Looking for something fun to do on a Friday night? Maybe a beer with friends after work, grab a bite to eat, catch a show or a late night hang at the bar… movie night with the kiddos…¬†how about driving back and forth across Durham County collecting temperature data?! It’s all for the fun of science.

But I digress… Thursday night was build-out. The idea is to get an external temperature and relative humidity sensor attached to the car at approximately 2 m (6 ft) height and out of the slip-stream of the car so I can drive the sensor around measuring the ambient environment. The data will be correlated to GPS data collected at the same time so I can start to unravel the correlation of temperature with urban form.

The design required an old 2-by-4, in this case harvested from an old renovation project around the house, and a two foot section of 4″ PVC pipe, also left over from a house project. Using a circular saw and drill, a seat for the external HOBO sensor and a notch for the chord were created in the end of the 2-by-4. The sensor was installed in its home and the chord stapled down to the beam using chord staples. Of course we have a chord stapler. Who doesn’t?


The PVC pipe was installed over the sensor after a bit of electrical tape was wrapped around the sensor for good measure to make sure it didn’t move in the wind. Then I drilled out two sets of 4 holes in line with the location of the roof rack bars so zip ties could be used to attach the rig. In addition to the zip ties, I ended up using a ratchet strap to keep the whole thing in place, again, just for good measure. Once installed on the car, the sensor body drops through the sun roof so I can launch and download data inside the car.

The set up looks a little home-made, and it is. I was thinking I needed to create a hashtag: #NotAGoogleCar… but it’s probably pretty obvious.


With the mount ready, and the weather window open, it was time to get transecting! (Yes, I made a verb out of transect.) Starting at 6 am on Friday morning, I drove the predefined route every 4 hours (6 am, 10 am, 2 pm, 6 pm, 10 pm, 2 am). Each run to Durham and back took an average of 1:45, and was 46 miles. That totals up to some 280 miles and about 10 hours of driving. On each run, I collected temperature, relative humidity and GPS data every second. That’s a lot of data! That’s a lot of NPR.


With raw data in hand, its time to get processing, but first I have to rest up for the next round of transecting… possibly as soon as Monday. Because the fun doesn’t stop when science walks in…!