Fleets of radar satellites are measuring movements on Earth like never before | Hacker News Reader

Fleets of radar satellites are measuring movements on Earth like never before

sciencemag.org

90 pointssohkamyung30 comments

Can you get this sort of data publicly?

I've been watching the Mina, NV earthquake storm(https://earthquake.usgs.gov/earthquakes/map/?extent=37.25766...) over the last year and I'm curious how the ground is moving in response to those quakes.

JPL releases some of their products of earthquakes/disasters world wide here: https://aria-share.jpl.nasa.gov/

Also a group a Leeds is trying to make a public portal: https://comet.nerc.ac.uk/COMET-LiCS-portal/

Neither place seems to have the recent Nevada earthquakes though.

There's a bunch of effort to make InSAR more available/accessible. Even though plenty of SAR data is available through ASF (https://search.asf.alaska.edu/#/), it takes a bunch more processing beyond the SAR images to get useful ground deformation.

ASF has also started to try providing some interferograms (single InSAR images) but without averaging a bunch over an area, or doing more advanced stuff, they often look like mostly noise on a given day.

The following can be used to process InSAR data: https://github.com/isce-framework/isce2

An update is in progress: https://github.com/isce-framework/isce3

ESA freely distributes all data acquired by the Sentinel program from here: https://scihub.copernicus.eu/ There are also multiple partner sites with the data, including cloud providers that can you a VM with the data files already mounted.

Software is a bit harder, for simple interferograms of two measurements you can just use open source Sentinel-1 toolbox ( https://github.com/senbox-org/s1tbx ) to align, diff, subtract geometric corrections based on an elevation model and convert the measurements to geographic coordinates. This will be enough for seeing large scale motion, like earthquakes. For better quality processing you need multi-temporal InSAR which can estimate atmospheric phase delay, height of persistent reflectors which will cause a geometric phase difference due to slight differences in measurement location (measurement baseline). There is some open source software available for this (e.g. StaMPS), but I can't vouch for its capabilities.

I would love to hear some perspective on this, but from my recollection one of the most significant early datasets ever released was about 20 years ago when I was in grad school -- I heard about this "Shuttle Radar Topography Mission" that was going to scan the entire Earth.

I remember it had a long extendible gantry (?) that was deployed out of the shuttle bay to give it more scanning baseline. I think it was done for the intelligence community (this being just post 9-11) but some part of it became public? I remember reading that the data was collected on tapes (!) that they had to swap hot to ensure continuity of the data.

Shortly after that, I noticed that Earth simulation flyby animations became almost common, planes started being equipped with simulated terrain maps for collision warning, etc. I imagine it was a revolution for a lot of industries.

The inspector-gadget arm sticking out of the space shuttle does look pretty funny https://www.usgs.gov/centers/eros/science/usgs-eros-archive-...

I don't think the arm was related to anything with intelligence- they just needed some way to have a second radar slightly moved away from the main one on the Shuttle. The further away it is, the more accurate they could measure elevation, so they stuck it way out on a pole.

But I think you're right that it's possibly one of the most widely used GIS datasets. Some applications obviously need finer resolution than 30 meters, and lidar is more accurate for the small areas that it can cover, but it's still cool that they could cover most of the globe at 30 meter resolution

The ASTER GDEM data set covers most of the world and is more accurate than non-USA SRTM.

THEY cover at 10m. You get the low-rez version. Better than the 90m before, so who am I to look a gift horse in the mouth.

I work on machine learning on remote sensing imagery, and the Shuttle derived Shuttle Radar Topography maps remain an important freely available Digital Elevation Map (DEM) source.

Initially only 90 meter resolution was made public and some time later 30 meter resolution was available. I think maybe even higher resolution is still classified but can’t remember.

Was shocked we still use 20 year old elevation data extensively.

Are we going to get a public DEM dataset better than SRTMv3 any time soon?

Already available:

https://copernicus-dem-30m.s3.amazonaws.com/readme.html

Pretty cool! Looks like this has the same resolution 30 meter resolution as the global SRTM DEM, but with coverage of the whole globe. Looking forward to using this.

to be fair, between SRTM, ArcticDEM and REMA, you could get all but maybe some southern islands already?

Hopefully the DLR will release the TanDEM-X dataset at some point in the future.

https://en.wikipedia.org/wiki/TanDEM-X

It could be the best global DEM if it's released, but now they're still charging for it. Maybe in a few years as new DLR satellites come on line.

I wonder if this will for some area's of science be a step like going from seeing the surface of your skin and then suddenly seeing it under a microscope and initially being shocked at what you see.

Maybe it will be like "The Core" (2003 box office failure) and we discover Earth is dying way faster than expected, then we have to send a squad of heroes to...

Using radar interferometry, you can get very high-resolution altitude and surface shift detection. I knew the researcher who was doing the synthetic aperture radar back in the 80s starting with SeaSat data, and he was getting very detailed elevation maps and earthquake change maps.

He was also measuring glacial flow.

Funny that the author of the piece doesn't mention the principal researcher on that project at JPL (and she happens to the same last name as one of the other JPL researchers).

Anyway, I can only imagine that this technology's gotten a whole lot cheaper now.

I understand the radar doesn't penetrate water much, is that correct? I worry whether this could be used to detect submarines from orbit - which would be very dangerous, as it could instantly break MAD, which currently hinges on submarines.

That's correct. However, SAR is used to detect vessels on the surface all the time. It's actually been used to discover vessels that are illegally fishing. https://globalfishingwatch.org/impacts/illegal-fishing-north...

At this point, my naive assumption is that all vehicles, boats, planes, and humans with a cellphone are tracked at all times, by someone.

An interesting, and successful regulatory use is detecting oil slicks from bilge dumping (and the vessel associated with them) Example: https://skytruth.org/2019/11/multiple-accounts-of-oily-pollu...

Submarines are first strike tools, that's kind of anti-MAD, so it would be great if they were rendered useless.

You can initiate effective first strikes with submarines, but they provide no counter to MAD. They're not anti-MAD in any sense -- destruction would still be mutual.

And one of the most assured scenarios of destruction is enemy submarines striking back, even if the other legs of nuclear forces are successfully taken out.

Submarines are the hardest to take out, so they are the most pro-MAD.

Submarines are the one thing that will survive the first strike, because your enemy doesn't know where they are. They thus provide a guarantee on the "Mutually" part of MAD.

Radar doesn’t but magnetic fields do. https://en.m.wikipedia.org/wiki/Magnetic_anomaly_detector

Radar penetrates pure water perfectly fine. But most water on Earth, particularly that in the ocean, has tons of dissolved material which results in very high radio absorption.

This brings up a question of how to communicate with a submerged submarine. At extremely low frequencies (that's the technical term...) radio waves can reach submarines that are submerged below a few m depths. That's why you have things like https://en.wikipedia.org/wiki/ZEVS_(transmitter) blasting at 82 Hz, detectable everywhere on Earth. The US stopped its equivalent (https://en.wikipedia.org/wiki/Project_Sanguine) in 2004, apparently, in favor of "improved" VLF systems.

Here's what one of those VLF systems looks like: https://en.wikipedia.org/wiki/Jim_Creek_Naval_Radio_Station

> At extremely low frequencies (that's the technical term...) radio waves can reach submarines that are submerged below a few m depths. > VLF [...] https://en.wikipedia.org/wiki/Jim_Creek_Naval_Radio_Station

Even the signal from https://en.wikipedia.org/wiki/VLF_transmitter_DHO38 can "officially" reach around 30m depth in seawater worldwide.

I wish we could use that for OpenStreetMap mapping :(

What a mediocre piece... Random walk among whatever quotes author could find....

Reply on news.ycombinator.com