Analyzing Attribute Data: Spatial Joins, Kernel Density, and Hotspots

Continuing on using some of the same airmen data from the last portfolio post, the objective of this project was to continue to develop spatial joining and analysis of data to develop a visual and statistical density and distribution.

Before beginning, once again, the correct projection must be used in order to maintain consistent and correct visual and distributable information.

The projection used for this project was Lambert Conformal Conic, a visually more correct map to the spherical shape of the Earth.


The continental U.S. was selected using the 'feature to feature' tool within the geoprocessing toolbox.

Next the state of Indiana was queried out using the select tool with 'intersect' being selected as well. A Kernel density of the airmen within Indiana was then ran. 

Above is the airmen data visualized by being overlain onto a correct projection of the state of Indiana using the Kernel density tool. Notice where the greatest densities lie, that being Indianapolis and the greater area, the area near Chicago, and near South Bend. 

Though a Kernel density can be a useful tool, it may lack an overall easy and simplified depiction of areas that are considered to be "hot spots", so with that in mind, next I will show you a hot spot map given the same airmen data entries for the state of Indiana.

First, a fishnet with grids at 10sq kilometers was made over the state using the clip feature.

Next the clip tool was used again and the zipcodes regions of the state via feature file, were made into a base layer.


In order to create a hotspot with base layers of the fishnet and zipcode boundaries, we need to do a joint count with the various airmen ratings and feature files selected to represent what these hotspots mean. Below are four maps with various different base layers and hotspots, along with a kernel density map.


Lastly, I made a map showing the population density and distribution of unmanned aviation pilots in the continental U.S. to show the scale of which unmanned pilots exist within the country.












Comments

Post a Comment

Popular posts from this blog

Defining More Features and Buffers

Image
 Features can also be defined measured with other forms of tools and highlighted with buffers. The pairwise buffer tool is helpful in creating zoning areas that can be visually appreciated. Below is a map that has highlighted roads of a flight area at Purdue Wildlife Area. The roads were manually configured using the draw polygon tool. That way, all roads can be accounted for with no error. The roads are still grass features, so the classification wizard querying did earlier would not be accurate, whereas the map above is. Now, I added a 20m buffer from all roads highlighted on the map. This gives you a visualization of the land that is further than 20 meters from all roads. Due, to the road grid pattern being tight, the amount of land is quite small. The buffer feature can also be used for other features. As seen below, I have put a buffer around mostly trees, but in total, all objects that exceed 10 meters in height. Another form of analysis can be found when the same buffers abo...
Read more

Pix4D Processing and Map Making with Ground Control Points

Image
 The usage of ground control points (GCPs) can be extremely useful when processing data and creating orthomosaic images of an area flown for data analysis. Below I have two instances of processing using GCPs, both of the Purdue Wildlife Area (PWA), and both used for independent and various analyses.  Once in Pix4D, the GCP data collected is set into a file location, and on a spreadsheet, the various figures collected from the GCP during the flight are brought and inserted into Pix4D to begin processing. Initial processing then is selected and let run completely. Make sure correct projection and GCP locations are used before processing is started. Processing varies in time length, but after completion a post processing evaluation should be done.  Next initial processing should be unselected and the next two steps should be run, that being point cloud mesh and lastly the creation of a DSM and orthomosaic model. Once processing is done, locate each GCP location and make sure...
Read more