Arc Collector Part 1

Introduction:

A modern smart phone has an amazing amount of computing power, and so it does not make sense to buy another stand alone device that also includes a highly accurate GPS to collect data. For very basic data collection simply a phone and its own GPS unit can be used in conjunction with the ArcCollector software, and if a higher degree of spatial accuracy is required for a project, then a bluetooth GPS unit that is compatable with ArcCollector can be used. This setup with a phone is even more powerful when it is considered that it is not limited to only its local storage and unlimited amounts of data can be uploaded in the moment via a cellular data connection to an online server and even viewed from there in real time by others also collecting data in the field or those who would like to watch what is coming in from afar possibly to give real time feedback.

In this project, the goal was to get microclimate data throughout the entire campus. A tutorial was given on how to create a geodatabase, feature classes, and fields that could be populated in the field using ArcCollector, then after each student connected to the project that was created by professor Joseph Hupy on the university ESRI website server, we each logged into ArcCollector using our personal enterprise accounts and went out and collected data for our assigned zones. Two students were assigned to each zone. Data was collected on Wednesday, March 29th between 3:30 and 5:00pm. The fields that were populated were temperature, dew point, wind chill, wind direction, wind speed, time, and group number.

Figure 1

Figure 2

Methods:

A map with a file geodatabase with a point feature class and group zone polygon feature class was created in ArcMap then uploaded to ArcGIS Online was supplied to everyone in the class. Everyone in the class then went through the steps shown in subsequent writing below to add their own data points to the point feature class after being added to the group containing the project on ArcGIS online. The data was then copied to our own content page on our ArcGIS online accounts so that individual maps could be made.

Each group coordinated to get a collection of points that covered the entire area of the group zone. Each member collected about 20 points. The ArcCollector app was downloaded to the smartphone, and after logging into a personal enterprise account the shared group map was opened. New points were added by tapping on the boxed button shown in Figure 3. Then, the boxed microgrp feature class shown in Figure 4 was selected for editing. Finally, every field was selected and edited to include accurate information. The fields are shown in Figure 5.

Figure 3

Figure 4

Figure 5

The meanings of each field name are shown below:

GPR: Group (Zone) Number

TP: Temperature

DP: Dew Point

WC: Wind Chill

WS: Wind Speed

WD: Wind Direction

Notes: Any extra notes about a data point. Interfering factors such as vents were noted here mostly.

Time: This was the time recorded as military time without any symbols. For example 1:20 PM would be recorded as 1320.

All weather related fields were found using a Kestrel 3000 pocket weather meter. This device is pictured in Figure 6, and the symbols at the bottom of meter screen which indicated different readings were deciphered in the field using the manual found online. A screen shot of the symbol key is shown in Figure 7.

Figure 6

Figure 7

After collecting the data it was ready to map. The data, copied into the personal content folder in the ArcGIS Online interface, was clicked on, then the Open in ArcGIS Desktop button was clicked. With the data in ArcMap basic overview maps (shown in Figure 1 and Figure 2) were able to be made very easily. Next, to get data files to work with stored on the local network instead of in the cloud, the two layers (the group zones and the points with the weather data) were exported to a locally stored file geodatabase. The point feature class that was exported was then used with the IDW Interpolation tool to create heat maps for temperature, wind chill, and dew point fields, using their respective fields. These maps can be seen in Figures 10-12. All maps were classified with equal interval and five classes. Figure 8 shows the IDW Interpolation tool used to make the heat maps. The wind speed and direction map was made by simply going into the symbolization settings for the point feature class and changing the setting to graduated symbols set to the WS (wind speed) value, then clicking advanced, then rotation to choose the WD (wind direction) field for rotation, and finally making sure to choose an arrow symbol that was pointing down so that the direction was correct on the map. The symbology settings are shown in Figure 9.

Figure 8

Figure 9

Results and Discussion:

Figure 10

The temperature map above (Figure 10) shows temperatures slightly cooler on upper campus. This may have something to do with higher wind speeds on upper campus.  

Figure 11

The dew point map above (Figure 11) shows higher dew points at the creek and near the river. This is because of the high moisture content of the air as would be expected. Other points besides on the bridge along the river were not shown to have high dew points but this is because the IDW interpolation was doing the best it could with the point that it had, filling in areas between high dew points with high dew points. Lower moisture content in the air was detected on upper campus where there was no running water and dry air could blow in.

Figure 12

Figure 12, showing the wind speeds and directions, shows higher wind speeds on upper campus, on the bridge, and near parking lots as expected. It also shows more variation in wind directions around buildings and trees, as expected.

Several points can be made in regard to the quality of the data. One of these points is that new and estimated data had to be recorded into datapoints after returning to the lab because several people used the wrong setting to record dew point. This could have been corrected with keys to the symbols on the weather meter screen being used by everyone or better yet taped to the meters themselves. Because of this mistake, the dew point data should not be trusted. Another point that can be made is that some people incorrectly recorded times. These had to be fixed later for standardization. A final point is that many of the fields would not record more that two significant digits even if more were typed in. This creates a problem with many records being rounded down because the third and fourth significant digit was never considered by the application. For example, a 49.9 would be recorded as a 49 instead of a 50 or a 49.9.

Conclusion:

ArcCollector, properly set up with correct field attributes, with a fully charged battery, and can be an amazing too for amassing data. Athough there were troubles with the data in this exercise, it is okay because this is just an exercise and is for learning.

Sources:

http://www.nkhome.com/pdfs/K3000_Instructions_7.23.10_WEB.pdf