John Hogland
Dr. Hogland is a Research Forester working for the Rocky Mountain Research Station. His research interests revolve around quantitative methods within geographic information systems (GIS) and understanding the relationships between landscape patterns and forested ecosystems processes. Current projects include: 1) quantifying forest characteristics at fine spatial scales, 2) designing, developing, and building new procedures that integrate machine learning and statistical modeling with fast raster processing (Function Modeling) to streamline spatial modeling and reduce storage space associated with GIS analyses, and 3) developing sampling strategies focused on reducing the cost of sampling while maintaining the characteristics of a representative sample.
Sessions
In this workshop we will explore how to set up a geospatial data science environment, use that environment to create a well spread and balanced sample, and estimate canopy cover from data derived from STAC, OSM, and REST services.
Raster texture is an important attribute for sample design and raster-based classification, regression, and clustering. We present an automated approach and companion Jupyter notebook that uses principal component analysis and convolution kernels to quantify orthogonal intensity and texture metrics.