Asteroid Research: Thermal Modeling
Figure: Left panel is a plot of the reflectance spectrum of asteroid Phaethon. Blue data points show the original spectrum while the orange and green data points correspond to thermally corrected spectra using two different asteroid thermal models (NEATM and STM). Right panel is the same plot zoomed in on the wavelength interval from 2.2 to 4 micrometers.
My introduction into studying asteroids started with the modification and development of IDL software to perform thermal corrections to asteroid spectra in the vicinity of 3 micrometers. Reflectance spectra, as seen in the figure above, are important in the study of asteroids as they allow us to determine the composition of asteroids. Reflectance spectra are the result of light emitted from the Sun reflecting off of an asteroid, which is then observed by detectors here on Earth or in space. Specifically, when sunlight interacts with the asteroid some of the sunlight is absorbed due to the composition of the asteroid. So, a portion of the observed reflected light is taken away by the asteroid. In addition, the thermal properties of asteroids produces light, intrinsic to the asteroid itself, and this light is added onto the observed reflected light. So an asteroid reflectance spectrum, like the "Uncorrected Spectrum" we see in the left panel of the figure above, is composed of reflected sunlight that has been slightly modified due to absorption plus light that is intrinsically coming from the asteroid. Well, the intrinsic asteroid light, that produces most of its light in the infrared portion of the electromagnetic spectrum, must be removed so that the spectrum may be used for analysis. Such corrections are performed by using a thermal model and can produce more accurate spectra like the "Corrected Spectrum" curves we see in both the left and right panels above.
Asteroid Research: Asteroid Taxonomy and Machine Learning
To enhance my understanding of machine learning and how it can be applied in the field of asteroid studies, I made myself a little sandbox composed of the eight-color asteroid survey data set along with the K-Means machine learning algorithm apart of Python's scikit-learn library. As the consensus appears to be that there are three basic types of asteroids, I decided to process the mean colors of each asteroid forcing the K-Means machine learning algorithm to identify three clusters. The results, seen in the figure to the left, generally suggest that clusters 0 and 1 are primarily C-Type and S-Type asteroids, as they are overwhelmingly the most numerous asteroids in those clusters respectively. I must note that some of the types are considered to be subtypes. Cluster 2 shows that D-Type asteroids are the most numerous; however, an argument could be made this cluster is primarily composed of M-Type asteroids as E, M, and P types are spectrally indistinguishable in the eight-color asteroid survey and the sum of their contributions is greater than the number of D-Type asteroids in this cluster. Overall, my sandbox was fairly rewarding.
Figure: Upper left panel is a bar chart of the class distribution of an asteroid data set. The remaining panels represent three clusters identified by Python's K-Means machine learning algorithm, where the clustering is determined by the "colors" calculated for asteroids apart of the Eight-Color Asteroid Survey,
CosmoQuest: Lunar Cratering Statistics using Crowd-Source Data
I had the greatest pleasure of working on a project in which the general public is offered the chance to do what some scientists do, count craters on the Moon. Counting craters is an important tool used in determining a timeline for objects in our solar system. The figure on the right is a GIF plot I produced as apart of a presentation I gave at the 2018 American Geophysical Union (AGU) fall meeting. The corresponding paper should be coming out soon, so keep an eye out for it on my publications page.
Figure: Overlapping images taken by the Lunar Reconnaissance Orbiter Narrow Angle Camera (LRO NAC). Images display lunar surface encompassed by Apollo 15 landing site and were used for CosmoQuest's MoonMapper Project.