| Basaltic lava flow morphologies provide insight into the eruptive history of the volcano and the
progression of a lava flow. However, classifying these morphologies is a subjective process. The
goal of this work is to develop a quantitative method of describing lava flow roughness. Using
Unmanned Aerial Systems, we created orthophoto mosaics and Digital Terrain Models. We then
performed qualitative a priori classifications using aerial images, and selected areas that
appeared to have a single morphology. We used the root-mean-square height and Area Ratio to
calculate the quantitative roughness of these areas in three-dimensions; we then clustered the
resulting roughness measurements using a clustering technique called the kmeans. We performed
this analysis on data resolutions of 0.1, 0.5, 1, and 2 m/pixel to better simulate satellite data, as
well as performing this analysis using both a scaling moving window and a static moving
window. Endmember lava flow morphologies smooth pāhoehoe and blocky-’a‘ā were easily
identified by the method. Other clusters included small-scale roughness (slabby pāhoehoe),
small-medium scale roughness (lobate, rubbly pāhoehoe), and medium-large scale roughness
(rubbly-inflated, hummocky). Our quantitative method of differentiating lava flows could be
applied to other lava flows, including those on Earth and other planetary bodies.
Keywords: lava flow, volcanology, planetary geology, basalt, roughness, terrain analysis |