Microclimate Survey with Arc Collector

Introduction

This week we would be attempting to do the microclimate survey of campus, but with ArcCollector rather than using a Juno Unit. ArcCollector does have some positive and negatives. Firstly, ArcCollector can be launched on anyone's cell phone, which makes it highly usable and deplorable in the field, with out additional costs of equipment if one already has a smart phone. The down side to using ArcCollector on a cell phone is that the accuracy of a cell phones GPS is definitely not what you would find in a survey grade GPS unit. Really the option of using ArcCollector on a smart phone really comes down to an organizations needs, if that organization needs the ability to have many people collect data points in broad areas, and that data does not need to be in the range of accuracy or precision of a field grade GPS then it can definitely be used to collect a lot of data points, such as we did. 

A microclimate survey is exactly what it sounds like, a survey of various climate conditions in a specific area. In this case our microclimate survey would be taking place on the UWEC campus. Our task was to survey points on the campus and record, Temperature, Wind Chill, Dew Point, Wind Speed, Cardinal wind direction, and relative humidity.

After all of these points are captured by everyone in the class they can be aggregated, and then added to one map, such that we can than make various theme layers describing the data in anyway an individual user would choose to do so.

Fig 1. The UWEC Campus dived into zones, in which we would be conducting our Microclimate Survey. The red lines indicate individual zone boundaries, which correspond to various groups. Data points have been aggregated with the merge tool , see below. Background Imagery from ESRI online.

Methods

The first thing we had to do was deploy a prefabricated geodatabase, which was made by Dr. Hupy for our microclimate survey. This evolved bringing the geodatabse into ArcMap and putting the zone layer on the map. Next, we had to sign into our ESRI online enterprise accounts, and publish and share our basemaps on ArcGIS online.

After this was done, we were left with the seemingly simple task of downloading ArcCollector to our smarphones, which actually turned out to be a bit of an issue. I was able to download ArcCollector successfully, but unable to log into ArcGIS online via the UWEC enterprise login, due to an unknown error. It took a while to figure out, but Dr. Hupy finally narrowed it down, and was able to get me logged in. Then it was off to the field!

Once out on campus, we were to go to the zones we had during the first microclimate survey (which did not get a blog post due to technical difficulties last week with the Juno Units). Data was collected and brought back by everyone in the class. This data was uploaded to our geography server so everyone was able to access it and actually turned out to be a substantial amount of data points for our small class size.

Fig 2. Microclimate data points from the entire class uploaded to geography server and then downloaded and put into ArcMap. Note the table of contents on the right hand side of the figure, which is the data from individual class members which exists all in their own feature classes. Green dots represent individual data points, while red lines represent individual zones.

Now came the tricky part. Once everyone's data was downloaded into ArcMap, all of the data points had a different range of values which made it impossible to represent the data in any meaningful way at all. Displaying temperature for example with class breaks and a color scheme meant that while one person had recorded temperature to be between 68-74 degrees for all of their data points another person could have recorded their temperature points to be between 64-78 degrees. This may not seem like a big detail but it turned out to be HUGE. When color coding these temperature between everyone's data points multiple temperatures had the same color, you could not tell which temperatures were hot or cold, due to the fact that none of them had the same scale, but different scales and colors based on their ranges.  

In order to actually utilize the data in a way in which we can map any data we have to normalize the data such that the data will all be on one continuous scale. How do we do this? The answer seemed simple, "how do you combine everyone's data into one large set, with one large table, normalizing everyone's data?". But I had absolutely no idea how to do this, so I went to "Help" and started poking around. Eventually I learned about the merge tool (Fig 3) and this is what Arc Help had to say, "Use this tool to combine datasets from multiple sources into a new, single output dataset. All input datasets must be of the same type (that is, several point freature classes can be merged, or several tables can be merged, but a line feature class cannot be merged with a polygon feature class)".

Fig 3. Desktop Help for ArcMap, image shows the function of the merge tool and how feature layers and tables are combined into a single layer. 

Now that all of the data was combined I could begin to map the data.

Results/Discussion

Once all of the data was merged together maps were created for every variable that was taken in the survey;

Fig 4.Data on microclimate dew point on the UWEC campus.

 Dew point seems to be highhy variable across the microclimate survey with multiple points across all ends of the spectrum recorded. Whats more interesting is that on the North side of the river we seem to have a general trend of higher temperatures while on the South side of the river we have considerably lower temperatures. 

Fig 5. Data of microclimate temperature on UWEC campus.

Depending on where the point was taken and the amount of black top in the area the temperature across the UWEC campus is relatively grouped together, between 63-74 degrees F. It is interesting to note that the areas with the lowest temperature are near the river in forested areas, and in Putnum Park  behind the school. Potentially shade or other factors are contributing to these variations in temperature. The one point that is captured at below 63 degrees F in green, is more than likely in error, or we would need more information to explain why that one point is so 10 degrees cooler than the rest of campus before accepting that point. 

Fig 6. Data on microclimate windchill across the UWEC campus.

Wind chill would be the most expected data value with to have the least variation, but whats interesting in the microclimate survey is that the variation seems to only be by the Chippewa River and the rest of the campus is very uniform except for the one blue dot (0-15 degrees F) and this is more than likely human error and could be thrown out.

Fig 7. Data on Wind direction the UWEC campus.

 Similar to wind speed below, the data on wind direction is expected to be variable, depending on when the data was captured. In some places there exists a general trend while other places go against that trend. It is interesting to see that wind is so variable on campus.

Fig 8. Wind speed data across the UWEC campus.

This map is interesting mainly because of the time in which the data was captured, as the wind speed changes when the depending on the time of data capture, it would be the measure with the most expected variation from point to point.

Conclusion

Again we have a sampling technique that did capture and paint an interesting picture of the UWEC, and again which can be scaled up to a variety of levels and uses. ArcCollecter is definitely easier to use and deploy in the field with a smartphone than by using a antiquated Juno unit. There were some technical details to overcome in the beginning but that was more to do with the enterprise log in system than anything else. I also now have a through understanding of the value of creating and implementing domains when attempting to gather data, it really makes me appreciate the effort others go to when I use data in ArcGIS.