Weather Station Data:

Underground Temperature and Frost Depth

(article continued from previous page)

I read somewhere that the ground freezes between one foot and three feet deep in the Chicagoland area. For some reason, this is amazing to me. It just doesn’t seem possible. So, I decided to prove it for myself.

My DIY weather station measures the battery voltage, the solar panel voltage, and six sensors. The sensors can be anything, but presently they are six thermistors to measure temperature. The weather station’s temperature sensors are usually configured like this:

1. Touching the soil under a rock
2. Inside the weather station box
3. Outside the weather station box (53 cm above ground)
4. A little higher up (116 cm above ground)
5. Even higher up (187 cm above ground)
6. Highest of them all (274 cm above ground)

But, for this experiment, the sensors are arranged in the opposite direction:

1. Inside the weather station box
2. Air temperature resting on grass just above the ground
3. 25 centimeters (about 10 inches) under ground
4. 50 centimeters (about 20 inches) under ground
5. 75 centimeters (about 30 inches) under ground
6. 100 centimeters (about 39 inches) under ground

Below is a graph of temperatures for September 2010. The inside of the clear project box heats up in the sun and cools off at night. The air temperature near the ground doesn’t get quite as cold ❶ as the inside of the project box about a foot off the ground. This is because the thermal mass of the ground moderates the temperature. This same phenomenon explains why a desert is colder than a jungle at night: less mass (wet vegetation) to moderate swings in temperature.

Air and soil temperature Chicago IL Sept 2010.

On September 7, the soil between 25 and 100 cm in depth is approximately 67 °F. Two weeks later, the temperature has gradually fallen to 63 ° F.

A warm spell ❷ raises the temperature of the soil for a few days. Notice the soil nearest the surface fluctuates the most in response to air temperature.

Jumping ahead to October, let’s add in the solar and battery information to the graph. It’s interesting that sunlight does not cause temperature fluctuations in the soil at 25 cm and deeper.

Soil temperature and light activity Chicago IL 2010.

There appears to be some error ❶ in the data for the deepest soil. I suspect it was caused by water condensation on the board or connector. Or, it might have been a passing gopher.

The spikes in the solar panel voltage readings ❷ are my wife letting the dog out. Seriously. This weather station is located near one of the backyard floodlights. My wife turns on the lights before letting the dog out at night. You can see that some nights she’s busy and leaves the lights on (Nov 19) for long periods.

This next graph is a month worth of data at the end of fall. There are another couple of data spikes early on, in the 50 cm soil readings.

Air and soil temperature Chicago IL Nov Dec 2010.

Interestingly, the soil temperature has really spread out based on depth. There is now a ten degree Fahrenheit difference from the deepest (46.3 °F) to the shallowest (36.3 °F) soil depth.

The top couple of inches of soil were likely frozen in mid-December. However, we are now at a fairly high 52 °F at the end of December. (That doesn’t mean I can’t have any Christmas hot chocolate, though.) I probably won’t get any records that indicate significant frost depth in soil until late January.

Ahh, but it is toasty warm inside -- isn’t it? Let’s see what the weather station records when I bring it inside to my office and the kitchen.