GIS 5100 Module 6 - Part II: Least Cost Path and CooridorAnalysis

In the second half of Module 6, we built upon the knowledge obtained in Scenario's One and Two: reclassifying raster datasets to generate a Suitability Map. From there, we were able to continue analyzing the datasets to obtain a Least Cost Path. A Least Cost Path analysis is a geoprocessing workflow that determines the lowest cost path from a source location throughout an entire dataset. As stated in the first post for Module 6, 'cost' is a relative term that does not necessarily refer to monetary units; however, in the map below, a Least Cost Path analysis was performed for an oil company wishing to install a pipeline through a southwestern portion of the state of Oregon. Once the datasets have been reclassified and combined using the Weighted Overlay tool, a Cost Distance function can be run to determine the cost from a set source location. As shown in the map below, the Least Cost Path is derived from the Cost Distance output, and the dashed line represents the lowest-cost route through the study area. Like the Suitability Map, factors can be weighted to give priority to some criteria over others. For instance, the bottom-left map in the image below provides a Least Cost Path based on slope alone. However, if cost is added for river crossings, the Least Cost Path is drastically altered to minimize the number of times the pipeline crosses the rivers in the area [see bottom-right map below.]

Another type of map that provides least cost information is a Corridor Analysis map; a Corridor Analysis map determines an area, or corridor, of least cost between two points; the Corridor analysis does not take direction into consideration. As illustrated in the Corridor analysis map, the Least Cost Path falls within the lowest classification of the corridor area. This would provide the oil company of an area to build the pipeline within, allowing them to make final decisions regarding pipeline installation while maintaining cost efficiency.

 Lastly, in Scenario 4, all of these processes were applied to create a Corridor analysis on travelling bears throughout Coronado National Forest. While no new processes were added to the workflow involved in creating this map, it was a great opportunity to apply all tools / functions in a cumulative effort to produce the deliverable below. I was quite happy with this deliverable and believe it to be an outstanding representation of my comprehensive understanding of Suitability Mapping, Least Cost Path Mapping, and Corridor Analysis Mapping.

GIS 5100 Module 6 - Part I: Suitability Mapping

Module 6 of Applications in GIS was very extensive and loaded with information; the Module was divided into two parts, with two scenarios for each part. This first portion of the module focused on Suitability Mapping; a suitability map identifies areas [within a study area] that meet some, or all, criteria for a given problem. For example, in the map below, regions within the study area are highlighted that are optimal for cougars to live in; these areas were determined by meeting four criteria: distance from roadways, proximity to rivers, areas with increased slope [mountainous or canyon], and forested areas. To determine which areas meet all [or none] of these criteria, a reclassification process had to be run on the provided datasets. For instance, the Digital Elevation Model was processed using the Slope Calculator function and reclassified into two distinct classes: areas with < 9° slope and areas with > 9° slope. Next, a reclassification was employed on the landcover dataset, distinguishing forested areas from every other landcover type. The same process was completed for proximity to roadways and rivers [a Euclidian Distance raster dataset was created prior to the reclassification process.]

After all datasets had been reclassified, the values for each cell could be added together using the Raster Calculator too, and the analysis could begin. Referencing the map above, three variations on the analysis are displayed. The two bottom maps illustrate regions within the study area that meet ALL four requirements, thus being determined as the likeliest places for cougars to dwell; the only difference between the two bottom maps is one is vector-based [comprised of polygonal geometries], while the bottom-right is raster-based [a grid array of pixel values]. The top map, however, displays the study area using a graduated symbology; this tells the map viewer how many criteria are met for the entire study area. A visual comparison of the top map with the bottom two clearly illustrates all three maps highlight the same optimal areas for a cougar to live.

The second scenario of Part 1 was a similar analysis, only determining which areas within the study area would be best for a future development. The criteria used in this analysis were: proximity to roadways, proximity to water, current land classification / land use, and the slope of the land. While there were some variations within the criteria, and what was determined to be optimal for development, the workflow for Scenario 2 was the same. As displayed in the map below, a graduated symbology was utilized to show how many criteria were met for each pixel within the study area.

The difference between the two maps in Scenario 2 is weighted value. The map on the left gives an equal weighted value for each criterion, while the map on the right does not. This is significant because some factors are going to affect the suitability of potential sites more than others. For instance, while proximity to water will have some effect on the location of a future development, it will not be the same effect as not having any available roadways to access the potential site. 

As outlined in this post, the applications of Suitability Mapping are endless, and can be used in a wide variety of ways. Although this lab was very time-consuming and covered a tremendous wealth of information, it was a great experience continuing to learn about the potential applications of Geographic Information Systems.

GIS 5100 Module 5 - Damage Assessment

This blog post will be short and sweet, as all details of this project are outlined in the ArcGIS StoryMap shown below. Due to the confines of the blog layout, the StoryMap is displayed as it would appear on a mobile device; a full-screen version can be viewed here. Also, larger images of the maps created throughout this analysis are displayed at the bottom of this post.

 

 

GIS 5100 Module 4 - Coastal Flooding

This module of Applications in GIS was the most time-intensive, complex, and in-depth assignment yet. The lab was broken up into three separate analyses, each looking at the effects of coastal flooding. The first part of the lab, compared two Digital Elevation Models and illustrated their differences via a raster output. The first DEM was a LiDAR dataset that was obtained pre- Hurricane Sandy and the second was a LiDAR dataset that was obtained post- Hurricane Sandy. This portion of the lab was fairly straightforward; it was essentially taking the values of the post-storm dataset and subtracting the values of the pre-storm dataset. The output was the raster displayed in the map below. Some noteworthy points are as follows:

• The red areas are negative changes in elevation. It is apparent that these are mostly found along the shoreline and are representative of beach overwash, beach erosion, and buildings that were demolished during the storm.
• The lighter shades of the raster output represent areas that had less [or no] elevation changes due to Hurricane Sandy.
• The blue areas are positive changes in elevation. This is most likely due to water buildup, sand buildup, and debris buildup. 
• The red strip on the westerly border was excluded from the visual analysis due to its location along the border; it would not make sense for the elevation to gradually build up with an abruptly defined edge of negative elevation change.

The second portion of the lab was also very straightforward. For this analysis, we were given a dataset of New Jersey and tasked with calculating the percentage of Cape May County that was affected [flooded] by the two-meter storm surge that occurred during the 2012 hurricane. To accomplish this, we reclassified the data into two classifications: elevations that were less than or equal to two meters and all other elevations. After the data was reclassified, the raster was converted into a vector-based polygon, the area of the polygon was calculated, and the percentage of affected land was obtained by clipping the dataset to the Cape May County feature class and then solving the following equation:

Area Affected = [Total Area ≤ 2 meters] / [Total Area of Cape May County]

The third, and final, analysis of this assignment was a culmination of all steps executed throughout the first two portions. The dataset was a study area that is located in Collier County, Florida [southwest region of the state]. For this analysis, we were provided with a LiDAR dataset and a United State Geological Survey DEM, or Digital Elevation Model. For this particular scenario, the objective was to see how much land would be affected by each DEM for a one-meter storm surge. This was obtained by following the steps of the second portion of the lab for each DEM, the one from USGS and the LiDAR. After these affected areas were created, quantitative analyses could be done to determine the extent of damaged buildings within the study area. The map is displayed below. 

Although I am not entirely sure what caused all the confusion within this portion of the lab, I do know that stepping away, focusing on something else, and taking a few deep breaths were the most effective remedies. On a few of the functions utilized throughout this lab, simply shutting down the software, reopening it, and rerunning the tool was the solution to fixing incorrectly derived outputs. At other times, taking a few steps back in the process, slowing down, and reworking the lab was the solution to fixing other incorrectly derived outputs. I am confidently pleased with the deliverables of this lab, but there are some tools I want to become more accustomed to working with: spatial joins and table joins. As many roads often lead to the same destination, knowing the correct tools to use can make the workflow more streamlined, thereby increasing efficiency.

GIS Module 3 - Visibility Analysis

Module 3 of Applications in GIS was a four-part assignment that covered different topics of the ESRI tutorials page. It consisted of four hands-on tutorials, four quizzes, and four introductions to the various capabilities of this extensive software platform. The four tutorials were: Introduction to 3D Visualization, Line of Sight Analysis, Viewshed Analysis, and Sharing 3D Content using Scene Layer Packages. Each are discussed in detail below.

3D Visualization

The first ESRI tutorial was an introduction to the 3D capabilities of ArcGIS. While all of the data was provided in 3D format, it was a neat introduction into 3D navigation [slightly different than navigating a 2D map], the difference in various 3D map scenes [Local and Global], and the 3D symbology that is included in the software package. The ESRI documentation provides a 3D Workflow page that provides a lot of introductory information; this page can be viewed by clicking here. As shown in the map below, we were provided with all the required data to create a realistic, three-dimensional map of San Diego, California.

Line of Sight Analysis

The second ESRI tutorial was a lesson on Line of Sight [LOS] analysis. LOS is a visualization of which points [along a line] can be seen from an observation point; the ESRI documentation for this geoprocessing tool can be view here. This tutorial was based on the 3D knowledge obtained in the first lesson, so it was important to work these in order. The subject of this lesson was to analyze how much security coverage there was on a parade route that progresses through downtown Philadelphia, PA. Essentially, two security observation points were chosen, and the tutorial provided an analysis of how much of the parade route was visible from those two security observation points. As shown in the image below, the two security observation points are symbolized by reddish colored dots on the roofs of two buildings located in the central business district. The LOS analysis was performed at 30 feet increments [a parameter of the geoprocessing tool], and the visible points are notated with lime green lights. The choice of these two security observation points provides a great deal of visual coverage while the parade is taking place.  

Viewshed Analysis

While Line of Sight Analysis function executes among points or along lines, the Viewshed Analysis is performed over geographic areas, as in 360° from the observation point [unless set differently in the user parameters]. Click here to view the ESRI documentation for this ArcGIS Spatial Analyst tool. In the third tutorial, property owners of a campground wanted to illuminate their property using two specific lighting fixtures; specifically, they knew where the fixtures were going to be placed but needed to know at what height the fixtures needed to be placed to achieve 50% illumination, or greater. As shown in the map below, a viewshed analysis was performed [according to the specifications of the light fixtures] at a height of 3 meters and at a height of 10; it is obvious that the 10-meter height exceeds the illumination requirement while the 3-meter height is substantially lower.

Sharing 3D Content Using Scene Layer Packages

The fourth and final tutorial was taking what was taught in the previous modules [specifically the Introduction to 3D Visualization Tutorial] and sharing it to the web. A great resource on this subject is the ESRI Share A Web Scene Layer documentation page, and this hands-on tutorial was a great introduction into Scene Layer Packages and sharing them to ArcGIS online. The final product was a 3D landscape of Portland, Oregon published to the web. A still image has been pasted below, and the 3D map can be viewed by clicking here.

GIS 5100 Module 2 - LiDAR and Forestry

Module 2 of Applications in GIS was an extremely different GIS application than the crime analysis study completed in Module 1; Module 2 focused on forestry analysis on a LiDAR dataset. LiDAR [Light Detection and Ranging] is a remote sensing technology where thousands, if not millions, of laser pulses are emitted from a sensor, bounce off the surfaces below, and the amount of time it takes for the pulse to return back to the sensor is measured to determine the distance between the landscape and the sensor. Once the pulse returns, the X, Y, and Z data is collected, and the entire dataset is compiled into a LiDAR point cloud data file. This compiled dataset can display all features that are present on the landscape. The LiDAR point cloud raster shown below is from a swath of forestry in the state of Virginia. As illustrated in the legend, areas of higher elevation are displayed by warm colors of the spectrum and transition to the cool colors of the spectrum as elevation decreases. 

After the LiDAR point cloud dataset has been imported and processed accordingly, the continuous raster dataset can be converted into a two-dimensional multi-point dataset. Once this has been accomplished, geoprocessing tools can be ran to transform this vector dataset into Digital Surface and Digital Elevation Models [DSM and DEM, respectively]. A Digital Surface Model is an elevation model that captures both the environment's natural and artificial features. Conversely, a Digital Elevation Model is a two-dimensional raster output representation of a continuous surface. As LiDAR sensors can have many returns from the same laser pulse, the processing software can distinguish between returns that are representing characteristics protruding from the landscape from those that are actually hitting the terrain. Once this entire dataset has been processed, a DEM can be derived, creating an accurate representation of the terrain that lies below any protrusions from the Earth's surface; this Digital Elevation Model is also known as a Bare Earth Model. The DEM of the Virginian forest swath can be seen below. The same color scheme used for the LiDAR point cloud dataset was applied to the DEM; this was an intentional choice to show the topographical similarities between the two datasets.

After the Digital Elevation Models and Digital Surface Models have been generated, further geoprocessing functions can be employed to conduct further analysis on the study area. Geospatial analysis examples could include vegetation height [map below, left] and canopy density within the study area [map below, right]. The information gathered by these data outputs can easily be displayed in a graphical form, such as the Distribution of Vegetation Heights histogram, displayed in the lower left-hand corner of the map below. 

Some noteworthy characteristics of this Virginian Forestry Analysis, is the majority of vegetation heights present [within study area], are within the range of 40 to 80 feet, with the statistical mean at approximately 54 feet. Also, a direct correlation between vegetation height and canopy density can be discerned upon initial observations of the map.

This module was very informative on how continuous, raster-based datasets can be used to gather further information on land classification and other characteristics of the natural environment.

GIS 5100 Module 1 - Crime Analysis

Module 1 of GIS 5100 [Applications in GIS] was very interesting and insightful; it was also a great introduction into one of the many ways GIS systems can be used to improve the world in which we dwell. The module focused on using different hotspot analysis methods to map crime in two cities; after the information was mapped, the data could by further analyzed to see which mapping technique would be more effective at preventing the same type of crimes from taking place in the future.  

The first half of the module was based on datasets from Washington D.C. for 2018. First, we created a thematic, choropleth map [see GIS 5007, Module 5 for more information on choropleth mapping] of burglaries throughout 2018 [per 1000 homes] that occurred in each census tract that lies within Washington D.C. [above, left]. While this map does provide insight into which census tracts had more burglaries, it can often be visually misleading to the user, due to the varying sizes of the tracts themselves. 

Next, we mapped all the assaults with dangerous weapon incidents that occurred throughout Washington D.C. in 2018. For this map, we used the Kernel Density Estimation model; this type of model places a user-defined grid over the city and searched each grid cell for how many assaults occurred within a user-defined radius. Each incident is then given a weight that is calculated according to its proximity from the grid cell's centroid. Finally, the density of each cell is mapped, creating a map that shows the assault hotspots of 2018 [above, right].

On the second half of the lab, we analyzed datasets for the city of Chicago, IL to determine which method would give the most accurate predictions of future occurrences of a specific crime [homicide].  To do this, we created three maps of all homicide hotspots that occurred in 2017 throughout the city using three different mapping techniques: Grid Overlay, Kernel Density Estimation, and Local Moran's I Spatial Autocorrelation [see below].

For the Grid Overlay map, a grid was placed over the entire city and a spatial join function was run to obtain a number count of homicides within each grid cell. Once this count was determined, the top 20% of grid cells with the highest homicide counts were mapped. The Kernel Density Estimation technique was described earlier in the post, and the areas having greater than 3x the mathematical mean were mapped as hotspots. Finally, Local Moran's I is a method that determines where homicide cases are random or where they are clustered. This intensive formula is beyond the scope of this discussion, but ESRI's documentation on this Spatial Autocorrelation method can be found here. The formula determined where the clusters [hotspots] exist, and they were mapped in the righthand frame on the map displayed above. 

The remainer of the lab focused on analyzing the statistics produced by each mapping technique. To accomplish this, the total area of each hot spot mapping method was calculated [square miles]. Next, all homicides that occurred throughout Chicago, IL in 2018 were overlayed on each map and a spatial join function was executed to determine the number of 2018 homicides that took place within the 2017 hotspots. Then, the percentages of 2018 homicides within each 2017 hotspot were calculated. Finally, the density was calculated by taking the number of 2018 homicides that occurred within 2017 hotspots and dividing that number by the total area of the 2017 crime hotspots [this was done for all three maps]. 

After a careful analysis of all these statistics, I would recommend the Kernel Density Estimation map as the most effective means to efficiently allocate resources, thereby reducing the number of homicides in the future. The reasons for this choice are as follows:

The spatial area of the KDE map is not the biggest or the smallest; too small of an area would place all resources closely together, causing an increase in response times to areas not within those confined spaces. Conversely, too large of an area, and all resources are spread out too thinly, also causing increased response times. Secondly, the percentage of 2018 homicides happening in 2017 crime hotspots was remarkably high compared to the Grid Overlay map, and only marginally smaller than Local Moran's I. The reason I chose KDE over Local Moran's I was due to the fact that Local Moran's I had a 38% increase in land area over KDE, which would force a sparser allocation of resources. Thirdly, the density of Grid Overlay was higher than that of KDE [Local Moran's I homicide was substantially lower], but this was based on a smaller geographical area and a lower count of 2018 homicides taking place within 2017 hotspots. A careful reasoning of this statistic concluded with KDE being the better option. 

This lab was very interesting and intriguing; an excellent example of how GIS is being used on a professional level every day. I look forward to advancing further in this class to see how GIS systems can be utilized in other real-world scenarios.

GIS 5100 Introduction - About Me Story Map

Hello, everyone! My name is Marc Wright and I am currently in my third semester of the M.S. in GIS Administration program. Outside of school, I love to ride my motorcycle [2008 Harley Sportster 1200], spend time with my family, work out at the gym, read, draw, and work on the lawn / landscaping. My wife and I are also very active in our local church [Oklahoma City Community Church] and have recently committed to service opportunities that revolve around feeding the homeless within our community. Professionally, I work full time as a Network Designer / GIS Technician / Mapping Technician for a fiber optics company based out of Norman, Oklahoma. I am thankful that I get to utilize some of the knowledge that I am obtaining from the coursework here at UWF, but I am also very excited to make the jump into some sort of GIS Analyst position. Academically, I received my B.S. in Environmental Design through the College of Architecture at the University of Oklahoma, but ultimately decided against obtaining my Master's degree in Architecture. My goal upon completion of this program is to go where the best opportunity arises [I would love to find myself in the environmental sector, something like species or land conservation], and just go with the flow from there! There are not too many GIS jobs located in Oklahoma City currently, so moving out of state after graduation is probably a very realistic option.

Some adjectives that would describe me are tenacious, determined, optimistic, and comical [I think I am hilarious, but my three teenage step-daughters do not seem to agree]. Good luck to you all this semester, and I look forward to working with you!

Here is the link to my story map: Wright Family Travel Log but I have also embedded it into this blog post below [the formatting is a little off, due to the constraints of the blog frame].