This is the second part of a post from earlier. You can read the first post here that explains the experiment in more detail:
I was finally able to have a successful run on our mountain comparison experiment...mostly.
The first run of our experiment was in August 2015. After the second day of the DRASTIC station being deployed on the mountain a pretty good storm rolled in and hammered the mountain. I posted a picture of the radar from the National Weather service on the previous post, it was pretty intense. The road up the mountain got pretty well washed out and my weather proofing on the housing for the station didn't withstand the elements. Water leaked in and fried the SD card and one of the pins on my micro controller. However it was not a total loss because all the sensors were still operational and I still had spare pins on the controller so I was able to do some rewiring on a couple of the sensors and be back up and running.
Once we got the station rebuilt and did some engineering on the housing we ran back up to the mountain to redeploy the station once again. This time I tried to engineer a better water proofing system for the micro-controller housing....a gallon sized zip-lock bag and a lot of tape. On our way to redeploy the station I set up my time lapse system that I built from another micro-controller that emits a infrared pulse that fires the shutter on my Nikon camera. I also included a couple of pictures of some wildlife we saw along the way...not a lot of things scare me but while walking down the trail (7,100 ft elev.) to set up the station a diamondback rattlesnake crossed my path...it made me a little jumpy for the rest of the trip. If you watch close towards the end of the time lapse part of the video when it is just about dark you can see our lights when we are coming down the mountain for part of the way.
For the most part the experiment was a success because I was able to gather data but I still had a few hiccups along the way. The first major hiccup was that the housing for the outside temperature was blown askew so my temperature readings cannot be trusted during the peak part of the days because the direct sunlight on the sensors gives increased temperature readings. The second hiccup was that halfway through the fourth day we had some pretty gusty winds and it shook my solar panel so hard that one of the wires came loose(nobody says science and engineering is easy especially for a hillbilly building this stuff in his shed in the backyard). However with a few light modifications to the original game plan I was able to still collect and analyse the data and come up with results that I have a high degree of confidence in.
The first modification that I made to the experiment was that I just used the data collected during the night and early morning hours when the sun wasn't shining directly on the station. I had a photoresister, which measures light, on the station so this was an easy modification to make within the analysis program that I wrote to comb through the data. I write all of my analysis programs in Python Scripting language. The second modification I had to make was to only analyse three days worth of data instead of 6. I ran some analysis on the data however and I felt like there was enough consistency in the data to say with high confidence that there was a clear trend in the data.
A quick recap of the experiment: we set up two identical weather stations to gather temperature data on the same line of latitude but at different elevations. One station was up on Virgin Mountain at 7,146 ft and the other was near my house at 1,401 for a difference of 5,745 ft of elevation gain and 21 miles between the two stations by way of the crow.
The finding of our experiment was that the temperature up on the mountain was consistently 24.69 degrees Fahrenheit higher than in Logandale. This difference was consistent between 12:00 AM to 8:59 AM with only a deviation of 3.108 degrees. This is about 0.004297 degrees per foot of elevation gain.
Although there were some hiccups along the way I am still going to chalk this project us as a success. Yes we had to make some modifications along the way, and we had to deploy the stations twice, but the data that we gathered was useful and we were able to pull some correlations out and find some patters in the data that help us to better understand the weather patterns within our area to a greater detail. We also learned a lot about how to better construct our stations so we can build better units for our next experiment.