Taking our Temperature and Thermometers

Much is said these days about temperatures, but very little is told about how those temperatures are measured. There are various types of thermometers used at established weather stations, each uses slightly different instrumentation.  Here I’ll describe the different types of thermometers used to take our temperature and how that instrumentation has changed over the years.

(Additional content was added to this post on July 31st, 2024)

A Brief History of Taking our Temperature

Daniel Gabriel Fahrenheit invented the first Liquid-in-Glass (LIG) thermometer in 1714. He started out using alcohol and then switched to mercury filled thermometers. An alcohol thermometer is preferred, especially in colder climates, because mercury freezes at minus 39 degrees.

Around 1850 the U.S. Weather Bureau began taking daily readings of temperature using Liquid-in-Glass thermometers at official weather stations. Most of those stations were in the eastern half of the United States at that time.

By the late 1870’s these weather stations began using two LIG thermometers, one that registered a maximum temperature and one that registered a minimum temperature. These thermometers were read once a day by a trained weather observer, to get a maximum and minimum temperature for the previous 24-hours and were reset manually each day.

By the mid-1880’s the standard was for these thermometers to be housed in a Stevenson Screen, later known as a Cotton Regional Shelter. These shelters were essentially large, white, louvered wooden boxes that protected the thermometers from wind, rain, and direct solar radiation. The shelters were also designed to allow good ventilation to negate any artificial overheating inside the shelter.

Old type Cotton Regional Shelter
Inside the shelter, max/min thermometers and a thermograph as back-up.

Fast forward another 100 years to the mid-1980’s, when the National Weather Service began changing the existing instrumentation over to an electronic thermometer. These could log temperature readings automatically, eliminating the daily routine of reading and resetting the thermometers. This provided a more consistent recording of the data, with fewer missed days in the daily weather records.

By the late 1990’s, about two-thirds of all official weather observing stations across the U.S. had been changed over to electronic thermometers. Today, almost 100-percent of the official thermometers maintained by the National Weather Service are electronic.

Electronic Thermometers

The modern Maximum-Minimum Temperature Sensor (MMTS) used today is an electronic RTD type thermometer, or Resistance Temperature Detector, which measures electrical resistance in a coiled metal wire.  As temperatures rise or fall, electrical resistance rises or falls, in a linear fashion.

The RTD-type thermometers used at official weather observing stations today have proven to be as precise as the LIG-types, both are accurate to within about 0.5 degrees F. The biggest difference is that RTD thermometers have a much faster response time than LIG thermometers. Which means, brief spikes in temperature, high or low, will be recorded by an RTD. An LIG thermometer cannot record those spikes, as it takes longer for the liquid to react.

RTD type electronic

RTD thermometers are more sophisticated, more accurate, and more stable than simple thermistor-type thermometers, like you would find in a home weather station or your car. Many of the remote weather stations in the mountains utilize a thermistor-type thermometer, which in many cases are not shielded properly and susceptible to erroneous readings, like your car’s thermometer is.

The new MMTS thermometers are not housed in the old type of shelter, but rather inside a beehive shaped, vented canister. Which, I have discovered, the size of the beehive canister is not standard from one weather station to another. That creates an issue, which I will expound upon shortly.

Jackson, WY Instruments

The earliest official weather readings here in Jackson, WY began in 1899, although only for about a month. By 1916, the primary weather station in town, located behind the Forest Service office on North Cache Street, began taking more regular readings. But it wasn’t until about 1950 that records became more consistent year-round. Back then, and until the early 2000’s, LIG-type thermometers were used.

In March of 2002 the Jackson Climate Station switched over to an electronic MMTS thermometer.  In October of 2012 the station moved a few hundred yards to the North, to behind the Visitor Center.

Because Jackson’s weather station has always bordered the natural setting of the Elk Refuge and has been in roughly the same location for over 100 years, I believe the Jackson Climate Station has one of the best long-term temperature records available anywhere in the country.

A weather station that has remained in a more natural setting, away from concrete, asphalt pavement, and large buildings, provides more realistic historical temperature readings. Jackson’s station is unlike most weather stations around the U.S., which have seen big cities grow up around them.

Jackson vs. Moose Wyoming Instruments

I have resided in the Jackson Hole valley for over 40 years now and in that time, I have observed and researched our weather and climate, about as extensively as anyone could. Over the years I have seen some changes, not only in the weather, but also in the instrumentation used to measure temperatures.

While analyzing monthly weather data from the Jackson Climate Station, located on the north end of town with the Moose Climate Station in Grand Teton National Park, I discovered that a change in equipment yielded different results. Somehow Moose had become warmer than Jackson, on certain days, which historically has never been the case. Allow me to explain why.

The Stations

Both Jackson and Moose are Cooperative (COOP) weather stations established and maintained by the National Weather Service. Each day, trained weather observers take readings from these weather stations and submit the data to the National Weather Service, which is then transmitted to the National Climatic Data Center.

Both Jackson and Moose have long-term records, and these records establish what we refer to as our climate, the compilation of monthly averages over a minimum of a 30-year period. They are also the stations referenced when a new record high or low temperature is established.

Jackson’s station has been in roughly the same location since the early 1900’s. The Moose weather station was originally located at Beaver Creek, then in 1958 it moved to Moose, behind the Park Headquarters building. In 2018 the station moved across the road, close to new Visitor Center.

Jackson switched from Liquid-in- Glass to a digital-type thermometer in 2002. Moose changed equipment when the station moved in 2018.

The Moose station now sits over dirt and sparse sagebrush, whereas the Jackson station sits over grass. The underlying surface can influence temperatures, as dirt heats up more readily than green grass. Initially, when I observed that Moose was recording higher daytime temperatures than Jackson, I thought it was because of the Moose station’s new location. Further investigation led me to a different conclusion.

Size Matters

Where these two weather stations now differ is in the housing that shields the thermometer. The “beehive” housing that protects the thermometer is supposed to be of a standard size. The Jackson housing, made by Airflo, meets the NWS standard. The Moose housing, made by Nimbus, is about half the size of Jackson’s housing. I am not sure why or how that happened. The NWS told me there was a vendor change before the Moose station was installed.

Why is this important? Because thermometers need good air flow around them to get an accurate reading of the true air temperature. A larger housing allows better ventilation, or aspiration. A smaller housing does not allow good aspiration and can lead to higher temperature readings, most notably on sunny days, with calm winds.

This became glaringly evident when I saw the average monthly high temperatures in August of 2018 that were nearly 5 degrees warmer in Moose than they were in Jackson. On the warmest and driest days of that month, Moose had daily high temperatures that were 9 to 11 degrees warmer than Jackson’s. That is just not realistic.

Going back to a time when Liquid-in Glass thermometers were being used, I looked at the records from August 1988, one of the hottest and driest summers up to that point. What I found was that Jackson’s average high temperature that month was about 2 degrees warmer than Moose. On the warmest days of that month, Jackson had temperatures that were 5 to 8 degrees warmer than Moose. That’s much more reasonable.

So, between 1988 and 2018, Moose changed from being about 2 degrees cooler than Jackson in August to being almost 5 degrees warmer than Jackson in August.

I can only speculate that the discrepancy is in the size of the housing and how well they are able to aspirate. A smaller thermometer housing yields higher temperatures on a sunny day than the same type of thermometer housed in a larger beehive.

There is one more thing that might have been affecting the Moose thermometer, and that is the fact that it is a wireless system, it requires a solar panel to keep the battery charged. Unfortunately that panel is very close to the thermistor housing and may be adding additional heat to the max temp readings, especially on those sunny days!
(See photo below).

Moose weather station, with thermometer housing next to solar panel.

Further investigation showed that overnight low temperatures, temperatures on cloudy or rainy days, and wintertime temperatures when there was snow on the ground, were less affected, and little discrepancy was noted.

The problem with this change in thermometer types the last 20 years, and more so with the discrepancy between thermometer housing sizes, is that our temperature records are no longer consistent. Those inflated high temperatures in Moose become part of the long-term record and comparing them to records prior to 2018 has become apples and oranges.

(Note: The Moose thermometer has been inoperable and not recording temperatures since May 23rd, 2024. I am told by the NWS they are working on fixing that. Hopefully they switch out the beehive to one that is closer to the standard size specified by the NWS, and also move the solar panel further away from the thermometer too!)

It makes me wonder what other weather stations around the country are experiencing a similar effect with their newer instrumentation as we are seeing here in Jackson Hole.

As Robert Mallet said in the Manual of Scientific Inquiry, back in 1859, “Nature, rightly questioned, never lies.”

Post by meteorologist Jim Woodmencey