The record of climate change for planets and moons with atmospheres can be recorded on the surface by geologic features that only form under specific climate regimes. Often, geologic features are observed in locations where the contemporary climate cannot explain their formation. These anomalous features can be explained through climate simulations of previous epochs that can then be compared to surface mapping.

This is typically accomplished by taking quantitative Globla Climate Models (GCMs), quantitative mapping of surface landforms, and comparing them separately and qualitatively. This leads to great uncertainty in the assessment of how previous climates produced contemporary landforms.

To address this, I developed Python software to directly integrate GCMs and surface maps within a GIS to directly measure correlations between model output and mapping results. This allows for hypothesis testing in a far more robust fashion than has been accomplished before.

On Mars, I have used this technique to determine where conditions for melting of water ice were achievable during past climates and compared those results to where relatively young sinuous channels called "gullies" are found. We found that the highest-elevation extent of recent gullies matches exactly with the highest-elevation extent of terrain on Mars that experienced conditions above the triple point of water within the last million years. This finding was not possible before the technical innovation of quantitatively integrating GCM models with mapping within a GIS.

Below is a movie of one of our simulations. Black dots are the mapped locations of gullies on Mars (Harrison et al., 2015). The color map represents pressure at the surface, with terrain below 6.12 mb transparent, so all colored locations are above the triple point. Contours represent temperature in units of Kelvin, with locations greater than 273K and thus above the triple point for water within the outside contour.