Formation and thermal evolution
My early work modeled heat and mass transport in comet nuclei and other icy bodies: when volatiles move or survive, how interior structure changes, and what present-day activity may remember about earlier conditions.
Research
I tend to work the same way in very different fields.
Start with a physical model, ask what is actually observable, and let the data push back. That approach has followed me from comet interiors and asteroid dynamics to calibration of Earth-observing instruments.
Research framework.
How I think about the problem
Forward models on one side; inference on the other.
The diagram uses “forward-engineering” and “backward-engineering” as my own shorthand. In more standard scientific language, the left side is forward modeling: start from physical assumptions and calculate what should happen. The right side is inverse modeling, retrieval, or inference: start from measurements and work back toward the physical state that could have produced them.
The interesting part is the loop between the two. A model that cannot reproduce the data needs work; a measurement that cannot be interpreted through the physics is not enough by itself.
Planetary science
Small bodies carry a long memory.
Comets, Centaurs, trans-Neptunian objects, and asteroids preserve clues about formation, later processing, and movement through the Solar System. My planetary work has mostly asked how those pieces fit together.
Planetesimal formation schematic.
Dynamical pathways
Later work coupled physical questions to orbital history, including the Centaur-to-Jupiter-family-comet transition and the Gateway region exterior to Jupiter.
Linked processes
Dynamics, surface activity, structure, and interior all talk to one another.
A small body does not have separate “dynamics” and “physics” histories. The orbit sets the thermal environment; the thermal history changes the interior and surface; activity changes what we observe; impacts and binary dynamics can redistribute material again.
The figure is a synthesis of modeling questions I have worked on at different stages of my career. It is not one single end-to-end model.
Selected small-body modeling summary.
Earth observation
Calibration is not housekeeping. It is part of the measurement.
For a satellite record that runs for many years, the instrument changes, the orbit changes, and the geometry changes. If those effects are not understood, they can be mistaken for changes in the Earth scene.
Reflective solar bands
I support calibration of the reflective solar bands on Terra and Aqua MODIS as a member of the MODIS Characterization Support Team. Recent work has focused on what happens as orbital drift changes the solar-diffuser viewing geometry late in the missions.
Day–night band
I also support calibration of the VIIRS day–night band on NOAA-20 and Suomi NPP. The DNB spans a very large radiometric range, so gain behavior and long-term calibration need to be tracked carefully.
Collaborative work includes NOAA-21 calibration/performance analysis, and I presented the DNB calibration update at the 2026 MODIS/VIIRS Calibration Workshop.
Scientific computing
The code is part of the scientific argument.
I use computation to make physical assumptions explicit, compare them with data, and repeat the analysis without relying on a one-off calculation.
Numerical modeling
Heat and mass transfer, orbital dynamics, ejecta evolution, and related physical systems.
Data analysis
Time series, trends, uncertainty checks, cross-comparisons, and diagnostics for large scientific datasets.
Research software
Python, C/C++, Fortran, shell scripting, Linux, and Git. I care about code that can be checked, reproduced, and maintained.