The article explains why the asteroid looks like a pixelated sprite taken from the era of Monkey Island.
For those that didn't want to bother reading both articles and just wanted to have a look at the image but then thought "WTF" after having a look at it:
"The individual pulses can be timed very accurately as well, so that the shape of the asteroid can be determined, too. If there is a bump on the asteroid, like a hill, then a pulse hitting that won’t travel quite as far as a pulse that hits a crater. It gets back sooner, and this can be measured. The spatial resolution of this method at the distance of YU 55 will be about 4 meters, so they’ll be able to make an image that’s about 100 pixels across of it."
I’m not really sure why that needs an explanation. All our imaging technologies have a certain maximum resolution. The asteroid is small and far away, ergo we are not going to get many pixels.
But, people looking at the image may not understand it is a radar image, not an optical image.
People might think that the highlights on the top of the image are sunlight, and the dark area below is shadow, but of course they're not, because radar doesn't care about sunlight.
In fact, I'm not sure how they pseudo-colored the images. In vanilla radar imaging, you get the highest return off surfaces that are perpendicular to line of sight (generally speaking). Here's a (model) plane radar image: http://www.fhr.fraunhofer.de/fhr/site/drucken_c336_f4_en.htm... A lot of the radar imaging work is classified.
I feel like I'm about to ask an incredibly dumb question, but why does the radar image show such a pronounced crescent shape on the asteroid? What's causing the radar "shadow"?
It's possible that those images show a side view generated from the radar response, rather than a front view. The pixel intensity could be sample intensity, so the more reflective surfaces there are at a distance Y, angle X, the brighter the pixel at (X, Y) will be. Since the back of the asteroid is shadowed by the front of the asteroid, the radar provides an image of the side facing the radar only.
This is how I generate side and top view images from Kinect data in my lighting automation software, anyway (Shameless plug: see https://github.com/nitrogenlogic/kinradar for an ancient ASCII art version you can play with, or http://www.nitrogenlogic.com/ for the real thing). I don't actually know how the NASA images were produced.
"On 19 January 2029, 2005 YU55 will pass about 0.0023 AU (340,000 km; 210,000 mi) from Venus. The close approach to Venus in 2029 will determine how close the asteroid will pass the Earth in 2041."
11 comments
[ 2.8 ms ] story [ 36.5 ms ] threadSeriously does this kind of image do anything for the scientists or is this just PR?
http://echo.jpl.nasa.gov/introduction.html
Briefly: asteroid geology, better orbit prediction through understanding of mass distribution, modeling for evolution of solar system studies.
There would have to be a rationale, because time on the telescopes is limited.
The article explains why the asteroid looks like a pixelated sprite taken from the era of Monkey Island.
For those that didn't want to bother reading both articles and just wanted to have a look at the image but then thought "WTF" after having a look at it:
"The individual pulses can be timed very accurately as well, so that the shape of the asteroid can be determined, too. If there is a bump on the asteroid, like a hill, then a pulse hitting that won’t travel quite as far as a pulse that hits a crater. It gets back sooner, and this can be measured. The spatial resolution of this method at the distance of YU 55 will be about 4 meters, so they’ll be able to make an image that’s about 100 pixels across of it."
But, people looking at the image may not understand it is a radar image, not an optical image.
People might think that the highlights on the top of the image are sunlight, and the dark area below is shadow, but of course they're not, because radar doesn't care about sunlight.
In fact, I'm not sure how they pseudo-colored the images. In vanilla radar imaging, you get the highest return off surfaces that are perpendicular to line of sight (generally speaking). Here's a (model) plane radar image: http://www.fhr.fraunhofer.de/fhr/site/drucken_c336_f4_en.htm... A lot of the radar imaging work is classified.
http://en.wikipedia.org/wiki/Doppler_shift
This is how I generate side and top view images from Kinect data in my lighting automation software, anyway (Shameless plug: see https://github.com/nitrogenlogic/kinradar for an ancient ASCII art version you can play with, or http://www.nitrogenlogic.com/ for the real thing). I don't actually know how the NASA images were produced.
- the asteroid is 400 meters in size
- it's a "C-type" asteroid (http://en.wikipedia.org/wiki/C-type_asteroid)
"On 19 January 2029, 2005 YU55 will pass about 0.0023 AU (340,000 km; 210,000 mi) from Venus. The close approach to Venus in 2029 will determine how close the asteroid will pass the Earth in 2041."