Join us February 25th: The Terroir of Landslides

[TC Haddad]

Greetings PDXOSGeo!

Join us once again on 4th Wednesday (February 25th), to continue the conversation with David Percy, aka "Percy" for more of his research on applying Reconstructability Analysis (RA) to categorical raster datasets.  Percy is preparing to defend his dissertation on Spatial Reconstructability Analysis, and led us in in a lively discussion in January. This month he will share another aspect of his work.

RA is a form of machine learning algorithm (MLA) that works exclusively with discrete data, so categorical data such as the variables in the SSURGO soils database, the geologic database (DOGAMI), and land cover (NLCD) can be used as inputs in their native format without conversion to dummy variables, as is typically done in other MLAs.

For his February demo, Percy will show how these variables are used along with topographic data to model landslide susceptibility in two of the counties ranked highest by FEMA for most landslide hazards: Lincoln and Tillamook. Percy will show that there is evidence to suggest that we can map a Terroir of Landslides, using combinations of soil type, geology, elevation,and terrane group, similar to the combinations of environmental factors that are used to characterize wines, chocolate, and coffee. Results will be presented that show how these data are useful in making hazard maps.

All of the extraction for analysis was done in QGIS and Python using open source GIS libraries Shapely and Rasterio. Code will be shared and discussed!

Time / Space Coordinates:

Wednesday February 25th, 6:30–8pm | Hot Lips Pizza @ the Natural Capital Center: 721 NW 9th Ave, Portland OR 

See you there!