Michell's GIS Journey

 In this analysis, I modeled potential movement corridors for black bears between two protected areas within the Coronado National Forest. Using data layers such as land cover, elevation, and roads, I developed a habitat suitability model that reflects black bears’ preferences and movement constraints. To begin, I reclassified land cover, elevation, and proximity to roads into suitability scores, where lower scores indicate more favorable habitat. These criteria were combined using a weighted overlay, giving the most weight to land cover, followed by elevation and distance to roads. The combined suitability map was then inverted to create a cost surface, representing the ease of movement across the landscape. Next, I converted the protected area boundaries into raster format and applied cost distance analysis to calculate the minimum movement cost from each park across the landscape. Summing the two cost distance outputs highlighted areas with the lowest cumulative travel cost, whi...

  In the first part of this week’s GIS assignment, I performed a land suitability analysis using a weighted overlay method to evaluate potential areas for development based on a combination of environmental and infrastructure-related criteria. The goal was to classify land into five suitability categories, from 1 (least suited) to 5 (most suited), using raster data and weights assigned to different input layers. The analysis involved two main scenarios. In the first scenario, I applied equal weights to each factor: land cover, soil type, slope, distance to streams, and distance to roads. This provided a baseline for comparison. In the second scenario, I applied alternative weights, assigning a higher importance to slope (40%) and reduced weights to the other factors. This alternative approach reflected a scenario where terrain steepness plays a more critical role in site selection. The final results were visualized in a side-by-side map layout, with consistent symbology and classif...

  In the wake of Hurricane Sandy’s landfall in New Jersey, we conducted a GIS-based structural damage assessment focused on a section of the impacted coastline. The goal was to digitize damage data using FEMA-relevant categories and determine patterns in relation to distance from the shoreline. This assessment provided crucial insights that could support insurance claims, emergency response planning, and future risk mitigation strategies. To complete the analysis, we first created a polyline feature class to represent the pre-storm coastline using historical imagery. Then, using post-storm aerial photos, we manually digitized points for each structure along a specific stretch of the coast and classified them into five damage categories: No Damage, Affected, Minor Damage, Major Damage, and Destroyed. We created buffer zones at 100-meter intervals from the coastline to segment the structures into distance bands and used spatial analysis tools to calculate the number of structures per...

  This week’s GIS assignment walked us through a complete spatial analysis pipeline, starting with post-disaster building damage detection and finishing with coastal flood modeling using LiDAR data. We began by creating a Spatial ETL Tool, which allowed us to automate data translation and transformation between different formats. Using this, we compared pre- and post-disaster building data to detect structural changes after a hurricane event. The differences were clear: some buildings were missing in the post-disaster file, which likely indicates they were damaged or destroyed. Visual comparison of layers helped us identify red polygons representing possible destruction zones. The second half of the assignment focused on coastal flooding risk in Florida using DEM rasters. We used both LiDAR and USGS elevation data to identify areas at or below 1 meter above sea level.  We also performed a building analysis on the flooding areas, which revealed key differences between the two e...

Over the past week, I completed four Esri web courses that expanded my understanding of 3D GIS in ArcGIS Pro, focusing on visualization, analysis, and sharing tools. The course Introduction to 3D Visualization provided a solid foundation in navigating 3D scenes, converting 2D features using elevation surfaces, and using extrusion and real-world base heights to enhance spatial understanding. In Performing Line of Sight Analysis , I learned how to analyze visibility between observer and target points, which is useful for assessing obstructions in fields like urban planning or security. The Viewshed Analysis course introduced raster-based visibility modeling, showing how elevation and observer location affect what areas are visible on a map—a critical skill for tasks like emergency response planning or site selection. Finally, Sharing 3D Content Using Scene Layer Packages taught me how to package and publish 3D data for web use through Scene Layer Packages (SLPKs), streamlining collabo...

LiDAR, or Light Detection and Ranging, is a remote sensing technology that uses laser pulses to measure distances and generate highly accurate elevation data. This week’s assignment focused on using LiDAR data to analyze forest structure and vegetation density. Starting with a LAS point cloud, I generated Digital Surface Models (DSM) and Digital Elevation Models (DEM), then used raster math to estimate tree heights across the study area. After filtering ground and vegetation returns, I converted them into raster layers to calculate point density and create a canopy density map. The result clearly showed areas of dense forest cover and open ground, which I interpreted in relation to human-made features such as roads and clearings. I also created a histogram to visualize the distribution of tree heights, revealing that most trees in the area range between 40 and 70 feet tall, with a mean height of about 54 feet. These tools and analyses are highly valuable for forest management, allowing...

  This week’s assignment focused on using spatial analysis techniques to create and evaluate crime hotspot maps. Our objective was to explore how different methods—Kernel Density, Grid Overlay, and Local Moran’s I—can be used to analyze 2017 homicide data in Chicago and assess how well those methods could predict the locations of homicides that occurred in 2018. The goal was not only to map past crime, but to consider how GIS can support proactive decision-making in public safety. In this analysis, I applied the three hotspot techniques using 2017 homicide data. For the grid and Moran’s I methods, I performed spatial joins to aggregate homicide counts by census tract or grid cell. I then normalized the data where needed, such as calculating homicide rates per 1,000 households, to make the results more meaningful. Each method produced a different pattern of “hotspot” areas across the city. After creating the hotspot maps, I evaluated their predictive power by comparing them to 2018 ...