3
$\begingroup$

This NASA/University of Arizona HiRise tweet shows this interesting image of a meteor impact crater that triggered a landslide (or maybe a "dust-slide") on Mars.

The tweet links to this University of Arizona HiRise page: ESP_054066_1920 which extensively documents extensive metadata associated this image including geographic coordinates 11.797° N, 14.763° E, HiRise satellite image data, solar illumination angle, etc.

At the bottom of the linked page is a google viewer for visible, infrared, and topography maps of Mars.

I can't yet see this crater and dust-slide in this viewer, and I don't see any controls for entering a the Mars coordinates or knowing the current magnification or scale of the view shown.

It shows a red box, so I'm wondering if that's the event and if the map viewer will not zoom in sufficiently?

Can this Google viewer be used to find this crater? If so, how? If not, is there another viewer that can zoom in sufficiently to see it, yet zoom out to see a larger view of the area?

Click images for larger views:

enter image description here enter image description here

enter image description here

$\endgroup$
2
+100
$\begingroup$

Q: It shows a red box, so I'm wondering if that's the event

The HiRise page gives some information about the image:

Latitude (centered) 11.797°, Longitude (East) 14.763°

If you use https://marstrek.jpl.nasa.gov viewer and then use the upper right corner airplane icon to go to 11.797, 14.763 you will find yourself centered on the same point as the red landmark in the HiRise page viewer, so I would say that the red landmark marks either the center of the HiRise image or the landslide.

If you open the "JPEG Black and white map projected" image on the HiRise page (or the non-map projected one), you will see that the landslide is almost centered on the image so there is not much different between the two possibilites. The red box is probably the image, but I haven't checked if the scale is correct. The "JPEG Black and white map projected" is 5km across.

Q: [I'm wondering] if the map viewer will not zoom in sufficiently? Q: Can this Google viewer be used to find this crater? If so, how?

The map viewer at the bottom of the page is the same as this one: https://www.google.com/mars/ In the "About" page (https://www.google.com/mars/about.html) it suggests that images for the viewer come from these data sets: http://www.mars.asu.edu/data/ (Themis data set among the others).

These data sets do not seem to include HiRise, so that might be why the landslide is not there.

A resolution problem might be as well, for example if you use this viewer https://marstrek.jpl.nasa.gov and go to the same spot (same Lat/Long) the image will soon become very blurry (and it is also from a Viking set, so probably too old to contain the landslide).

Another problem might be that by using the viewer at the bottom of the page (or also use this http://jmars.mars.asu.edu/maps/?layer=thm_nightir_100m_v8 to directly browse Themis images) the spot indicated by the landmark seems to be in a "dark area" on the right of a "ridge". So maybe this is an additional reason of why we can't see the landslide there.

To get a better view of the landslide, I would follow Conelisinspace suggestion and check the high resolution JP2 image from the HiRise page.

However, viewing these JP2 images is not easy, for two reasons:

  1. the JPEG2000 format is usually not natively supported by the default image viewers on your operating system and installing a JPEG200 viewer is not always straightforward (based on my experience with some Linux and Windows computer, I haven't tried on Mac).
  2. even after installing a JPEG2000 viewer, these images are HUGE! (i.e. 20.000 x 15000 pixels or even bigger!). So opening these images can be very slow / exceed available memory.

My best solution so far involved installing some libraries/Python packages so that in the end I was able to open them using Python.

In the following, I share the code I used to open the ESP_054066_1920_RED.JP2 image for the landslide. In the code, I also cut the image so that the part showing the landslide is shown and the rest is left out. The resulting smaller image is saved to .PNG format and I've uploaded it so that it's possible to see the landslide in this post without running any code.

Link to image on Imgur: https://imgur.com/a/KkjtdQh

Enjoy!

import matplotlib.pyplot as plt
from PIL import Image
import numpy as np

jpgfile = Image.open("ESP_054066_1920_RED.JP2")
data_array = np.array(jpgfile)
print(data_array.shape)

#uncomment the following to see whole image in Python
#plt.imshow(data_array, origin='lower', cmap='gray')
#plt.show()

reduced_img = data_array[7500:12750,10500:13000]
plt.imsave('landslide.png', reduced_img, origin='lower', cmap='gray')
$\endgroup$
  • $\begingroup$ yay! for Python, I'll give this a thorough read when I get to a non-mobile location, thanks in advance! $\endgroup$ – uhoh Sep 5 '18 at 12:48
  • 1
    $\begingroup$ Great picture especially because you can zoom in. Only, the picture is upside down also because landslides go downwards. $\endgroup$ – Cornelisinspace Sep 5 '18 at 12:54
  • $\begingroup$ Good catch, landslides should always go downwards! ;) I think it's matplotlib (imsave/imshow) that displays the array inverting the y-axis... I'm going to check and fix it :) $\endgroup$ – BlueCoder Sep 6 '18 at 8:01
  • $\begingroup$ Fixed the code: the image now display in correct y-axis orientation. I have also changed the code to further reduce the image so that it was possible to upload as it is from the script to Imgur. However, I noticed that Imgur actually converts the image to jpg (lossy compression!) so maybe I could have just left the previous one reduced by GIMP... :) $\endgroup$ – BlueCoder Sep 6 '18 at 8:25
  • $\begingroup$ BTW, did you notice the pattern on the right of the landslide? It originates from a small hill (?) in the north... is it an older landslide that has lost its black color? See this image, that includes the origin of this "grey" landslide up north: imgur.com/a/CtifOT0 $\endgroup$ – BlueCoder Sep 6 '18 at 10:26
2
$\begingroup$

The image below is from the The crater itself is only 5 meters across link on the HiRISE page ESP_054066_1920.

enter image description here

NASA/JPL/University of Arizona

This answer from @BlueCoder shows this image viewer which can be used also to find HiRISE images.

To find the landslide event type in the longitude values 14.7 in the Min box and 14.8 in the Max box.
Also type in the latitude values 11.7 in the Min box and 11.9 in the Max box and Run Query.

This gives you 4 images of the landslide event with which you can zoom in and can toggle the full page.
After that, you can zoom in again and move the image !

$\endgroup$
  • 1
    $\begingroup$ Thanks; this appears to be just a empty-zooming of part of the image in the question, and it doesn't qualify as finding it in a viewer. $\endgroup$ – uhoh Sep 8 '18 at 12:32
  • 1
    $\begingroup$ Right,, just added some colorfull information. $\endgroup$ – Cornelisinspace Sep 8 '18 at 12:41
  • $\begingroup$ yep, thanks! I felt a little bad because I mentioned the bounty here; I was hoping you wouldn't feel slighted. $\endgroup$ – uhoh Sep 8 '18 at 13:59
  • 1
    $\begingroup$ I added information how to use the viewer that @BlueCoder found to zoom in on the landslide event. $\endgroup$ – Cornelisinspace Sep 10 '18 at 9:50
  • $\begingroup$ Neat! That's now a lot more convenient than having to manually download the HiRise J2 image. I wonder if the maximum resolution is the same (for sure it's very good anyway). $\endgroup$ – BlueCoder Sep 10 '18 at 10:25

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.