#inverse problems

[minty-ibuprofen]

21.01.2021, 10/100 days of productivity

Looking over feedback from a report my friends and I wrote. The exam will essentially be a defense of this and a couple other reports we wrote together. Thankfully, they were well received enough that I’m confident we’ll do well.

[dishesfromthedish]

Root vegetable or space rock?

I picked up a sweet potato at the grocery store last week that looked suspiciously like a binary asteroid.  Little did I know it looked just like (8567) 1996 HW1 until I coincidentally started reprocessing data from this binary space rock, two lumps of stones stuck together and orbiting the sun.

Robert made some crack about how you couldn't distinguish it from an actual asteroid, I made some remark about how you could try to take a radar image of it, or 3D model it, and see if people could guess its provenance as being terrestrial or cosmic, and then...

I went to our local CAD expert, Rhys.

"Do you have your camera on you?"

"... yes..."

"Could you cad up this sweet potato?"  

I handed it to Rhys before lunchtime, who reluctantly agreed and was appalled at the presence of a sweet potato on his desk as a paperweight.

After lunch, he came into my office, looking entirely disgusted.

"This is the most STUPID graphics project ever," he practically spat, "come pick up your potato." 

I followed him back to his office where a 3D model of asteroid "Batatas" awaited on his monitor, done up in Blender via photographing it from every conceivable angle and then using some AutoCAD software to create the vertex-based model.  Tremendously excited, we summoned Patrick and watched as Rhys applied a rocky texture to the rendering of the tuber.  It looked just like a contact binary asteroid, or 1996 HW1!  Could we get synthetic radar images from this model?

El Jefe joined us, and suggested that this was a great idea.  We could indeed make synthetic radar images from Rhys's 3D model—rotational lightcurves as well—add some noise, and then turn this into an inverse problem and see if we could coax a sweet potato shape out of 2D radar images.  (Other near-Earth asteroids look like muffins, dog bones, and even roast chickens.  Why not add sweet potatoes to the mix?)

We could create fake observations for asteroid "Batata", and put it in the logs to troll our JPL colleagues!  What will they think of asteroid 2013 T8TR?  Will they try to process and observe it?  What should we do for the ephemeris file?  At some point Rhys was laughing hard enough that I justified bothering him to CAD up a potato.

Go forward on the inverse problem!  Give it to summer students to tackle!  Projects for everyone to establish how well radar works for imaging space rocks.

Perhaps my idea of a website, "root vegetable or space rock" isn't that far-fetched?  Will this be a DPS poster?  Or two?  "Digging Asteroids out of Noise: Modeling the Radar Signatures of Sweet Potatoes," perhaps. An April 1 arXiv submission? Stay tuned!

And there you have asteroids for vegetarians and Paleo people.

[shonk]

The publishable part of my dissertation has finally been published. And in a reasonable journal, even.