when no one expected me to come back afterward.

The sprites media frenzy has settled down but is not yet back to zero. I had over 70,000 views on my Flickr page in one day, which easily beats my previous record of 20,000. All in all I got around a quarter million views last week, and I'm still running several thousand per day. If I had charged a penny per view I'd be able to buy a D600 off it.

I imagine I'll have something interesting to post once I'm back on the move again. For now, here's a funnel web spider:

The Weaver

Wednesday, August 14, 2013

Sprites 2013: Final Update


We flew the final mission of the campaign on the night of 12-13 August. It was pretty much a bust: The storm which looked active died out right when we got off the ground. We only captured two sprites to the high speed cameras, neither of which was in the field of view of the dSLR. Geoff's camera, facing out the opposite side with a wider angle lens, caught three wimpy sprites, and that was it for dSLR shots - 3 sprites in 11,000 images. Not good at all. However, two of those three I actually managed to see with my naked eyes, so that's pretty cool. Here they are:



I now realize that I saw one the other night when I was seeing all the jets, and just didn't recognize it as such, so that's three I've seen by naked eye. Not bad.

I guess that wraps up the campaign. Pretty anticlimactic. I was hoping to use what I'd learned so far to get a shot that eclipsed all the previous ones, but if there are no sprites when we're up there I can't do much. Sprites 2013 is at an end.

Tuesday, August 13, 2013

Sprites 2013: Update #6


We flew again Sunday night / Monday morning, flight path above in green. The weather was marginal, but it went much better than expected. We got 9 or 10 sprites on the high speed, two of which were concurrent with the dSLR. This one - the best of the night - was on the final leg.


Most of what we saw were C-sprites, short for 'Column sprites' or 'Columnar sprites' - it just refers to their shape as tall, single columns. Here's a really good example of C-sprites:

Sprites, with high speed video

And here are the same sprites in the high speed:

And here's another set, grouped in a ring called a crown:

Sprites, with high speed video

With high speed video:

There's these strange groups of columns stuck together at the end like a bundle of sticks:


A less impressive / more noisy image:


And lastly, this sprite was pretty dim so I boosted the exposure pretty hard, and what popped out? 


See the (faint!) green striations across the lower half of the image? Airglow! Never expected to get that in a three-quarter second exposure. It's a chemiluminescence - emission of light by chemical reaction - effect common across the world, but very dim and unnoticed by virtually everyone. The primary color is the same 557nm green line from the aurora, which indicates we're looking at the emissions from excited atomic oxygen, up above 100km altitude.

For this flight Geoff brought his own camera (D5100) aboard on a freshly purchased GorillaPod, to mimic my setup from last time on the opposite side of the aircraft. So now we have coverage on the left (science) side with my D7000 and a 'normal' 35mm lens, and coverage on the right (non-science) side with the D5100 and a wider 24mm lens. On the previous flight I burned through my 64Gb memory card really surprisingly quickly, so we brought 3 new 64Gb cards aboard (thanks Ryan!), for a total of 2x64Gb + 1x32Gb memory per camera, and we switched to shooting JPG fine rather than RAW images. This was a mistake: Once the JPG compression crunches the dark noise, I can't subtract it back out again. Fortunately I recognized we had memory to burn on the last leg and I switched back to RAW for the final data run, which is when I captured the best sprites.

As I write this, we've actually just finished flying another mission, our last of this campaign. But since it's 5 in the morning and I have to meet in 6 hours to remove the cameras from the aircraft then drive 13 hours, I'm going to save that update for later - when I have time to go through the 11,000 pictures (seriously!) and look for good sprites. I will say that I finally saw a sprite naked-eye; two of them, both recorded on Geoff's camera.

Until then.

Sunday, August 11, 2013

Hey look


Meteors change color as they come down. I've never noticed this before. I looked up a few other pictures and they all seem to have the same color profile - green on the top changing to pink. The American Meteor Society says "The majority of light from a fireball radiates from a compact cloud of material immediately surrounding the meteoroid or closely trailing it. 95% of this cloud consists of atoms from the surrounding atmosphere" And I know meteors occur in the same altitude regime that auroras occur. I therefore postulate that the green top end is the molecular oxygen emission changing to the pink nitrogen emission at the 100km cutoff level. So I suppose now I can actually locate a meteor in Lat/Long just from a picture, provided the meteor was bright enough to show color.

All's quiet on the sprites front. Just waiting for the right storm.

13Aug2013 edit: Nope, that's not quite right...

I took this Sunday night from the Gulfstream:


It's a meteor in two consecutive frames with a one second cadence. And look, the part of the trail that was white in the first frame turns green in the second. I mean, duh. The green is not a prompt emission, it takes about a second after the oxygen atom is excited before it spits out a green photon. The meteor comes through and excites the oxygen (top frame), then a second later the oxygen de-excites and spits out some green (bottom frame). So looking at meteor images, the point where the trace turns from green to pinkish white is not the 100km level, it's just where the meteor was when the exposure began, and the green part is where the meteor was before the exposure began, just now emitting green. It's actually a pretty neat illustration of a forbidden transition, and I'll probably use it as an example in the future.

I should know these things.

Thursday, August 8, 2013

Sprites 2013: Press notes

No flights the last few nights due to unfavorable weather, but the internet (the cool parts of it anyway) is somewhat abuzz with the image and video I put up with the last post. This has taken off a bit more than I thought it would. Universe Today and were expected, but now more mainstream media is jumping on board. I'm going to use this post to archive some of the articles, beware that below the page break lies a bunch of copypaste from news sites. And if you know of one I'm missing, please comment or send it to me.

Tuesday, August 6, 2013

Sprites 2013: Update #5

We flew again tonight, but kind of got burned by the storm. There was a large number of positive charge-moment-change around 8:00 when we were planning, and just the lightning map in general looked like an unusually high number of positive strokes. Normally we see about 1 positive stroke out of every 10 or so, but tonight it seemed like it was around 50/50. And that was reflected among separate storms across the US, making you think there must be something larger controlling the fraction of positive strokes. It's interesting.

Anyway, we took off around 9:00pm, based on the fact that there was a very cold storm with lots of positive CMC, which is exactly what we look for. But it just died almost as soon as we were in the air. There was still a lot of lightning, but the big positive CMCs almost disappeared. I don't know what happened, and I actually don't think anyone knows enough about what sets the fraction of positive lightning strokes to really explain what happened. We had a long discussion about this over post-flight beer - someone should really figure this out.

But long story short, we only recorded two sprites to high speed tonight, neither of which was very good. One we only got the upper halo part of (which is pretty useless as spectra goes), and the other saturated the intensifier, so we can't use its spectrum either. We called the flight early, and now have about 10 flight hours left. Enough for two 5-hour flights, or, if our next flight is good, we'll just stay in the air as long as we can and call that the end.

On the plus side, I got another sprite in my dSLR. This whole time I've wanted to use a tripod to place my camera in a window and just let it shoot, but the NCAR folks are pretty strict about stuff like that. Well everyone was interested enough in my shots from last time that I was able to ride the good spirits into putting my camera into the window with a gorilla pod, using my bandanna as a dark cloth. Looks like this:


The first of the two sprites tonight was too far forward to catch in the Nikon, but I did manage to catch the second one:

Sprite Image

And, since we caught it on high speed as well, here's the same sprite at 10,000 frames per second, slowed down by about 666x:

So that's pretty cool. Since we still have 1 or maybe 2 flights left in this campaign, I'm still hoping to capture something that puts this to shame :P but if this is the best I get, it's still pretty damn good. And, the PI's are so impressed that they want me to put some real effort into this on the next campaign. It's a great public outreach opportunity, as demonstrated by the fact that my Flickr page got more than 10,000 views on the day I put the first pictures up. So if I don't get the shot to end all shots this campaign, I'll be well ready for it next time...

Edit: Oh yeah! In the sprite excitement I almost forgot we had some quite nice evening light while fueling the aircraft and waiting for dark:




Saturday, August 3, 2013

Sprites 2013: Update #4

Three jets

Last night we got out on the north edge of the storm and ran 'racetrack' paths back and forth. Scientifically, it wasn't a great night. The storm was nice and productive when we arrived on station, but the activity was spread out along a line, rather than focused in a hotspot, which makes pointing the cameras much more difficult - we basically wait to see sprites in the wide angle context cameras, then point in the spot they just happened. Well when the sprites are occurring all over the place that doesn't work so well. We only captured 7 sprites to high speed, and only one had a really usable spectra. After the first two racetracks the storm died off and we didn't really see anything for the next two, so we called it a night and went home.

Flight path in green.

However, I'm thrilled about last night for other reasons: I managed to capture both jets and sprites with my dSLR! I only know of a couple examples off the top of my head where people have captured sprites in SLRs, and I can't find any examples of jets captured with SLRs - am I the first? Since jets tend to hug the top of the clouds it's understandable that they're more difficult for a ground observer to see/photograph, so it makes sense that being up in a sprite-chasing aircraft would give me a serious advantage. The photo at the top of this post is, IMO, the cream of the crop, despite the strong motion blurring (the pilot would pick that moment to hit a bump...) It shows three jets shooting out of the cloud top just above some lightning, and it's even fairly well composed. Magic. Here is a good shot of a single jet - it looks like a blue butane lighter flame sticking out of the top of the cloud:

Blue jet

And here's the sprite I got, it's the red alien jellyfish on the top right:


Since I have a good starfield in both images, and I know we were about 200 kilometers from the electrically productive part of the storm, I can estimate the size of these things... but it's far easier to wait for the solver to give me a degrees/pixel solution than to figure it out myself, so we'll just wait for that.

For the technical aspect (skip this paragraph if you don't care about photography), I was shooting 1 second exposures, from a moving aircraft, handheld. I butted the camera up against the window glass and put my weight on it to get rid of most of the wobblies and light leaks, but the motion of the aircraft itself still showed up, especially when we hit a patch of turbulence  (we are, you know, flying right next to a thunderstorm). The city lights on the ground showed quite a bit more motion trailing than the stars so I cropped them out, but it was interesting to notice. For purposes of focus you think of both the city lights and the stars as being 'at infinity', but for purposes of parallax they really aren't. The instantaneous phenomena, the lightning, jets and sprites, show no motion blur. I used a 35mm f/1.8 lens held wide open, ran the camera at ISO 6400, subtracted a dark frame from all of them to get rid of the systematic noise, then ran strong noise reduction to help with the random noise. Some light dodging on the TLE's to make them a little more visible, but they are otherwise pretty much as-shot.

I was also able to see quite a few jets with my naked eyes! That's a first for me, and I'm always excited to see a new sky phenomenon for myself. I still haven't been able to see a sprite naked-eye, and it impresses me just how difficult that actually is. I'm flying in a private jet, right next to a thunderstorm, for the specific purpose of imaging sprites. I have very good low light eyesight, and I've watched tons of sprites in real time on the context cameras so I know exactly what and where to look. I was watching intently out the window while I snapped these shots, and the camera caught a sprite that I didn't see. Garggghhh! Since a typical sprite only lasts a couple of milliseconds, it's entirely possible it happened during a moment I looked away, or blinked. But still... Takeshi did manage to see one by eye last night, his first, so I'm not giving up yet.

Friday, August 2, 2013

Sprites 2013: Update #3


We flew again last night, the route shown above. More trouble with high cirrus clouds on the south and east sides of the storm, but eventually we came into clear air on the north side and were able to record 13 sprites in the high speed cameras. Fiddling about in the clouds for a couple of hours seems to be par for the course for this campaign. We used another 6 hours of flight time, leaving us around 18 more hours in the campaign. We're currently sitting at NCAR RAF and waiting to see how the weather will be tonight before making a decision to fly again: There currently isn't a lot of storm or electrical activity out there, much less the large positive charge-moment-changes we want to see, but the CAPE indicates huge instability out over the middle of Kansas, so we're hoping something will pop up.

I drove up through some of the canyons in the mountains today: It's really quite nice up there away from all the traffic and subdivisions. Because I have nothing else to add, here's a barn I saw today:

That's a fine lookin' barn you got there.

Update: As hoped for, a storm has formed in the middle of Kansas. It's very cold in the IR satellite imagery and the cirrus clouds are pretty confined this time. Pretty much exactly what we're looking for. Planning to take off at 10:30.