The rectangular structure in the images is matching exactly the pixel direction of the image itself. That gives a clear signal that something might have been done with the image itself.

As you can probably see already, the image is composed of a background which shows height differences and a foreground that is red to blue. I do not know the exact meaning of the colors, but it seems that the red-blue image is an infra-red image projected onto the image with the height differences.

The rectangular structures in the image are large and overlapping each other. That gives an hint that this might be a very low resolution infra-red image that has been zoomed in and filtered. The infra-red is usually of much less resolution than normal images. Still we can see some interesting details on some places.

So what happened here?

It seems that they have used a height-image and have calculated the heights of the area. Because that operation produces a lot of small errors, these errors are filtered away at places where the height is difficult to calculate. That is mainly at the flat areas. An often used filter that is used in height calculations is the median filter, which produces rectangular structures. It can even generate pyramid like structures. These are all filter artifacts.

The infrared image is probably taken from a different angle. To project the infra-red image onto the height-image they did a calculation using the height of the terrain. But both data are so inaccurate that it produces a strange image like this.

I wonder if the infrared image is even aligned with the features on the height-image, because they are so different. Even the hills on the left, that show the most details, do not form regular patterns in the infra-red.

If we can retrieve more data from the NASA we can better see what filters and calculations they actually used.