5 ms·

NASA Maps Surface Changes from California Quakes

89 points by infodocket 7y ago

breck 7y ago
> Each color cycle represents 4.8 inches (12 centimeters) of ground displacement in the radar line-of-sight. Is there a legend somewhere, or at least a max? I'm not sure what is meant by "color cycle".
- mehrdadn 7y ago
  I think they just mean how the maps cycles through the colors. The contours for the same color are apart by multiples of a cycle.
- jofer 7y ago
  You're looking at an interferogram. The method used here can't measure displacement directly. It measures a phase difference between two radar images. Turning this into a map of displacement is non unique in the presence of noise and limited spatial resolution. That's why you'll see this type of data displayed in this way. The rainbow palette is mostly convention, but either way, it's the most direct view of what actually measured. It's also kinda pretty, i.m.o...
  - mirimir 7y ago
    It is certainly pretty. But I'm still not clear what it means. OK, so the radar is measuring distance between the satellite and reflecting surface. They're compairing data from July 8, 2019 and April 8, 2018. I'm guessing that the two images look pretty much the same. Especially given limited spatial resolution. But ELI5, what does the interference pattern show? I mean, are there 12 cm amplitude waves of vertical surface displacement? Something like frozen S waves?
    - jofer 7y ago
      An earthquake is rocks sliding past each other. What you're looking at here is a measure of how much they moved. (It doesn't "slide back" afterwards -- the motion is permanent.) For an earthquake of this size, the motion will be on the order of a few meters. ------------ In a bit more detail, the radar can't measure distance precisely enough to detect the movement. The distance measured before and after by radar is the same, within error. However, there's another part of the radar signal beyond just how long it takes to travel. That second part is the phase of the returned signal. Imagine the first time we imaged a small area, we got a return waveform that looked like this (zero phase): /\ /\ /\ \/ \/ \/ but then the next time we got back a slightly different result (270 degree phase): \ /\ /\ / \/ \/ \/ The difference is shape of the returning signal is a phase shift. The radar wave is shifted slightly We know that it moved at least three quarters wavelength in the ascii art example above. However, we'd get the same result if it moved ten and three quarters, though. We can measure part of the change very precisely, but the bulk of the motion looks the same to us. We're looking at that fined-grained part of the motion (phase difference) not the overall motion itself. In programming terms, we're looking at the result of a modulo operator.
      - mirimir 7y ago
        Thanks. I see that the ALOS-2 SAR uses L band, which seems to mean 1-2 GHz (30-15 cm). So maybe the ALOS-2 actually uses 12 cm? So does that mean that the pattern shows something like contour lines?
        jofer 7y ago
        Yep! You can think of the bands as contour lines of deformation.
        mirimir 7y ago
        Hey, thanks.
      - DuskStar 7y ago
        > we'd get the same result if it moved ten and three quarters And also the same if it moved one quarter wavelength in the opposite direction.
- antoineMoPa 7y ago
  If I understand the right thing, from one red "line" to another, there has been a 4.8 inch displacement between the two photos.
  - cardiffspaceman 7y ago
    Since the description says each color cycle represents 4.8 inches, I would say if you trace a line from one red pixel across the blue pixels to another red pixel, you've traversed a cycle, and you've increased or decreased displacement from the reference by 4.8 inches.
SubiculumCode 7y ago
The aftershocks going on here (most small, but higher frequency than usual) is interesting to look at, if not a little disturbing :). A 4.0 in the last hour too (6:48 PM). http://scedc.caltech.edu/recent/Maps/118-36.html http://scedc.caltech.edu/recent/Maps/118-36.html
golem14 7y ago
At first glance, this looks very much like some Julia sets.
blululu 7y ago
That is a very cool finding. Though I must say that visualizing the data with a jet colormap and no legend is a little disrespectful to anyone who is sober.
- nemetroid 7y ago
  This was my initial reaction as well, but the article mentions that > Each color cycle represents 4.8 inches (12 centimeters) of ground displacement either toward or away from the satellite. ...so what we're actually interested in is counting the number of cycles through the color map between two points, for which jet actually seems like a perfectly good choice. jofer explains in their comment why the data is shown this way: https://news.ycombinator.com/item?id=20398703 https://news.ycombinator.com/item?id=20398703
- enriquto 7y ago
  > a jet colormap and no legend this is very standard for interferometry (that represents a height difference modulo some step). There is no useful "legend" that you can associate to this colormap. The only information missing is the step between cycles of the colormap, which is given.
mturmon 7y ago
Relatedly, InSAR images of the Kilauea volcano area: https://www.jpl.nasa.gov/spaceimages/details.php?id=PIA13910 https://www.jpl.nasa.gov/spaceimages/details.php?id=PIA13910 And land subsidence due to groundwater pumping in California’s Central Valley, showing motion of up to 70 cm (!): https://www.jpl.nasa.gov/spaceimages/details.php?id=PIA16293 https://www.jpl.nasa.gov/spaceimages/details.php?id=PIA16293

89 points by infodocket 7y ago

> Each color cycle represents 4.8 inches (12 centimeters) of ground displacement in the radar line-of-sight. Is there a legend somewhere, or at least a max? I'm not sure what is meant by "color cycle".

mehrdadn 7y ago
I think they just mean how the maps cycles through the colors. The contours for the same color are apart by multiples of a cycle.
jofer 7y ago
You're looking at an interferogram. The method used here can't measure displacement directly. It measures a phase difference between two radar images. Turning this into a map of displacement is non unique in the presence of noise and limited spatial resolution. That's why you'll see this type of data displayed in this way. The rainbow palette is mostly convention, but either way, it's the most direct view of what actually measured. It's also kinda pretty, i.m.o...
- mirimir 7y ago
  It is certainly pretty. But I'm still not clear what it means. OK, so the radar is measuring distance between the satellite and reflecting surface. They're compairing data from July 8, 2019 and April 8, 2018. I'm guessing that the two images look pretty much the same. Especially given limited spatial resolution. But ELI5, what does the interference pattern show? I mean, are there 12 cm amplitude waves of vertical surface displacement? Something like frozen S waves?
  - jofer 7y ago
    An earthquake is rocks sliding past each other. What you're looking at here is a measure of how much they moved. (It doesn't "slide back" afterwards -- the motion is permanent.) For an earthquake of this size, the motion will be on the order of a few meters. ------------ In a bit more detail, the radar can't measure distance precisely enough to detect the movement. The distance measured before and after by radar is the same, within error. However, there's another part of the radar signal beyond just how long it takes to travel. That second part is the phase of the returned signal. Imagine the first time we imaged a small area, we got a return waveform that looked like this (zero phase): /\ /\ /\ \/ \/ \/ but then the next time we got back a slightly different result (270 degree phase): \ /\ /\ / \/ \/ \/ The difference is shape of the returning signal is a phase shift. The radar wave is shifted slightly We know that it moved at least three quarters wavelength in the ascii art example above. However, we'd get the same result if it moved ten and three quarters, though. We can measure part of the change very precisely, but the bulk of the motion looks the same to us. We're looking at that fined-grained part of the motion (phase difference) not the overall motion itself. In programming terms, we're looking at the result of a modulo operator.
    - mirimir 7y ago
      Thanks. I see that the ALOS-2 SAR uses L band, which seems to mean 1-2 GHz (30-15 cm). So maybe the ALOS-2 actually uses 12 cm? So does that mean that the pattern shows something like contour lines?
      - jofer 7y ago
        Yep! You can think of the bands as contour lines of deformation.
        mirimir 7y ago
        Hey, thanks.
    - DuskStar 7y ago
      > we'd get the same result if it moved ten and three quarters And also the same if it moved one quarter wavelength in the opposite direction.
antoineMoPa 7y ago
If I understand the right thing, from one red "line" to another, there has been a 4.8 inch displacement between the two photos.
- cardiffspaceman 7y ago
  Since the description says each color cycle represents 4.8 inches, I would say if you trace a line from one red pixel across the blue pixels to another red pixel, you've traversed a cycle, and you've increased or decreased displacement from the reference by 4.8 inches.

SubiculumCode 7y ago

The aftershocks going on here (most small, but higher frequency than usual) is interesting to look at, if not a little disturbing :). A 4.0 in the last hour too (6:48 PM). http://scedc.caltech.edu/recent/Maps/118-36.html http://scedc.caltech.edu/recent/Maps/118-36.html

golem14 7y ago

At first glance, this looks very much like some Julia sets.

blululu 7y ago

That is a very cool finding. Though I must say that visualizing the data with a jet colormap and no legend is a little disrespectful to anyone who is sober.

nemetroid 7y ago
This was my initial reaction as well, but the article mentions that > Each color cycle represents 4.8 inches (12 centimeters) of ground displacement either toward or away from the satellite. ...so what we're actually interested in is counting the number of cycles through the color map between two points, for which jet actually seems like a perfectly good choice. jofer explains in their comment why the data is shown this way: https://news.ycombinator.com/item?id=20398703 https://news.ycombinator.com/item?id=20398703
enriquto 7y ago
> a jet colormap and no legend this is very standard for interferometry (that represents a height difference modulo some step). There is no useful "legend" that you can associate to this colormap. The only information missing is the step between cycles of the colormap, which is given.

mturmon 7y ago

Relatedly, InSAR images of the Kilauea volcano area: https://www.jpl.nasa.gov/spaceimages/details.php?id=PIA13910 https://www.jpl.nasa.gov/spaceimages/details.php?id=PIA13910 And land subsidence due to groundwater pumping in California’s Central Valley, showing motion of up to 70 cm (!): https://www.jpl.nasa.gov/spaceimages/details.php?id=PIA16293 https://www.jpl.nasa.gov/spaceimages/details.php?id=PIA16293