Tuesday, October 17, 2017

GIS 5935: TINs and DEMs


This week we began to learn about DEM and TIN models in both ArcMap and ArcScene. This lab was very enjoyable and easy to work with. I personally really enjoyed the lab this week, feeling that I was gaining hands on experience with GIS applications that I would want to use in my future career. In this lab we focused on different elevations with TINS in ArcMap and ArcScene, comparing the differences between them. TINS can be really great for supplying detailed characteristics of elevation in small study areas. However for covering a greater area, DEMs are a better choice. DEMs contours are not as pointed as TINs (do to the triangles) and can bring more detail to the table for larger areas of analysis. 

The screenshot above shows how detailed TINs can be in smaller areas. In this TIN model I set up to have different angles of elevation, slope, and edges displayed differently to show the different textures in the elevation. By using applications like this you could locate ideal slopes for ski resorts, ideal location for observance towers, and a plethora of other applications.


Wednesday, October 11, 2017

GIS 5935: Location-Allocation Modeling


This week I gained more experience with Network Analysis by learning how to use the Location-Allocation tool under the Network Analyst Tools. This tool is helpful when trying to conduct an analysis for situations such as distribution centers and shipments. Location-Allocation allows you to analyze different scenarios for distribution, site location, and other attributes. You would ideally use this as a business manager to find the most ideal way to set up certain orders with certain distribution centers to reach the most customers. This of course can be used for other scenarios as well. In the ESRI exercise, I gained hands on experience running multiple location-allocation analysis to become more confident with this tool.

I then had to take what I learned from the ESRI tutorial and conduct my own location-allocation to find the most ideal distribution center for different market areas. This began by inputting my facilities (distribution centers) and demand points (customers) into a new location-allocation analysis, setting all 22 distribution centers to be included in the analysis. By conducting this analysis I was able to reassign market areas to their more ideal distribution center and provide a greater customer count. I then compared the market areas prior to the analysis to see which markets assigned to different distribution centers prior to the location-allocation. I could also back track by joining tables to find out the max number of customers for each distribution center and which market areas changed after the analysis.

All in all this lab was smooth sailing till Deliverable 9, which was a little confusing with joining all the tables. Luckily I learned a new skill within Python to make the joining and analysis much easier. As you can see in the map above, certain market areas were changed after the analysis to a more ideal distribution center.

Monday, October 2, 2017

GIS 5935: Vehicle Routing Problem




This week in GIS we worked on Vehicle Routing Problems. The goal was to be able to have all trucks cover all orders to increase customer revenue and service. Before I included two more trucks into the system (Trucks 15 and 16), I was unable to deliver 12 different orders. By increasing the number of trucks used, I was able to increase revenue by over 1000. In the image above you see the different routes that were covered. In the upper left hand corner you will see that unassigned orders is at 0, which before I introduced Truck 15 and 16 I had not route that would achieve the delivery of 12 orders!

All in all I really enjoyed this lab. It was very easy to follow and also very informative. There are a lot of things I took advantage of when getting upset with my package coming in a little later than the estimated time. There definitely is a lot more to a delivery business than I thought!






Wednesday, September 27, 2017

GIS 5935: Building Networks


This week in GIS 5935 we learned how to build functioning road networks and incorporate various "road blocks" that could change the route. The process began by creating a new data set through the SD gdb. Transportation file. This created the transportation junctions that allowed me to view the different speeds at certain times of the day. To build a route within the network data set, I then created a set route through the Network Analyst tool. I then calculated the set time to travel through the route by conducting an analysis of the route I created. Once the route was created I conducted several analysis where I included and excluded certain attributes. I noticed that when I increased the amount of attributes (i.e. adding restriction sites, removing or including one-way roads, etc), I noticed that my travel time would increase. My original time was around 94 minutes. By the time I added other items, it bumped up to 99 minutes.

Wednesday, September 20, 2017

GIS 5935: Completeness of Road Networks


This week in GIS 5935 we dove deeper into the quality of road networks by analyzing the completeness of road networks. In this lab, we had to determine the quality of two road networks in Jackson County, Oregon: Jackson County's Street Centerlines and TIGER's 2000 road network analysis. When calculating the sum length of the road networks, TIGER had more completeness with 509.4 km more road.

After calculating the total sum I then had to compare for each grid code. This lab was very challenging, coming across a lot of issues when it came to comparing the two roads. After quite a few trial and errors I decided to spatial join the road network to the grid that was supplied. I did this for both networks separately. After I joined them I then went into the attribute table to summarize the sum of their lengths in each grid code. This created separate tables that I could join together and create a field to calculate the differences per grid polygon.

As you can see in the picture, TIGER did have more completeness in many grids, but there are some grids where the Jackson's County dataset was more complete for grids. This could be tied to smaller bike lanes, downtown pathways, and other unique attributes to the county that the TIGER dataset would not have accounted for.

Wednesday, September 13, 2017

GIS 5935: Determining Quality of Road Networks



We have all been using a GPS and find that a road doesn't exist. Just this week my mother drove around every other road but the one that actually lead to my apartment complex. And the GPS was stating that she was already at the apartment. Obviously a major issue!

This week we learned about the spatial quality of Road Networks. In this lab we had to determine how accurate two different data sets were for the same area in New Mexico. I had to begin by creating a New Network Dataset for both street shapefiles to allow myself to see the junctions for both data sets. Afterwards I had to randomly select at least 20 points on the map to create reference test points to compare the data sets to for accuracy of the roads. I began by layering 20 orthosphotos throughout the study area (covering all corners and areas of the study site). I then zoomed to each orthos file and selected an ideal reference point.

Once I had set up my reference points I then selected the closest point of the data sets being compared to the reference point, creating a new shapefile with 22 points that were "closest" to the test points. I did this for both data sets. I then conducted an NSSDA Accuracy Assessment via Excel. The results showed that one data set (ABQ_Streets) was much more accurate compared to the other (StreetMapUSA).

I had a lot of trouble with this lab. Once again, lab instructions were very vague. It was to the point I could not distinguish what was even wanted as an outcome from the lab. After quite a few hours of struggling I was able to finish the lab with these my accuracy statements (which are posted below).



NSSDA Accuracy Statement for StreetMaps_Sample Spatial Data



Horizontal Positional Accuracy

Digitized features with areas with high amounts of perpendicular intersections. Selected 21 test points for reference and compared the Spatial Data Quality of StreetMaps data file to assigned reference points. Using the National Standard for Spatial Data Accuracy, the data set tested 384.6364429 feet horizontal accuracy at 95% confidence level.

Selected testing points were selected at random and spread across assigned study area. Horizontal Data (x, y) were attached to both the test points and Spatial Data points. StreetMaps_Sample data points nearest to reference test points were selected for conducting the NSSDA Accuracy Assessment.

Although tests of randomly selected points for comparison may show lower accuracy compared to ABQ_Streets Data Set between field and parcel map content, variations between boundary monumentation and legal descriptions can and do exist. Caution is necessary when using land boundary data shown. Contact Kerri Foote (via UWF GIS 5935 course) for more information.




NSSDA Accuracy Statement for ABQ_Streets_Sample Spatial Data



Horizontal Positional Accuracy
Digitized features with areas with high amounts of perpendicular intersections. Selected 21 test points for reference and compared the Spatial Data Quality ABQ_Streets data file to assigned reference points. Using the National Standard for Spatial Data Accuracy, the data set tested 219.6172706 feet horizontal accuracy at 95% confidence level.

Selected testing points were selected at random and spread across assigned study area. Horizontal Data (x, y) were attached to both the test points and Spatial Data points. ABQ_Streets_Sample data points nearest to reference test points were selected for conducting the NSSDA Accuracy Assessment.

Although tests of randomly selected points for comparison may show higher accuracy compared to StreetMaps_Sample Data Set between field and parcel map content, variations between boundary monumentation and legal descriptions can and do exist. Caution is necessary when using land boundary data shown. Contact Kerri Foote (via UWF GIS 5935 course) for more information.




Tuesday, September 5, 2017

GIS 5935 LAB 1 - Calculating Metrics for Spatial Data Quality


Well, this semester I started my second GIS course: Special Topics in Geographic Science. In Lab 1 we learned how to calculate metrics for spatial data quality, seeing how precise and accurate our measurements actually were. In this lab experiment we worked with 50 points tagged on the same GPS for the same exact coordinates. As you can see in the picture, the waypoints are all over the place. By calculating the average and a true reference point we were able to find out how accurate and precise even the average of all these numbers were.

In my lab I ended up with a high horizontal accuracy but very poor vertical accuracy. Having taken a break over the summer from GIS it was definitely a struggle in the beginning. I didn't realize how much I had forgotten in a few months. But it was like riding a bike, soon I began to pick up on more stuff like before. My only struggle in the end was the phrasing in the lab manual, which could be very confusing at times. All in all I enjoyed this lab. It was a nice refresher of old methods. And I also gained some new skills in GIS!