Stuart Robbins of the Southwest Research Institute in Boulder does that others don’t even dare to start: he counts the craters of the Moon.
How many craters are there on the moon?
“It’s a question that cannot be answered. You have to define a diameter first. And once you do, any number quoted must be taken with the caveat that every individual will vary in what they consider to be an impact crater (e.g., Robbins et al., 2014
). Based on how I
identify craters (which appears to be similar to many others in the field), I can say there are about 9300 craters larger than 15 km on the moon — I’ve mapped all of these, globally. My projection
for craters ≥1 km on the moon is approximately 1.1 million
. There are several ways to estimate this based on the areas I’ve completed so far (35% of the moon — see this poster
), but in the end, a simple linear extrapolation at this point is about as good as a more detailed one based on fraction of maria versus highlands. I should know the answer in a few months.”
Is there any crowd-sourcing involved or you are counting one by one?
“I am involved with a crowd-sourcing crater effort (CosmoQuest
), but this global moon database is only me, it does not involve any crowd-sourcing.”
For morphological attributes, if there will be funding, are you going to inspect all craters individually or is there an automation for that?
“Individually. I am aware of one or two automated methods, but they are even worse than automated crater detection and it would be more work to correct them than to just do it all manually.”
What was the fun in this work, if any?
“To be blunt, the only fun is when it’s done and you can look back and see the fruits of your labor. Tracing circles is incredibly tedious and boring.”
Did you have to redo some parts?
“I have re-done a few very large impacts that I originally did several years ago (in support of some CosmoQuest work). I have also re-done parts of the south pole within a few degrees of the pole with LOLA GDR that were originally done with LROC WAC mosaics because the GDR is significantly better quality there. I learned from that for the north pole and started with LOLA GDR rather than LROC WAC within 5° of the lunar north pole.”
Major discoveries so far or predicted using this dataset?
“I have had little time to do a look-back and actually analyze the data beyond producing basic density maps which are, at the moment, only for about 1/3 of the lunar surface. The only thing I think I’ve found that I have not seen reported elsewhere is that the moon’s north pole is saturated with kilometer-scale secondary craters.”
When you finish it, are you going to throw out all craters smaller than 1 km?
“Craters smaller than 1 km will be available on an as-requested basis, considered supplementary material. People misuse and abuse datasets an extraordinary amount, and crater populations that are not “complete” are abused more often (this has been encountered and is still being encountered by Nadine Barlow for her 1980s Mars crater database). The “just contact me for the rest” is a trivial hurdle but it would allow me to personally, individually remind people of all the caveats of those craters. This is also the model followed for the Mars database published in 2012 (Robbins & Hynek, 2012).”
What’s the main difference compared to the making of the Mars crater catalog? (not the result, but the process of producing the catalog).
“Mostly this comes down to experience. I had to re-do a non-trivial portion of Mars because it was done early in my graduate student career and I got “better” at it and didn’t want there to be regional biases based on how I had changed. I also did different “passes” through Mars where I went to different diameters for completeness or other things, and these were saved as individual files that lead to a book-keeping nightmare and numerous very large files due to duplicating topography datasets. This time, things are streamlined with “passes” corresponding to what datasets were mapped on ONLY (so all WAC is done together), and just the basic “how am I going about this” did not require trial and error. Additionally, I’ve grown faster with how quickly I can measure and catalog each crater, so while constructing the Mars database I started out at 1500-2500/day, and by the time I was done was up to 2000-3000/day, at this point with the moon, 5000/day is below average, and 10,000/day is a good day.”
A portion of the Moon, showing all cataloged craters in green (outlines for crater ≥15 km in diameter, dots for 1-15 km). Source