James Larsson

This is one page from a field packet I created while working at Clean Streams on a contract for Maryland DNR's Maryland Biological Stream Survey.

This was my final project for Spatial Modeling in my Master's program at University of Maryland. The model parameters and design were selected from other examples I found in research literature for explaining physical/chemical measures of stream quality, and I wanted to explore whether these same tools could be applied for biological stream quality. This model was carried out using regression tools in ArcGIS Pro, including the Random Forest algorithm. I did extensive data preparation for this analysis, including an ArcGIS Model Builder workflow and a custom Python script to generate watersheds from a set of points and a digital elevation model raster, as no adequate tool for this specific case existed already in ArcGIS Pro. The conclusion of my project was that the factors I had data available for were not able to adequately explain stream biological quality, and that further research would be needed to answer the question I posed.

This was my final project for the Spatial Analysis course in my M.S. GIS program. Having recently seen the Total Solar Eclipse of 2024, I had wanted to do my project on a topic related to the next solar eclipse in the US. For my analysis, I considered several different factors, including land use, land ownership, climate patterns, and light pollution, and then ranked the top 20 locations sorted from most suitable to less suitable. My conclusion was that, barring the effects of climate change, Trinity National Forest in Northern California would be the most suitable viewing location by these criteria.

This is a site summary report I wrote for a wetland sampled by myself and one other coworker while working for Maryland DNR. These were part of a project to test MDE's new wetland assessment method for use in stream restorations, the Maryland Wetland Ecological Integrity Assessment.

This is a site summary report I wrote for a wetland I sampled independently while working for Maryland DNR. These were part of a project to test MDE's new wetland assessment method for use in stream restorations, the Maryland Wetland Ecological Integrity Assessment.

In a project for my undergraduate Remote Sensing course, a partner and I used data from Landsat to explore the relationship between surface temperature and calculated NDVI and IBR indices, in both warmer and cooler years, and classified by Landuse type. Google Earth Engine was used to produce the data, and R was used to analyze and chart it. This chart displays the relationship between NDVI and Temperature in Baltimore in 2016.

This was a personal project I did between semesters at UMBC. I had wanted to teach myself how to generate watersheds from a Digital Elevation Model raster using open-source SAGA GIS software, and I decided to do so using my home county as an example.

This was a personal project I did between semesters at UMBC. UMBC has some fields and forests towards the northeast side of campus which other students have informally named after fantasy novel settings, and none of the small trails linking them had been mapped by any online sources, so I had fun taking a couple hours to run around the woods on campus with a GPS in hand, and then when I got home I opened up QGIS on my computer and processed all of my GPS tracks into the attached fantasy-style map. This project was purely a creative endeavor, but gave me the opportunity to practice some of the skills I had been picking up in my GIS classes.