Arc Collector Part II: Bumper Stickers

Introduction

           This lab required a more sophisticated understanding of Arc Collector then the previous exercise, including how to set up the software and then collect data. This lab was much more elaborate than the introductory Arc Collector assignment and required the class to come up with unique study questions that could be answered by collected data in the form of points with attributes, with at least one of the attributes involving a numeric field. For this lab the connection between bumper stickers and automobiles will be explored. More specifically, the data collected sought to see if there was a relationship between the year of a car and the number of bumper stickers, as well as the type of bumper sticker. Additionally, the idea that a type of vehicle being a car, SUV or truck might determine the type of bumper stickers on the vehicle was explored.

        The study area for the data collection took place on the University of Wisconsin Eau Claire, to the south of Philips and Davies hall in the Davies parking lot. In total 122 vehicles and there bumper stickers or lack there of were recorded. This was done because it was a convenient location that was a wide age range, as the university was a wide spread of ages. Replicating this data collection in different parking lots around town would most certainly show different results. Data collection was conducted on April 25th from 10 am until 12:30 am. Setting up the project correctly in Arc Catalog is the first step to successful and accurate data collection. If a step is skipped or a value is displayed incorrectly, data integrity could be hindered.

Figure 1

       
         Figure 1 above shows the how the data points looked when they were collected in the Arc Collector map. They were not even remotely accurate as the GPS was likely having troubles getting a good signal due to the large hill located to the south and the tall buildings to the north. Each data point was collected directly behind the cars in each stall to ensure that all the stickers of the rear of the car were clearly visible. Once the data was opened in ArcGIS online the points were corrected to where they were actually taken.

Figure 2

            Figure 2 above shows the corrected data collection points. The points were corrected by simply clicking and dragging in ArcGIS online.

Figure 3

        Figure 3 above shows the study area in greater detail as highlighted by the red box. The Davies center and Philips hall are shown to the north of the data collection area. The specific area was the middle two rows of parking in the Davies parking lot, the outside rows were not done due to time constraints and the time that took to collect nearly 500 attribute values for the 122 data points. 

Methods

         The process of setting up Arc Collector begins with setting up a geodatabase, and then selecting domains that are applicable to the data collection as well as setting up the feature class. In total there were four domains created in the database properties. The domains were type of bumper sticker, age of vehicle, type of vehicle and number of bumper stickers. The type of bumper sticker had 6 different codes, religious. sports, outdoors, political, other or none. For the age of the vehicle a short integer was selected, then for type of vehicle, car, SUV and truck were the three options. Finally the number of bumper stickers was set up, with values from 0 to 15, though no car displayed more than 7 bumper stickers during data collection. Next, the sign in to ArcGIS online was done using the UWEC enterprise account. The set up was shared in the form of a map, which made it accessible on the mobile devices. Tags were added so the project could be found by others, the tags being Geog 336 and bumper stickers. After this was set up, the project was published. From here the data can then be collected using the mobile app.

          The question that was sought to be answered by doing this data collection was whether or not the age and type of vehicle have a relationship to the number and type of bumper stickers that one has on there car. To deploy a data collection for this question, Arc Collector was used, which is a very user friendly, real time way to collect and upload data. Figures 4-9 are screenshots of the program in action that display how each of the different attributes appeared when being collected.

 

Figure 4

Figure 5

Figure 7

Figure 6

     Figure 4 to the right shows what the initial screen appeared as each time a car was selected for data collection. Every car that was in the Davies parking lot in the middle section was collected, none were skipped. The reason for some holes in the data is because there were no cars in the parking spots. Figure 5 to the right shows what the screen displaying the type of vehicle looked like. It states that the vehicle is either a car, truck of SUV. Figure 6 to the left shows the attribute displaying the age of the vehicle. As stated, the value was collected according to the number of years old the vehicle appeared. In order to get the actually age of the vehicle the owners would most likely have to be present of an extensive amount of research would have to be done about the cars. The reason a rough estimated was used was because of time constrains as well as the fact that it was not highly important to answer the question. Figure 7 to the right shows the attribute with the type of sticker. Either religious, sports, political, outdoors, or one that does not fall under any of the predefined categories. A car that recorded a no value were cars that did not have any type of stickers displayed.

Figure 9 above shows the corrected data collection points on the device they were collected on which was an iPhone 6. The quality is rather good and the points show up nicely.

Results

         Below are a series of maps that show the attributes that were collected throughout this exercise. Figure 4 below shows the estimate of the ages of the vehicles that were in the Davies parking lot at the time of the data collection. The was this is set up is that the ages 2-5 mean the vehicle is within 2-5 years old, while 6-8 means the vehicle is 6-8 years old and the same for 9-16. These were just ball park estimates which factored in a large knowledge of cars. It is important to note that if the data collection was done again, this field would be fixed to display the values in years, for example 2000-2005. An error was made when setting up the project and it was not realized until data collection began. Though this was not ideal it still gave a good idea of what the age distribution of the cars in the Davies parking lot were.

Figure 8

         Figure 5 below displays the type of vehicle that were recorded, being either a car, SUV or truck. Only five of the vehicles of the 122 in the data were trucks. This number is very low, though it is to be expected as gas is not cheap and a good percentage of the people parking in this lot are likely college age students, though there are exceptions. 84 of the 122 vehicles were cars, while 33 were SUV's and as stated above only five were trucks. This makes sense that a large majority of the vehicles were cars because they are cheaper and more fuel efficient.

Figure 9

          Figure 6 below is a map that displays the types of bumper stickers that were recorded. Surprisingly only 30 of the vehicles did not have bumper stickers, 7 were outdoor themed, 11 political, 6 religious, and 15 were sports related. The largest number of vehicles fell in the category of other, meaning that there was not a pre designated category that they fell under. 53 cars had stickers in the other category, if the data collection was re structure for use another time, some of the other categories would be added to get a more accurate reading on what people use for bumper stickers.

Figure 10

        Figure 7 below shows the number of bumper stickers that each car had. 50 of the vehicles only had 1 bumper sticker, which is just over just over 40 percent of the vehicles. This is likely because not many people want to degrade the look of there vehicle, but still want to show what there hobbies or views are, hence why one sticker was the most popular. 24 vehicles had 2 stickers, while 10 had 3.

Figure 11

     


          Throughout this lab there was a number of different things that were learned and that would be implemented if the exercise was done again in greater detail. After completion of the lab it was clear that the categories set up for collection were not the most popular bumper stickers. First and foremost a large majority of the vehicles that only had one bumper sticker were sticker of the dealership the automobile was purchased at. This was not thought about prior to completion and it should definitely be a category of bumper sticker if the lab was replicated. Also, a family category would be added that would include stickers such as "baby on board," of which there was more than five, as well as stick figure families, of which there was also a considerable amount. With a few simple corrections this lab could have shown more of the relationship with the fact that many of the newer cars only had the sticker showing where the car was bought. An example of this can be seen below in figure 14, this car had no additional bumper stickers other that the one that advertises the place it was bought.

Figure 12

Conclusion

        In conclusion, there is no notable association between the age or type of a vehicle and the number or type of bumper stickers on the vehicle. The number of bumper stickers varied from none to seven and did not show an association to the age or type of vehicle. In the beginning of the lab it was thought that older vehicles would have more bumper stickers under the premise that owners of new cars would not want to put bumper stickers on a new car. It was also thought that vehicles such as trucks and SUV's would have more outdoor bumper stickers, this was also disproved in this lab, though the sample size was small and there can not be a definitive relationship established for anywhere but the parking lot of Davies during data collection.

Arc Collector Weather Data Collection

Introduction

          The purpose of this lab is to become familiar with gathering geospatial data on a mobile device using Arc Collector. The purpose of using Collector is the fact that smartphones have more capabilities than most GPS units, that being said there is no reason to not be familiar with collecting data using it. Additionally smartphones can access online data in real time and allow the user to gather on the fly and upload data instantly. Additionally photos of the study areas can be uploaded instantly, this helps give the data viewers an idea of where the data was collected.
           The study area for this exercise was on the University of Wisconsin Eau Claire. The class was split up into seven groups and each group was assigned as area, a map showing the seven different data collection areas is displayed below in figure 1. The study area that will be discussed in this report is area seven, which was the study area that was assigned by Professor Joseph Hupy. Each group member was to collect around twenty data points and the group members were instructed to split the zones up into smaller areas to ensure that data was collected evenly throughout the assigned area. Data collected for this lab was on Wednesday, March 29th between 3:30 and 5:00pm. The attributes that were collected in this lab were temperature, dew point, wind chill, wind direction, wind speed, time, and group number. There was also a notes section available, though there were a few issues experienced with that and they will be discussed later in the results section.

Figure 1

Methods

          Arc Collector is a very straight forward method of data collection. For this exercise the data collection table was set up previously by Professor Joseph Hupy and the data recorded was entered and then submitted online where all other groups could see the data points and data recorded at those points. The attributes were selected by professor Hupy and that was done by creating feature class within the geodatabase.

Figure 2

          Figure 2 to the left shows the pocket weather meter that was used to record the weather data for this lab. This instrument has many uses, as it collects temperature, wind chill, dew point, wind speed and it also had other features that were not recorded. When in the field the points for weather collection were random, the only stipulation was that the data collection points were spread out within reason. While collecting data the points were instantly displayed after submitting the points, also as the data collection progressed the more points from other groups were also visible. Figure 3 to the bottom left shows what the screen looked like after opening the app, from there the white plus was clicked on in the top center. From there Figure 4 below shows what the data collection screen looked like. The data being collected was abbreviated, GRP is group number, TP is temperature, DP is dew point, WC is wind chill, WS is wind speed, WD is wind direction, then notes and time are the two final attributes. The white symbol of a camera in the upper right was used to attach photos.


       

         As far as units go, the temperature was recorded in degrees Fahrenheit, as well as the dew point and wind chill. Wind speed was recorded in miles per hour (MPH) and the wind direction was recorded in degrees and entered as a number from 0-360. The time was entered in military time from 0-2400 and it was important that there was no semicolon between the numbers in the time.
        From here a number of maps were made that displayed the data that was collected. ArcMap 10.4.1 was used. The temperature, wind chill, and dew point maps were all interpolated using the IDW tool to give a better idea of how those variables change with elevation.

Figure 3

Figure 4

Results


           Below in figure 5 is a map that shows points at which the weather data was collected. With the fact in mind that there was no specific place to collect points, the data points are reasonably evenly distributed. As discussed in the methods section, each student was to collect data at twenty different locations.

Figure 5

         Figure 6 below is a map that shows the points collected by each group, very similarly to figure 5, though figure 6 shows how close each group got to the other groups areas. For example, the data collectors in group seven had a point collected that was very near group six, this could be explained from a possible error in GPS because of the hill, or the data collector was slightly off. Group one did a nice job of covering the northern side of there area, they made sure to record data at both the northeast and northwest corners. Group five should have covered a much larger area in the southern third of there study area. At first glance it appears the group only covered about 60% of there assigned area. Certain groups such as group four were very limited on where there data points could be recorded as there is a lot of buildings in there study area.

Figure 6

        Figure 7 below is map of the wind chill that was interpolated using the IDW tool in ArcMap. The study area on the west of the map has a high wind chill, this can be explained because that location is the top of the hill, therefore the wind hits the exposed area much stronger. The lowest wind chill recorded was 45.0592 degrees and the highest recorded was 62.892 degrees. The area in the southeast portion of the map had the warmest wind chills, likely because the wind was out of the northeast and it was blocked by Philips hall which is to the northeast of the data collection there. 

Figure 7

          Figure 8 is a map that shows the direction of the wind along with the speed in miles per hour. The wind direction is shown by the way the arrows are pointing. The bridge is a good place to look to get an accurate wind direction, all of the arrows are point to the southwest, which means the wind direction was from the northeast. Variations in the wind direction on campus can be accounted for because of the buildings. Wind tunnels can be created as the wind hits the side of buildings and funnels ways that it was not actually blowing.

Figure 8

          Figure 9 below shows the dew points collected throughout the lab. The lowest dew points recorded were 30.0257 degrees and the highest recorded were 57.9726. Almost all of the high dew points were recorded on lower campus in lower elevation areas. In comparison the east side of the study areas were almost exclusively done on upper campus which is a much higher elevation.

Figure 9

         Figure 10 below is a map displaying the temperatures that were recorded. There are some warm spots on the north side of the Chippewa river, in contrast there are a few cold spots in the south portion of the study area near the towers halls.

Figure 10

       

           It would be interesting and beneficial to complete this exercise again and to collect all the data with one or two data collectors in order to ensure that the data was all collected the same way. Much of the variation in wind speeds and temperatures could be because the collectors did not let the units get an accurate reading.




Conclusion

           In conclusion, Arc Collector was a very useful method of data collection. It is very efficient, easy to use and allows a number of users to collect data and report it in real time. This lab demonstrated that a smartphone or tablet can indeed do all of the operations that a GPS can when used correctly. It was very useful to see the data collected displayed in real time. One issue that would need to be resolved if this was a data collection for company is making sure that each data collector is clear on how to collect the data. To clarify, some students entered time the wrong way, also some were not familiar with recording wind directions and finally not everyone held the meter long enough for it to get an accurate reading on wind speed. As professor Hupy discussed before data collection, as our supervisor he could see where all the students were during class time, assuring that no one was skipping out. This is applicable in the real world because an employer could track down time and ensure that employees stay on track, which saves time and money.