232-1 Sensors for Precise Measurement of Air Temperature on Automated Weather Stations: Precision Thermistors Vs. Prt's.

See more from this Division: ASA Section: Climatology & Modeling
See more from this Session: Symposium--Recalcitrant Problems and Emerging Solutions in Biophysical Measurements and Sensors: I
Tuesday, November 4, 2014: 8:05 AM
Long Beach Convention Center, Room 103B
Share |

Bruce Bugbee, Crop Physiology Laboratory, Utah State University, Logan, UT, Mark Blonquist, Apogee Instruments Inc., Logan, UT and Jobie Carlisle, Utah State University, logan, UT
Platinum resistance thermometers (PRTs) are widely considered the most stable and accurate temperature sensor and are commonly used. Thermistors are also commonly used to measure temperature, but they don’t have the same reputation as PRTs, despite improvement in construction materials to improve stability and accuracy.  Long-term stability has been an ongoing concern with thermistors, but most of these concerns apply only at sustained temperatures above 100 C.  At cooler environmental temperatures, moisture can diffuse into poorly sealed thermistors and cause drift.  For this reason, the thermistor sealing material of choice is either glass or water resistant epoxy.  Studies by the National Bureau of Standards indicate that well-sealed thermistors are as stable as the best PRTs, and drift less than 0.002 C per year at environmental temperatures (-60 to +60 C). Reviews often list poor linearity as a disadvantage of thermistors, but with modern programmable data acquisition systems non-linearity is no longer an issue.  As long as the resistance/temperature curve for the sensor is repeatable, the Steinhart-Hart equation can be used to accurately convert resistance to temperature.    Thermistors are typically smaller than PRTs, which means that they have reduced errors from thermal radiation, even when housed inside radiation shields.  We are conducting long–term tests with replicate epoxy-coated thermistors in a temperature cycling chamber; and replicate PRTs and thermistors in an aspirated shield in an extreme temperature mountain environment.  These tests indicate thermistor drift of less than our detectable limit of 0.02 C.  The thermal-induced error from solar heating of the 3-mm diameter stainless-steel PRTs was 0.2 C greater than the smaller thermistors.  These data indicate that properly constructed epoxy thermistors can be a better choice than PRTs for precise measurement of air temperature on automated weather stations.
See more from this Division: ASA Section: Climatology & Modeling
See more from this Session: Symposium--Recalcitrant Problems and Emerging Solutions in Biophysical Measurements and Sensors: I
<< Previous Abstract | Next Abstract