Monday, October 16, 2017

Lab 7: TINs and DEMs

Terrain Models

Lab 7 involved working with terrain models. Two types of continuous surface terrain models are digital elevation models (DEMs) and triangulated irregular networks (TINs).  A TIN is a vector based version of a 3D surface and is made up of triangles created from a density of elevation points containing elevation information. DEMs are a raster based network where the elevation data is stored within the cell values of equal size throughout the raster surface.

TIN networks can have a higher degree of resolution and are used for high precision modeling in smaller areas. DEMs are more widely available and work well over larger applications. TIN data models are unique in the fact that there elevation data is contained in elevation points which are used to determine the overall shape of the surface. The more points grouped together the greater the detail and accuracy. Generally in areas of abrupt change there are more elevation points.

In the example below the mass points are visible in black and connected by edges to create a triangle network. Within each triangle the elevation can change but the slope and aspect stays constant. In the upper portion of the image there are less points used due to the constant nature of the terrain. In the middle of the image there are a greater number of points as well as smaller more irregular triangles signify steeper elevation.

Triangulated Irregular Network

The example below is a DEM raster image of the same area. The surface contours are smoother but lack the detail of the TIN data model.
Digital Elevation Model

Tuesday, October 3, 2017

Lab 5

Vehicle Routing Problem

The lab 5 scenario utilized a Vehicle Routing Problem (VRP) analysis to assist a trucking company in creating optimized routes for a days worth of pickups.  The pickup locations were provided along with the distributions center, route zones, truck information, and the network dataset of highways for the United States. 

To get started the Pickup locations and parameters were were added to the Orders in the Network Analyst window. The Depot location and Route were also added to the Network Analyst window.  Next the Route Zones were selected to ensure route/order continuity. The VRP analysis was ran and the results were inspected. After inspection, there were12 orders not assigned to a route, and only 14 trucks were used when 22 are available. In addition several trucks had time violations that did not aling with the customer requested pickup time.

The original scenario excluded 12 orders because the routes were set so that only trucks assigned to a particular route zone could pick up orders that were within there assigned route zones. 

For the second VRP we were instructed to change the properties of two existing routes allowing two more trucks to pick up orders. Once the VRP was solved with the two new routes there were zero orders not assigned to a route and only one time violation. The total revenue was $33,625 and the total coast was $16, 991.

VRP analysis of optimized routes

Tuesday, September 26, 2017

Lab 4-Building Networks

Network Building

For lab 4 we were introduced to adding functionality into a network analysis. We were provided with a geodatabase of San Diego streets containing information about traffic patterns, traffic laws, and street restrictions.

A network data set had to be created inside the Network feature dataset containing the participating Streets feature class. The network dataset was completed and network elements that were connected were determined providing drawn street edges and street junctions.

The goal of the lab was to create a route through the city incorporating stops at 19 facilities. The new network of streets were added to ArcMap and the Network Analyst Extension was turned on. The Facilites shapefile were added as stops. A Network Analyst window is added to ArcMap and a new route is selected. Under the Analysis Settings tab parameters such as reordering stops and setting restrictions can be determined.

The first route contained no restrictions and took 105.5 minutes with a travel distance of 100,631 meters. 

The second route contained turn restrictions which increased the travel time to 228 minutes and the distance to 271,236 meters.

Thursday, September 14, 2017

Lab 3

Completeness of Road Networks

Lab 3 assignment was to determine the completeness of two different road networks in Jackson County, Oregon. Determining the length of road networks assumes a relative measure of completeness. The following data sets used for comparison was a TIGER 2000 road data set and a set of street centerlines for Jackson County. A polygon grid of Jackson County was also provided. 

The first analysis was to determine the total length of roads for the two road networks. The TIGER 2000 data set contained 509 km more road network then the street centerline dataset. By this comparison the TIGER 2000 data set is more complete.

The second analysis determined the total length of the two road networks within each of the polygon grids. In order to count only the  roads that fell within a polygon grid the clipping and intersecting tools were utilized to segment the roads per polygon grid. The final results were tallied determining how many of the grid polygons contained a more complete street centerline data set or TIGER 2000 roads dataset.

In addition a map was created summarizing the difference in total length between the two sets of roads per polygon grid using a percentage.

Geographic pattern in the differences in completeness for the two road networks.

Tuesday, September 5, 2017


Determining Quality of Road Networks

Lab 2 assignment used the National Standard for Spatial Data Accuracy (NSSDA) to determine the quality of road networks in Albuquerque NM. Several data data files were provided to determine the positional accuracy of the road networks. The two road networks were provided by the City of Albuquerque and a TeleAtlas product distributed by ESRI. In a addition we were provided with 6-inch 2006 orthographic images covering the study area. 

1.The provided data was displayed in ArcMap to visually determine the accuracy of the two road files in conjunction with the orthographic images.
2. A basic road network for each data set was created to generate all the junction points for the street layers.
3.Twenty test points were selected with five test points in each quadrant. All test points were collected at intersections and a new shapefile was created with these test points.The road shapefile provided by the City was used first because it appeared to have a better accuracy than the street shapefile provided by ESRI. A new shapefile was created for the street shapefile at the same junction points as the City road shapefile. An additional reference shapefile was created to digitize the true location of the test points.
4.   X and Y coordinates were added to the three new shapefiles.
5. The data was then calculated to determine the accuracy statistics using a horizontal accuracy statistic worksheet in Excel.

Test point locations throughout the city

Final Accuracy Statement

City Results:

positional   Tested 36.478 feet horizontal accuracy at a 95% confidence level.

positional   Not applicable

US Street Map Results:

positional   Tested 334.531 feet horizontal accuracy at a 95% confidence level.

positional   Not applicable

Sunday, September 3, 2017

Fundamentals of Spatial Data Quality

Lab 1- Calculating Metrics for Spatial Data Quality

Lab 1 involved determining the accuracy and results of a set of GPS waypoints as well as determining the Root Mean Square Error (RMSE) and the cumulative frequency distribution (CFD) of an error using a large dataset of waypoints.

In part A, we worked with the precision and accuracy of measurements. The difference between accuracy and precision is that accuracy measures how close an object is to the true value. Precision refers to how consistent an object is in its spacial placement.  

The below image represents the lab objective of creating a map layout with the distance of way points from the average location and buffers representing precision estimates.

In part B we were provided a large dataset and determined the RMSE and created a CFD chart. 

Below is a the CFD chart used to show the error distribution.

Friday, August 4, 2017

Module 11

Sharing Tools

The final module for GIS Programming 5103 ended with an introduction to sharing tools.

There are three ways that a tool can be shared.  

1. The first method is compressing the original folder structure using a ZIP file. This ZIP file can then be emailed or posted online.

2. The second method involves users having access to the same local network. The folder containing the file structure can be copied into the folder accessible to all users.

3. The third method is to publish the toolbox as a geoprocessing service using ArcGIS.

The final modules assignment involved working with a created toolbox and a completed script tool. The tool was edited and updated in ArcMap and the script was embedded into the tool and password protected. The toolbox was then submitted with no other files.

Below are screenshots of the final product. The tool dialog box was created from the provided script. The purpose of the script/tool is to create randomly placed points inside of a feature, then create buffers around those points.The results include the boundary feature, random points and buffers.
Tool dialog box
Results of running the tool dialog box.