GIS Portfolio: Ross Wygmans

GIS Day is an annual global event dedicated to celebrating and sharing Geographic Information Systems with the world. Since GIS Day isn't until November 16 this year, I decided to create my own event. The audience for this event will be my close friends and family because I oftentimes have difficulty concisely explaining what GIS is to them. This will be an informal gathering hosted at my friend Michael's house since he has a huge projector screen set up in his living room. This event would start off with an introductory video of a TEDx talk by Dan Scollon explaining what GIS is and how it is revolutionizing the way we view the world. I would then show some of my own examples of some really cool applications of GIS and explain some of the types of questions that GIS can help us solve. Finally, I would serve an amazing taco dinner! 

LinkedIn is an online social network where professionals can connect, share, and perhaps even find that dream job (or dream employee if you're recruiting). It's like a more courteous and less political Facebook for your career. One of the tasks for our internship this week was to update our LinkedIn profiles to include our GIS internships and highlight skills learned throughout the GIS certificate program. This was helpful for me as I did not even consider adding my GIS internship to my LinkedIn page. I can definitely see the value of including this practical GIS experience since employers are oftentimes looking for experience, even for entry-level positions.  One challenge I have with LinkedIn in finding the balance between being original and creative without forfeiting professionalism. Many profiles I have seen are very generic with few distinguishing characteristics. One opportunity to stand out immediately is with a catchy headline. My current headline is:  GIS Nerd | Tree...

Over the last two weeks, we have been learning more about the various industries using GIS and exploring specific positions within some of these industries. It is inspiring and exciting to see how GIS is utilized by such a wide range of industries, from natural resources to energy to defense. I spent some time looking for GIS opportunities within the forestry/conservation sector since this is the direction that I hope to take my career. I came across an Enterprise GIS Specialist position with F&W Forestry Services, an international forest management and consulting firm, that I decided to apply for. I may not be entirely qualified for this position, but I figured it never hurts to "toss my name in the hat". Besides, I previously worked for F&W as a Field Forester after completing my undergraduate degree in forestry, so perhaps this will give me a slight edge over other applicants with more GIS experience than me. 

My internship this semester is being merged into my current Environmental Specialist position with Alachua County's Office of Land Conservation. Since GIS is used regularly within our program to produce maps and perform basic geospatial analysis, completing my GIS internship through my employer was the logical choice for me. I'm grateful that our Director and Program Manager were so supportive in facilitating me in this arrangement.  One of the main projects I will be working on throughout this internship is creating an updated feature class of all the properties that have been nominated for acquisition through our conservation program, Alachua County Forever, since the program's inception in 2000. One of the previous Program Managers was maintaining a database with this information until 2015, but his database was so convoluted and cobbled together that once he left the program, no one was able to decipher how to use the database. I will need to dig through a lot of data a...

One of the most common uses of remotely sensed imagery is to determine the extent and distribution of various land uses and land cover classes across a particular area. There are two broad categories of digital image classification -- unsupervised and supervised. Unsupervised classification relies on clustering algorithms to determine which land cover type each pixel represents, while supervised classification uses carefully selected training sites to guide the computer's classification.  In this weeks lab, we experimented with both unsupervised and supervised classification techniques. Below is a land cover map of Germantown, MD, that was created using supervised classification within ERDAS Imagine. Training sites were generated using the provided lat/longs for various land cover types including agriculture, fallow fields, water, urban areas, deciduous and mixed forests, and grass. The smaller inset map shows the Output Distance file. Brighter pixels in this image indicate areas t...

There are several techniques that can be employed within ERDAS Imagine to identify specific features using multispectral imagery. These features can then be displayed using different color band combinations to make them easier to distinguish visually. In general, the four steps used to identify features in ERDAS are: Examine the histogram for shapes and patterns in the data. Visually examine the image as grayscale for light or dark shapes and patterns. Visually examine the image as multispectral, changing the band combinations to make certain features stand out. Use the Inquire Cursor to find the exact brightness value of a particular area. In this lab, we were asked to identify three features within a multispectral image using these methods of interpretation based a given set of criteria. We were then asked to select an appropriate color band combination to display these features in a way that makes them clearly stand out from their surroundings.  The map below displays the first ...

This week's lab introduced us to ERDAS Imagine, a raster-based software used to extract information from aerial imagery. We learned the basics of adding layers, examining metadata for each layer, creating subset images, and exporting images to be further manipulated in ArcGIS Pro.  The map below shows the land cover of an area within Olympic National Park. This area is a subset of a larger Landsat Thematic Mapper (TM) image that was resampled to 30m and had the thermal band removed. This image was pre-processed using ERDAS Imagine and was exported to ArcGIS Pro to create the final map output. Each color on the map represents a different land cover class, and the total acreage represented by each cover class is provided in the legend.   

Land cover is the biophysical description of the earth's surface and is directly observable, whereas land use is the documentation of human uses of the landscape and is not directly observable. There are many uses for Land Use / Land Cover (LULC) classification, such as urban planning, natural resource inventories, Utilities, and many more.  In this lab, we were given a natural color aerial photograph of an area in Pascagoula, MS and asked to create a LULC map based on the aerial imagery. We then generated 30 sample points and used Google Street View to "ground-truth" these points to see if our classifications matched what was observed on the ground. Below is the completed LULC map along with 30 sample points. Green points indicate areas that were mapped correctly, and red points indicate areas that were mapped incorrectly. The overall accuracy of this maps is 66.66%. 

Tone and Texture Identifying features in aerial photographs is not always a simple task. Fortunately, there are several principles and techniques available to analysts that can assist in interpreting aerial photos and help make sense of the images captured. First of all, it is important to be familiar with the various tones and textures one may encounter when interpreting aerial images. The map below highlights areas with five different tones (very light, light, medium, dark, and very dark) and five different textures (very fine, fine, mottled, coarse, and very coarse). Considering the tone and texture of an area or object can be a good starting point when interpreting aerial photographs.  Identifying Features There are also several identifying criteria that can be helpful when interpreting aerial images -- shape and size, shadow, pattern, and association. Shape and size define what an object "looks like", shadows provide an extra angle of vision of objects that cannot be see...

Selecting an interpolation method for sample data can result in vast differences in the surface layer output. In this lab, we compared surfaces derived from four different interpolation methods based on water quality data for Tampa Bay. These interpolation methods are Thiessen polygons (or Nearest Neighbor), Inverse Distance Weighted (IDW), spline regularized, and spline tension.  Below is a screenshot of the surface layer created using the spline tension method. Of the four interpolation methods, I believe this method is the most effective at representing the Biochemical Oxygen Demand (BOD) concentrations throughout Tampa Bay because it most accurately reflects the values of the sample points.  

Triangulated irregular networks, or TINs, are a vector-based models used to represent surface morphology. Vertices (or points) are triangulated using various interpolation methods and are connected by a series of edges. Unlike raster-based digital elevation models, or DEMs, nodes can be placed irregularly over a surface, which can contribute to finer detail in more elevationally complex areas. The image below represents an area near Big Bear Lake, CA. The slope of each triangle within the TIN is symbolized using graduated colors -- triangles with the greatest slope are shown in dark purple, and triangles with less slop are shown in light purple. The horizontal lines crossing the landscape are contour lines. The white lines represent regular contours, while the blue line represents an index contour. 

Data completeness is another important metric used to assess data quality. In this week's lab, we compared two sets of road networks for Jackson County, Oregon. The first dataset was from the county itself, and the second dataset was from the TIGER 2000 data. Since there is no standard methodology for measuring data completeness, we used total length of each road network as a proxy for data completeness.  We first divided the county into a 1km by 1km grid so that each grid cell could be analyzed independently, thereby showing the distribution of data completeness across the entire county. Each road network was divided along the grid lines to facilitate the analysis of total road length within each individual grid cell. We then calculated the percent difference between the two road networks using the county dataset as the baseline.  The map below shows the results of this analysis. The TIGER 2000 road network was more complete than the county network in 162 of 297 total grid po...

In this lab, we were asked to compare the horizontal accuracy of two street datasets for the city of Albuquerque, NM. One of the datasets came from the city itself, and the other came from StreetMap USA. Below is a general outline of the process followed.   Selected 20 test points for both datasets evenly distributed throughout the study area. These test points were located at clearly defined intersections.  Created a reference for each test point based on orthomosaic images. Generated XY coordinate data for test points and reference points and calculated accuracy statistics using the following table.  Figure from the Positional Accuracy Handbook. 1999. Minnesota Planning, Land Management Information Center, St. Paul, MN The map below shows the location of the 20 test points selected.  Results City of Albuquerque Dataset Tested 13.206 feet horizontal accuracy at 95% confidence interval.  StreetMap USA Dataset Tested 241.367 feet horizontal accuracy at 95% c...

The ability to describe, quantify, and understand spatial data quality is an essential function of all GIS practitioners and professionals. Two of the most common metrics used to describe spatial data quality are accuracy and precision.  The terms accuracy and precision are sometimes used interchangeably. While these metrics of spatial data quality are certainly related, these two terms describe two distinct attributes. Accuracy refers to how close a mapped representation of an object is to the object's actual location, and precision refers to the consistency of a measurement method.  Precision The map below shows 50 waypoints that were collected at a single location using a Garmin GPSMAP 76 unit. The yellow star represents the average waypoint, which is the average XY coordinates of all 50 waypoints. The buffers extending out from the average waypoint represent the 50th, 68th, and 95th percentiles. These areas contain 50, 68, and 95 percent of all the waypoints taken. The m...

As the Earth's climate continues to change, hurricanes and other tropical events are becoming more frequent and more intense. A significant cause of damage associated with these events is linked to storm surge -- an abnormal rise in sea level above the normal tidal level. In the first part of this exercise, we analyzed damage that occurred from Hurricane Sandy along a barrier island off the coast of New Jersey called Mantoloking. Red indicates areas of erosion and blue indicates areas of debris buildup or sand accretion.  In the second part of this exercise, we predicted coastal flooding that would occur due to a 1 meter storm surge using two different digital elevation models (DEMs) -- a high-resolution LiDAR DEM and a regular resolution USGS DEM. We then analyzed the number of buildings that would be impacted based on each of these DEMs. The total number of buildings impacted based on the USGS DEM was nearly twice the amount based on the LiDAR DEM. This significant difference is ...

A crime hotspot is an area with higher than average crime or risk of victimization. There are several methods used to determine crime hotspots areas. In this week's lab, we looked at three of the most common methods - grid overlay, kernel density, and local Moran's I.   Below is the analysis process implemented for each of these three methods.  Here are the resulting hotspot maps produced by each method.  Grid Overlay Kernel Density Local Moran's I

Within ArcGIS Pro, there are two options for viewing 3D content -- global scenes and local scenes. Global scenes are most suitable when working with large extent content or when the curvature of the earth is an important factor, while local scenes are better for smaller extent content in a projected coordinate system (PCS). Another difference between these two scene options has to do with available visual enhancements. Basic illumination properties, such as ambient occlusion and light source azimuth and altitude, can be adjusted within local scenes. However, global scenes contain one additional feature where the position of the sun can be set to a specific date and time, adding more realism to the scene.  Local scene Global scene One interesting feature I learned about this week in ArcGIS Pro is the ability to link 2D and 3D views within the same project. Oftentimes, viewing a scene in 3D can reveal spatial patterns and relationships that may otherwise be difficult or impossible to...

LiDAR, or light detection and ranging, is a common technology used in remote sensing. Aerial LiDAR systems mounted on aircraft or satellites send pulses of laser light to the surface of the Earth, and the reflected energy is recorded. When used over vegetation, LiDAR signals will typically return multiple values representing the top of the canopy and the ground. These values can be used by foresters, scientists, and other natural resource professionals to calculate variables such as canopy height, canopy density, and canopy structure.  In this lab, our task was to create layers showing canopy density and tree height with LiDAR data from a small section of the Shenandoah National Park in Virginia. This data was originally captured by the U.S. Geological Survey. The first step was to convert the ground and vegetation data from point cloud (.las) to raster, using a multipoint layer as an intermediary step. We then used these rasters in conjunction with several other geoprocessing tool...

For another scenario in this week's lab, I was playing the role of a Park Ranger in the Coronado National Forest. My task to model the potential movement of black bears between two protected areas. This potential corridor was based upon three criteria associated with bear habitat suitability:  Distance to Roads Elevation  Land cover type  In order to complete this task, I first used the Euclidean Distance tool on the roads layer to calculate the shortest distance from each cell to the nearest road. I then reclassified each of the three layers (distance to roads, elevation, and land covery) based on a suitability scale of 0 - 10, 0 being the least suitable for bear habitat and 10 being the most suitable for bear habitat.  I then combined these three reclassified layers into a single habitat suitability model based on the following relative weights:  Land cover (60%) Elevation (20%) Distance to Roads (20%)  Next, I created a cost surface layer by "inverting" ...

For one of the scenarios in this week's lab, I was playing the role of a GIS Analyst for a property developer. My task was to create a map showing suitable land for development based on the following criteria:  Land Cover Soils Slopes Streams Roads  My general process for this analysis was to first convert all of these layers into rasters if they weren't already. For the rivers and roads layers, I used the Euclidean Distance tool to calculate the shortest distance from each cell to the nearest river or road. I then reclassified each of these layers based on a suitability scale of 1 - 5. Lastly, I combined the results of each of these five criteria using the Weighted Overlay tool. Below are my results.  The first map is an equally weighted average where each of the five criteria have an influence of 20% (or 1/5th). The second map shows an alternatively weighted output where each of the five criteria were weighed as follows:  Land cover (20%)  Soils (20%) Slope (4...