Microclimate? Interesting Temperature reading from Air Unit

Last week, during our first batch of cooler weather, I got some very cold readings from my Air unit. My unit and a relatively close neighbor’s unit ( 3 miles) were reading a few degrees lower than other neighbors. We were reading around 29F when neighboring units and the nations weather service station about a mile away were closer to 38F. Some of the units were within a few hundred meters of mine. I also noticed that the units that were close to the NWS station readings were all newer tempest units. The two units that read low were Air units. This disparity was evident even during non daylight hours.

The disparity is not evident right now.

My Station is Excitable Toy on map below:

Hi @peter.vanamson, thanks for the report! We are not aware of any issues that could produce anomalously cold readings from AIR devices so wanted to dig into this a bit more. I pulled data for yours and some of the surrounding devices and indeed see the cooler readings from the AIRs that morning. AIR 1 is yours, AIR 2 is shbc to your southwest, Tempest 1 is Kelley Ln and Tempest 2 is Pond View Lane both just to your northeast.

The fact that both AIRs agree on those low temps makes a sensor issue seem unlikely, and I am actually wondering if you are observing a true meteorological phenomena. Here’s why: on the morning of the 31st, there was a very steep inversion in place in a very shallow layer right at the surface. The Chatham radiosonde from 8 am Saturday morning (12 UTC Oct31) shows this nicely:

This means that temperatures would be colder lower to the ground and then increase with height. The elevations of each station help tell the story:

AIR 1: 5.18 m
AIR 2: 4.33 m

Tempest 1: 7.62 m
Tempest 2: 11.35 m

CHH Airport: 14 m

So, my hypothesis is that the dense network of stations near you which are at different elevations are sampling the low-level inversion in place that morning, and the two AIRs which happen to be at lower elevations were in the coldest air while it was comparatively balmy for your neighbors’ Tempests up the hill! Does this seem plausible based on your knowledge of the area?

Thanks again,

-Tony McGee


Hi Tony;

Thank you for the note and work to bring additional effort to bear. I think its an interesting theory. The differential between my unit and the airport is consistent with your theory for sure. In terms of elevation, shbs, Kelly Ln and Woodcarver’s are quite low. Pretty much right at sea level. My station is probably just a bit higher than these two. So the difference between these two and mine would not be consistent with your theory. The area in general is known for having really tiny microclimate variation particularly when it comes to visible fog.

In the chart above what am I looking at? Assuming, X axis is temp in Fahrenheit, not sure what Y is or the green, red and blue plots.

It is a complicated plot showing all of the weather balloon data from the surface to near the top of the atmosphere but the important bit here is the red line which is temperature. In the lower atmosphere, this line would usually start towards the right at the bottom of the red trace and then move to the left as you go up (Y is height), indicating falling temperatures. However notice at the very very bottom, the red trace actually goes from left to right which is opposite of “normal”. That is called an inversion, where temperatures are increasing with height. It is more common to see when it is cold and clear.

As for the elevations, I was not expecting it but Pond View Lane is set to 11.35 m above sea level, now it is possible that a user is including height above ground in that. Meaning even if the elevation isn’t truly that high maybe the Tempest is on a taller pole that is getting it up higher?


Hi; Tony;

I have been noticing a now relatively consistent difference between my air unit and the tempest units nearby. For instance, right now I am at 28.2 and a nearby station is at 31.5. The nearby airport and a half dozen stations visible on weather underground are in the 32 - 33 F range. Could I have a faulty device?



Hi Tony;

I looked at the comparative graphs of my unit vs. some local units and now it seems to be right on the screws with them again. So, in short, I may have sent up a red flag by mistake. It really must be that my house is in a local cold spot. In addition, My unit is exceptionally well shaded so there is really very little chance of solar heat gain messing up my readings. Maybe some of the other local units suffer from that problem.

Chatham Wind and Time (the closest local unit):

My unit:

all the best,