http://www.nasa.gov/image-feature/africa-a...lion-miles-away
Full Version: DSCOVR
We do live on a beautiful world. I expect Aliens will come to visit just for the aesthetics.
This is so very cool! I have been waiting to see this; didn't realize they would capture it so soon.
We live in a very beautiful system!
http://www.nasa.gov/feature/goddard/from-a...g-face-of-earth
We live in a very beautiful system!
http://www.nasa.gov/feature/goddard/from-a...g-face-of-earth
QUOTE (Ron Hobbs @ Aug 5 2015, 07:49 AM) 
I wonder what the source of the fringe around the Moon is? Color channel misregistration?
Time lag between different color images--doesn't matter as earth rotates little--bigger problem for moon flying past.
My attempt to correct for the motion of the Moon, in addition to whatever processing caused blue filter signal to leak into the red and green channels:
http://postimg.org/image/6j1xbyua7/full/
http://postimg.org/image/6j1xbyua7/full/
Wow! I was actually wondering if being at L1 would allow transits to happened regularly. Apparently this is twice a year (plus the Moon being occulted two weeks before/after).
QUOTE (Explorer1 @ Aug 5 2015, 01:24 PM)
Wow! I was actually wondering if being at L1 would allow transits to happened regularly. Apparently this is twice a year (plus the Moon being occulted two weeks before/after).
Fascinating... This image does give a layperson such as myself an idea of the brightness/albedo of the moon compared to earth. It is darker than I expected!
Good find on the transit. Neat that we are seeing the far side of the moon. I wonder how the lunar occultation will look in late September, when we are also having a total lunar eclipse? Will we see any of the penumbra (or even umbra) on the moon? This might depend on the fact that DSCOVR varies a bit from being exactly along the Sun - Earth line.
I liked this headline.
Click to view attachment
http://www.bloomberg.com/news/articles/201...t-of-the-camera
Click to view attachment
http://www.bloomberg.com/news/articles/201...t-of-the-camera
I found a reference that DSCOVR's L1 orbit varies the Sun-Earth-DSCOVR angle by 4-15*. I assume like Soho the ellipse is mostly East/West. September's eclipse is a partial in the southern hemisphere. I expect it will be difficult to notice since it will be over mostly dark water, far from the subsolar point. The March 8th total solar eclipse next year should be very noticeably, since it will cross the central Pacific. Since the umbra will stay on Earth's surface for over 3 hours, I think we should get a couple of pics of it.
http://www.timeanddate.com/eclipse/list.html
http://www.timeanddate.com/eclipse/list.html
The difference between Earth-Moon albedo can also be seen in person when a gibbous Moon rises in the afternoon or sets in the morning and can be compared to distant hills, clouds, etc. on Earth, although there are many factors that can ruin the comparison.
The gibbous moon is an interesting comparison. The land surface of the Earth can be relatively close to the lunar albedo, compared with the clouds. When looking at specific locations relative to the sun, we can consider the reflectance since the albedo is more of an averaged quantity. At some point I may test my rendering software to show the DSCOVR view with the moon included.
No news on the proposed daily 'pipeline'? There haven't been any images released in September at all (not trying to sounding ungrateful, just wondering about what the holdup is!)
Just noticed the DSCOVR website had a news update yesterday. More or less it talks about the camera on board and mentions that a new website was created to host the available data.
http://epic.gsfc.nasa.gov/
Doesn't seem to be able to finish loading for me at this moment in time however.
http://epic.gsfc.nasa.gov/
Doesn't seem to be able to finish loading for me at this moment in time however.
It took me three tries to load the website yesterday, but when it finally loaded I found that I could pull up whole-Earth images from two days prior -- which happened to be my 60th birthday. It lets you scroll through images taken a few hours apart throughout a given day.
I was able to capture the image of the western hemisphere, featuring my home continent, as it appeared in the middle of my birthday. Kewl! I now have it as my desktop.
-the other Doug
I was able to capture the image of the western hemisphere, featuring my home continent, as it appeared in the middle of my birthday. Kewl! I now have it as my desktop.
-the other Doug
Yeah, it finally loaded for me as well. Probably took 10m to load the page. Either they are getting pounded with site hits or they dug up a clunker from the basement to host the site. 
At least! Very nice, and apparently archives going back a few months. Nice to see things like the distance information and angle from the sun too.
Eventually the Moon will pass behind/in front of our blue marble again....
Eventually the Moon will pass behind/in front of our blue marble again....
The timing of the pictures seems really random, and the Moon transit sequence isn't included. Are we only going to get handfuls of images for the time being?
Does the moon always transit as seen from DSCOVR? There's enough deviation of DSCOVR from the Sun - Earth line, and the moon's orbital inclination effect that could cause a miss.
The timing of the pictures seems to be every 2 hours or a bit less. Maybe some days they aren't downlinking as many.
The website has an email contact that may help with some of our questions.
The timing of the pictures seems to be every 2 hours or a bit less. Maybe some days they aren't downlinking as many.
The website has an email contact that may help with some of our questions.
I think they mentioned in the first press release on the lunar transit that it doesn't happened very often (think of eclipse frequencies). All three (Earth, Moon, DSCOVR) have to be on the same plane, so it's about twice a year.
None of the images show the Moon setting behind the Earth either, however; it might be a matter of timing the image to get the Moon in frame before and after it passes behind Earth's disc....
None of the images show the Moon setting behind the Earth either, however; it might be a matter of timing the image to get the Moon in frame before and after it passes behind Earth's disc....
I don't have direct answers to the questions being asked here, but I do have a little information that might be useful to people. When they first launched the website, I sent this inquiry to the media contact:
My inquiry was forwarded to Adam Szabo, who replied:
QUOTE
I'm enjoying browsing the new DSCOVR EPIC website; it's a beautiful camera. I am wondering if there is a place where I can view a table or text file or something that contains basic metadata for the images (primarily a list of links to images with their timestamps). The reason I ask is because I'm interested in trying to make animations of Earth over time from a fixed longitude, but the way the website is set up, it's laborious to identify which days contain images taken at roughly the same times. Such a table would also allow people to invent different ways of displaying/searching the data.
My inquiry was forwarded to Adam Szabo, who replied:
QUOTE
EPIC Daily "Blue Marble" API
The API URL is:
http://epic.gsfc.nasa.gov/api/images.php
This gets you a list of the latest day's images & metadata.
OR
http://epic.gsfc.nasa.gov/api/images.php?d...amp;w=X&e=Y
The second form's optional parameters allow you to focus in to geographical regions that were in view (technically, longitudinal bounding points) on a given date.
For instance: Using North America's boundaries of east = -53.034, and west = -170.859 and choosing August 24, 2015, the URL would look like this:
http://epic.gsfc.nasa.gov/api/images.php?d...9&e=-53.034
The date parameter is unpadded(e.g. 2015-9-1, vs. 2015-09-01 for September 1, 2015), and optional. Leaving the date off will default to the latest image set.
Leaving out the coordinates gets you every image for whatever date is returned.
Adding either longitudinal parameter returns all images for the given date, for which the supplied longitude is in view of the camera.
The JSON data looks like this:
ImageData object:
{
"image": "epic_1b_20150826231708_00", // image name sans extension.
"caption": "About an image", // Will contain a caption
"coords": "{}", // Contains a JSON string representing a Coordinates object.
"date": "2015-08-26 23:17:08" // The date the capture sequence for the image was initiated
}
ImageData notes:
For the image field to be useful, it has to be added to a URL. Image URLs differ depending upon purpose.
Given an image name of "epic_1b_20150901205648_00" (taken Sept 1, 2015, with the camera aimed
at 3 typhoons in the pacific (my favorite)), we have:
For compressed JPG thumbnails, the path would be
http://epic.gsfc.nasa.gov/epic-archive/thu...01205648_00.jpg
For full-size, compressed JPG previews, the path would be
http://epic.gsfc.nasa.gov/epic-archive/jpg...01205648_00.jpg
For full-size original PNG images, the path would
http://epic.gsfc.nasa.gov/epic-archive/png...01205648_00.png
Coordinates object
{
"centroid_coordinates": { // Geographical coordinates that the satellite is looking at
"lat": 4.076132,
"lon": -169.648562
},
"dscovr_j2000_position": { // Position of the satellite in space
"x": -1439710.750000,
"y": 659227.437500,
"z": 113316.414062
},
"lunar_j2000_position": { // Position of the moon in space
"x": 153199.062500,
"y": -319797.531250,
"z": -104905.093750
},
"sun_j2000_position": { // Position of the sun in space
"x": -134918656.000000,
"y": 62555808.000000,
"z": 27119770.000000
},
"attitude_quaternions": { // Satellite attitude
"q0": 0.976360,
"q1": -0.212080,
"q2": -0.039540,
"q3": 0.013750
}
}
The API URL is:
http://epic.gsfc.nasa.gov/api/images.php
This gets you a list of the latest day's images & metadata.
OR
http://epic.gsfc.nasa.gov/api/images.php?d...amp;w=X&e=Y
The second form's optional parameters allow you to focus in to geographical regions that were in view (technically, longitudinal bounding points) on a given date.
For instance: Using North America's boundaries of east = -53.034, and west = -170.859 and choosing August 24, 2015, the URL would look like this:
http://epic.gsfc.nasa.gov/api/images.php?d...9&e=-53.034
The date parameter is unpadded(e.g. 2015-9-1, vs. 2015-09-01 for September 1, 2015), and optional. Leaving the date off will default to the latest image set.
Leaving out the coordinates gets you every image for whatever date is returned.
Adding either longitudinal parameter returns all images for the given date, for which the supplied longitude is in view of the camera.
The JSON data looks like this:
ImageData object:
{
"image": "epic_1b_20150826231708_00", // image name sans extension.
"caption": "About an image", // Will contain a caption
"coords": "{}", // Contains a JSON string representing a Coordinates object.
"date": "2015-08-26 23:17:08" // The date the capture sequence for the image was initiated
}
ImageData notes:
For the image field to be useful, it has to be added to a URL. Image URLs differ depending upon purpose.
Given an image name of "epic_1b_20150901205648_00" (taken Sept 1, 2015, with the camera aimed
at 3 typhoons in the pacific (my favorite)), we have:
For compressed JPG thumbnails, the path would be
http://epic.gsfc.nasa.gov/epic-archive/thu...01205648_00.jpg
For full-size, compressed JPG previews, the path would be
http://epic.gsfc.nasa.gov/epic-archive/jpg...01205648_00.jpg
For full-size original PNG images, the path would
http://epic.gsfc.nasa.gov/epic-archive/png...01205648_00.png
Coordinates object
{
"centroid_coordinates": { // Geographical coordinates that the satellite is looking at
"lat": 4.076132,
"lon": -169.648562
},
"dscovr_j2000_position": { // Position of the satellite in space
"x": -1439710.750000,
"y": 659227.437500,
"z": 113316.414062
},
"lunar_j2000_position": { // Position of the moon in space
"x": 153199.062500,
"y": -319797.531250,
"z": -104905.093750
},
"sun_j2000_position": { // Position of the sun in space
"x": -134918656.000000,
"y": 62555808.000000,
"z": 27119770.000000
},
"attitude_quaternions": { // Satellite attitude
"q0": 0.976360,
"q1": -0.212080,
"q2": -0.039540,
"q3": 0.013750
}
}
The seasons are changing; yesterdays image (November 27th at 12:00 GMT) is giving a sense of deja vu....
http://epic.gsfc.nasa.gov/epic-archive/png...26081200_00.png
A little over a week until the anniversary of the Blue Marble that started it all.
http://epic.gsfc.nasa.gov/epic-archive/png...26081200_00.png
A little over a week until the anniversary of the Blue Marble that started it all.
Apparently the lunar eclipse images were saved for AGU 2015, so they weren't on the website (makes sense; the craft had to track the Moon, not Earth, so it would have been a confusing sequence).
http://www.bbc.com/news/science-environment-35076402
http://www.bbc.com/news/science-environment-35076402
Yes, this was shown yesterday here at AGU in a panel discussion including Al Gore, Adam Szabo and several additional science team members. It was suggested by the panel that the images from DSCOVR would look much nicer with some motion interpolation applied to make a continuous movie. Sounds right up UMSF's alley 
Also being discussed are followup missions with improved cameras.
Also being discussed are followup missions with improved cameras.
I'd love to try some motion interpolation with these... However, I can't seem to find the actual images from the eclipse anywhere - the EPIC site still shows no images for 2015-09-27. Has anyone else been able to locate them?
I haven't, either. I just sent an email and also asked about PDS release plans.
That was quick!
QUOTE
The July 16, lunar transit mages are available at the http://epic.gsfc.nasa.gov web site.
At the top level menu, select "Galleries" and "Lunar Transit". All 20 full resolution images are there
for downloading.
The September 27 images will appear shortly at the same location.
We are serving all of the EPIC RGB images through the http://epic.gsfc.nasa.gov site and will continue to update as new
images become available. The single wavelength, Level 1b images will be made publicly available through the Langley
Atmospheric Science Data Center (https://eosweb.larc.nasa.gov) our official archival site, starting in late February.
Adam Szabo
NASA DSCOVR Project Scientist
At the top level menu, select "Galleries" and "Lunar Transit". All 20 full resolution images are there
for downloading.
The September 27 images will appear shortly at the same location.
We are serving all of the EPIC RGB images through the http://epic.gsfc.nasa.gov site and will continue to update as new
images become available. The single wavelength, Level 1b images will be made publicly available through the Langley
Atmospheric Science Data Center (https://eosweb.larc.nasa.gov) our official archival site, starting in late February.
Adam Szabo
NASA DSCOVR Project Scientist
I'm curious whether there is any possibility of improving the downlink capability. I understand that the red and blue channels are being reduced to 1K x 1K pixels prior to transmission to Earth. The net result is that the images we see are the equivalent of a Bayer filtered color image, even though the original images onboard are 2K x 2K RGB.
I've been doing some renderings of the Earth from increasing altitudes (up to 40000km so far), and comparing with DSCOVR imagery. I'm starting to think the DSCOVR images (have insufficient gamma correction or are too contrasty) as the ratio of green counts between bright clouds and ocean is about 6. This is between the linear and gamma corrected values. Same with blue with about a 3 or 4 ratio in the DSCOVR images. Thus the images are too contrasty between clouds & oceans compared with reality. Here is a DSCOVR image I'll elaborate on in the next post.
Click to view attachment
Click to view attachment
Next we can look at the modified DSCOVR image, though only with a 0.70 gamma adjustment instead of the 0.45 I thought would be needed to generate fairly realistic results:
Click to view attachment
This can be compared with a rendering I did using 3-D cloud hydrometeor gridded data, surface albedo derived from NASA Blue Marble, and other atmospheric assumptions. The time and geometry should be pretty close. I'm still refining things like the sun glint, cloud phase function for single & multiple scattering, and atmosphere near the limb.
Click to view attachment
Click to view attachment
This can be compared with a rendering I did using 3-D cloud hydrometeor gridded data, surface albedo derived from NASA Blue Marble, and other atmospheric assumptions. The time and geometry should be pretty close. I'm still refining things like the sun glint, cloud phase function for single & multiple scattering, and atmosphere near the limb.
Click to view attachment
Using the automated server containing the json file of all images (here: http://epic.gsfc.nasa.gov/api/images.php), it appears that the images containing the lunar transits are not present there. Considering Emily's post, I'm not sure what it does mean. Does the folder http://epic.gsfc.nasa.gov/epic-archive/png/XXX.png contain only the non-lunar transit images ? Only featured images ? I am a bit confused, I could download more that 1800 EPIC images automatically except the most dramatic ones...
In the meantime I struggle a bit with the quaternion of the spacecract. According to it, the image should be oriented like this:
Click to view attachment
But when forcing the "up" of the image to "North" in earth frame, the image becomes inconsistant with the given quaternion but oriented as expected:
Click to view attachment
Any idea if the images were post-processed to make them point north ?
In the meantime I struggle a bit with the quaternion of the spacecract. According to it, the image should be oriented like this:
Click to view attachment
But when forcing the "up" of the image to "North" in earth frame, the image becomes inconsistant with the given quaternion but oriented as expected:
Click to view attachment
Any idea if the images were post-processed to make them point north ?
When I click on the first link above, I can see the 'json' file for just the most recent day's images. Generally the images I've seen are all pointing north. I wonder if they are also being centered in the frame?
No, they are not centred and the quaternion clearly lacks of precision : the centre of the earth according to the quaternion is usually several dozens of pixels off-centre from the actual centre of the earth.
To get the json files for other dates, you just need to add the date in your query, e.g. http://epic.gsfc.nasa.gov/api/images.php?date=2016-1-12
My conclusion is that the provided quaternions are almost useless, I needed to recalculate my own.
To get the json files for other dates, you just need to add the date in your query, e.g. http://epic.gsfc.nasa.gov/api/images.php?date=2016-1-12
My conclusion is that the provided quaternions are almost useless, I needed to recalculate my own.
Thanks for the updates. I can note that the longitude range as per this example simply shows if the box is in view. Is it possible to extract for when the centroid longitude lies within a certain range?
http://epic.gsfc.nasa.gov/api/images.php?d...9&e=-53.034
http://epic.gsfc.nasa.gov/api/images.php?d...9&e=-53.034
Have you tried contacting the EPIC team about what errors you are seeing in the quaternions?
No. I could recreate my own with a precision better than 1 pixel, so I don't need to bother them.
QUOTE (Stratespace @ Feb 28 2016, 02:31 PM)
No. I could recreate my own with a precision better than 1 pixel, so I don't need to bother them.
If you're going to say that their product is "almost useless" it only seems fair to give them an opportunity to fix it or at least explain what the issue is.
It's almost useless to me, I don't want to make them think that what they propose is "bad" or anything like this. The rest of the metadata is pretty good and useful to me.
As long as I could find absolutely no documentation on the quaternion (frame definition, where is the scalar and the vector, etc) I had to guess everything by my own. I finally found what I thought is correct, then I deduced that I was not useful to me... I am very far to question the intrinsic value of the quaternion they propose, maybe for example it is not supposed to represent directly the orientation of the image.
Anyway, I don't feel the "right" nor the confidence to contact people that have other things to do that being bothered by someone from the other side of the planet, who speaks poor english, who just plays a bit with the data they are very kind to propose and who maybe doesn't understand what is proposed to him due to lack of information. If you feel well doing it, please feel free to do it.
As long as I could find absolutely no documentation on the quaternion (frame definition, where is the scalar and the vector, etc) I had to guess everything by my own. I finally found what I thought is correct, then I deduced that I was not useful to me... I am very far to question the intrinsic value of the quaternion they propose, maybe for example it is not supposed to represent directly the orientation of the image.
Anyway, I don't feel the "right" nor the confidence to contact people that have other things to do that being bothered by someone from the other side of the planet, who speaks poor english, who just plays a bit with the data they are very kind to propose and who maybe doesn't understand what is proposed to him due to lack of information. If you feel well doing it, please feel free to do it.
Stratespace, I think that the EPIC team might actually appreciate the inquiry. I don't think that you would offend them (and it's obvious that's not your intent at all), and feedback from users of their products would doubtless be valuable.
I checked the website and it looks like they've shifted the date for the availability of the archival data from "end of February" to "spring 2016". There is a mailing list you can be signed up for to be notified of project status.
Yes I think this leap day qualifies as the end of February
Generally I think it's difficult for spacecraft attitude information to be accurate enough to navigate image position to the pixel level. I also end up simply fitting empirically (and somewhat automatically) to the limb within the image. Might be nice though for them to know about our discussions in general.
QUOTE (nprev @ Feb 29 2016, 02:14 PM)
Stratespace, I think that the EPIC team might actually appreciate the inquiry. I don't think that you would offend them (and it's obvious that's not your intent at all), and feedback from users of their products would doubtless be valuable.
Generally I think it's difficult for spacecraft attitude information to be accurate enough to navigate image position to the pixel level. I also end up simply fitting empirically (and somewhat automatically) to the limb within the image. Might be nice though for them to know about our discussions in general.
For the first images in particular, the quaternion attitude error is much more than 1 pixel, even sometime more than 1 Field of View !
My observations are:
- The "Earth north is toward image top" is roughly true, but not exactly. There is always are ~1° to 5° typically (with peaks at more than 15°) error between real north and image "up".
- The latest the images, the best their orientation wrt north.
- Some images a completely black, for some reason.
- There are long missing periods in the data that we know exist.
- There is uncorrected or residual image distortion. I could struggle estimate the polynomials, but it would be a lot of effort for a faint effect (~1 to 2 pixels).
I could re-calculate the quaternion according to what is visible in the images, update the image "up" toward a more accurate "north", and when corrected everything seems much prettier. I'll try to post a video showing this result.
My observations are:
- The "Earth north is toward image top" is roughly true, but not exactly. There is always are ~1° to 5° typically (with peaks at more than 15°) error between real north and image "up".
- The latest the images, the best their orientation wrt north.
- Some images a completely black, for some reason.
- There are long missing periods in the data that we know exist.
- There is uncorrected or residual image distortion. I could struggle estimate the polynomials, but it would be a lot of effort for a faint effect (~1 to 2 pixels).
I could re-calculate the quaternion according to what is visible in the images, update the image "up" toward a more accurate "north", and when corrected everything seems much prettier. I'll try to post a video showing this result.
Here we are:
https://www.youtube.com/watch?v=V5YZnXZs_v4
This is not intended to be beautiful nor pleasant to see, but just to provide an idea of what is possible in terms of projection after the metadata has been a bit "improved".
https://www.youtube.com/watch?v=V5YZnXZs_v4
This is not intended to be beautiful nor pleasant to see, but just to provide an idea of what is possible in terms of projection after the metadata has been a bit "improved".
Wacky. Sort of like passing by time on a subway train. But a cool start to something; I'm looking forward to where you go with this.
Yes, very cool! The gradual season change is so clearly evident, as are individual storm systems. Keep it up!
Ok, I've just put a bit of persistence on the map, so you won't need no aspirin when watching this.
But there is so much work yet to do to calibrate both the extrinsic and intrinsic parameters of each image, it is a bit discouraging... particularly considering all the processing already applied on the image. I find it much easier to play with classical raw PDS images !
https://youtu.be/7N63ucC1lXY
I'll need to find those parameters, and then tune some image morphing algorithms. I've already tested some of them on those images, and they provide poor result.
But there is so much work yet to do to calibrate both the extrinsic and intrinsic parameters of each image, it is a bit discouraging... particularly considering all the processing already applied on the image. I find it much easier to play with classical raw PDS images !
https://youtu.be/7N63ucC1lXY
I'll need to find those parameters, and then tune some image morphing algorithms. I've already tested some of them on those images, and they provide poor result.
As the morphing software didn't give me what I wanted, I designed an algorithm (based on some publications) dedicated to image temporal interpolation for planetary images (understand: it can deal with clouds warping, tearing and whirling).
The next video is based on only 2 images, processed in a fully automatic way (not a single command nor help brought to the algorithm, except "run"):
- Image 1: http://epic.gsfc.nasa.gov/epic-archive/png...07130002_00.png
- Image 2: http://epic.gsfc.nasa.gov/epic-archive/png...08130501_00.png
Result is this video (Watch in HD): https://youtu.be/M2CiR4WCaPo
Please tell me if you find it satisfactory... or not.
The next video is based on only 2 images, processed in a fully automatic way (not a single command nor help brought to the algorithm, except "run"):
- Image 1: http://epic.gsfc.nasa.gov/epic-archive/png...07130002_00.png
- Image 2: http://epic.gsfc.nasa.gov/epic-archive/png...08130501_00.png
Result is this video (Watch in HD): https://youtu.be/M2CiR4WCaPo
Please tell me if you find it satisfactory... or not.
Interesting to see, though a day is a long time to interpolate clouds for. I wonder how the Butterflow algorithm that Dan Delany has been showing us would do with this? This can somewhat be validated by comparing the interpolated DSCOVR frames with actual DSCOVR or Himawari frames that were remapped. Will be interesting to see how the time interpolation in post #98 can be applied to the movie shown earlier, including between frames just an hour or two apart.
BTW, a bit of a gap in real-time DSCOVR images, so we'll see when they can start to get some updates.
BTW, a bit of a gap in real-time DSCOVR images, so we'll see when they can start to get some updates.
Yes, obviously the objective is to run it on the map with all frames. But the 1-day gap seemed to be a good benchmark as a worst case scenario.
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.