Mt. St. Helens DEM Time-Series — USGS Partnership

Overview

Created for the GIS Project Management Capstone course at PCC as part of an ongoing project digitizing hand drawn contours contemporary to the events, documenting the geomorphic evolution of Mt. St. Helens from pre-1980 eruption to the 6 years of dome change within the caldera through physical 3D prints of 5000 m by 5000 m geographic area derived from USGS digital elevation data that all nest together like matryoshka nesting dolls. A Storymap detailing the entire process is available.

An interactive 3D viewer was produced using the same 3D objects as the 3D prints produced for the capstone project, to have a path to share the work and offer an opportunity for anyone to interact with each of the survey epochs: the pre-eruption “snowcap” originating from 1952 (the last time the mountain was surveyed before the eruption), the 1980 post-eruption surface, 1982 caldera dome rise, 1984 caldera dome rise, and the 1986 dome rise that ended the dramatic series of major geomorphic changes to the mountain.

Data

The project works with:

• 1952 pre-eruption DEM — derived from USGS topographic surveys, representing the full pre-eruption summit
• The main base model is a DEM from 1981 with coverage of the entire AOI that was lacking in the other dome rise DEMs, however that surface is only represented outside the crater rim. The surface within the crater rim is the October 1980 surface.
• Four post-eruption dome DEMs (1980–1986) — capturing the progression of dome growth at intervals ranging from weeks to months in the immediate post-eruption period

Each DEM represents a different stage of dome development that built a new lava dome hundreds of meters high inside the crater.

Methodology

Data Ingestion and Prep

The historical hand drawn contours were discovered in the map cabinet of a retiring USGS employee, and were produced using different traditional photogrammetric methods, contour intervals and reference datums. Previous PCC student projects had digitized these into DEM rasters, but they were still in three different resolutions and two different PCRSs. Before any change analysis can be done, each raster must be warped to a common spatial reference and downscaled as close to a unified resolution detail as possible.

QGIS, Python and GDAL were used for reprojection, cubic spline upsampling, gaussian smoothing and alignment. Then a unique QGIS plugin, DEMto3D, which converts any elevation surface into an STL file ready to be input into slicing software and then print. For this project, DEMto3D’s outputs were brought into Blender to check outputs, align and modify.

3D Printed Dome Sequence

A nested sequence of 3D-printed terrain models was produced to represent key time steps, scaled physical models that can be held, manipulated, stacked or displayed in sequence to illustrate dome growth as a tangible artifact rather than a raster on a screen.

The prints are paired with an interactive 3D viewer website produced using the same 3D objects as the 3D prints, to have a path to share the work and offer an opportunity for anyone to interact with each of the survey epochs.

Status

This project is ongoing in collaboration with USGS. Research into Empirical Bayesian Kriging showed promise for successfully interpolating the DEMs to create more natural transitions between survey epochs. A number of opportunities for talks and presentations are being pursued as well.