My day job is to analyze geospatial data on the computer. The data, usually from cameras looking down at Earth and Mars, is really interesting to look at: lava flows, sand dunes, river channels, among many other things. It’s value is primarily to science, but the world’s patterns and abstractness have artistic value. Many people recognize this, and some celebrate this by contributing to The Art of Planetary Science Exhibition in Tucson, AZ. There are many submissions this year from almost a hundred artists, and I’m happy to be one of those contributors this year in 2018. My submission this year is a carved HiRISE DTM, and is my first completed 3D carve of topography. I’m happy with the way it turned out!
Description of Piece: A place on Mars in miniature Fissure and Channel Southeast of Olympus Mons. Carved by a homemade computer numerical control (CNC) milling machine; surface tone painted by hand. Image data were provided by The High Resolution Imaging Science Experiment (HiRISE) camera in orbit around Mars, which were processed into a digital terrain model by the HiRISE Science Team.
The location of the data is from a location (square mark) east of Olympus Mons, Mars.
Perspective digital rendering of the HiRISE visible image data showing the fissure and channels.
Perspective digital rendering of the HiRISE DTM data.
This is the table top of my CNC machine. On the right is a practice carve, aka, attempt No. 1.
Carving in progress. You can’t see the spindle and milling bit because of the dust shoe. A 4″ hose is connected to a blower, pulling dust off the piece as it carves.
This is the completed carve after roughing (1/4″ square end mill, 1500 mmpm, 3 mm doc) and finishing (1/16″ ball end mill, 1500 mmpm, 0.254 doc, 0.33 mm stepover). I would definitely do it differently next time as I made many mistakes. Even though a machine is doing the carving, there are still plenty of “artistic” choice to make.
Because of the scale of the model, topography is subtle. The large cut in the middle is actually about 500 meters wide. On the model, it’s only about 1.5 cm.
Finished it with a black border, a typical view when viewing satellite data on a computer. Hand painted the surface. Ready for this weekend’s exhibition.
Just another view. I was experimenting with the piece upside down. (The bottom is North.)
I have to admit: Flying an unmanned aerial vehicle (UAV) such as the Trimble UX5-HP is a lot easier than flying a kite over lava flows. We covered several square kilometers in no time with this bird! There were two location: the south of Iceland (the Laki lava flow) and in the north of Iceland (the Holuhraun lava flow). We worked with a great group of students in the field participating in the Keck Geology Consortium. They helped us run our mobile UAV airport! I’ll post more details about our field work and the student projects soon. For now, here are a few images of us doing our volcanological mapping. The data will result in orthoimage data at 1-4 cm per pixel, and digital terrain models at 10 cm per pixel!
This is a taste of the data we’re creating at the Lunar and Planetary Laboratory. On the left is an orthoimage of a series of cones along a fissure in the Laki lava flow. On the right is a colorized digital terrain model (DTM) of the same area. No scale; map oriented north.
Setting up for launch near the Holuhraun lava flow
Ground control; our mobile UAV port.
Stepping through launch preparations with Dr. Christopher Hamilton.
Removing the pitot tube cover just before launch.
Somewhere out there is an expensive bird that we’re tracking.
Going through post-flight checklist.
It took several flights to cover the area surrounding the vent of the Holuhraun lava flow with the UX5-HP.
Flying a kite for aerial photogrammetry was difficult because Iceland has relatively unpredictable weather, including the wind. I was able to make a couple of flights in the northern part of Iceland near the Holuhraun lava flow, which finished erupting just a few months before our arrival. Most of our mapping work during this summer of 2015 field campaign was accomplished using DJI Phantom 3 Professional unmanned aerial vehicles (UAVs). You can see more details here at this website, including publication. Much of this work is still underway.
It takes a few people to make the camera system and the kite work together, especially the radio controlled gimbal.
Launching the kite in this area was challenging due to inconsistent winds and rocky terrain.
Here you can see that I have the rig holding the camera suspended from the kite line.
Once the kite and camera are up in the air with good wind, mapping is relatively simple. The camera takes images automatically at an interval I set, and my job is then to just fly it over areas of interest.
This image was taken from a radio-controlled gimbal attached to the kite. Here is a place where glacial, fluvial and volcanic processes all take place.
I worked on a project on lava flows in Hawaii with Dr. Christopher Hamilton. One of the goals was to study and understand the morphology of the December 1974 flow from Kilauea. A couple years prior, I had started a hobby of kite aerial photography (KAP). Because of the uncertainty around flying unmanned aerial vehicles (UAVs) in sensitive places, I was able to leverage my new hobby as a skill to collect over 10,000 kite aerial images of our study site. I used a computer vision technique called multiview stereophotogrammetry to build a digital terrain model at cm-scale spatial resolution for our research team. Read more in the excellent press release written by Daniel Stolte at University of Arizona.
Following our publication in the Journal of Geophysical Research, a few news outlets picked up the story and featured the work on their webpages. Check out the original press release HERE and the research article HERE. I particularly liked the take from AGU Blogosphere and Science Daily. Although apparent thermal inertia (ATI) is typically used for methods in planetary geology, it can still be a useful tool for Earth surface processes.
This time-series of data show how apparent thermal inertia (ATI), calculated from ASTER data, varies across a section of playa and dune surfaces in the White Sands Dune National Monument. High values shown here typically correspond to the wettest areas, where as low values are quite dry.