Typical meteorological measurements are performed on moving air. But if one wants to measure the temperature of air, then the small parcels of air moving around the sensor can all have different temperatures. This results in the sensor producing a signal that probably doesn’t correspond to the temperature of any of the air.
Modern sensor such as MMTS have very little damping volume and sensors that respond very quickly to changes in temperature, resulting in temperature spikes . Data from such instruments is not directly comparable to the large damped volume of previous Stevenson screens with traditional, mercury bulb thermometers as the older instruments took many times longer to respond to temperature change; and the heat content of a parcel of air needed to be much greater to heat the glass and mercury of the old thermometers. A little parcel of hot air wouldn’t have had a significant effect.
One solution is to sample and hold a constant volume of air periodically and to measure its “temperature” after a setting period.
This invention is comprised of an outer, well-insulated cylinder (shown here, sectioned) with an internal, insulated piston onto which are mounted the two sets of sensors (yellow). The cylinder is slowly moved from one end of the shaft (blue) to the other, the insulated end-stops (brown) also serving to close the end of the sampling chamber at the left or the right.
Measurement is taken some time after the cylinder stops moving, allowing the air within to settle. Electrical connections to the sensing elements would be through the hollow shaft.
Note that the sensors of the opposite chamber are exposed to the external, “volatile” air in the surrounding protective structure so that the sensors will be largely acclimatised to the next sample of air even before it is taken.
When the cylinder is once again set into motion, it “exhales” the contents of the sampling chamber while “inhaling” a fresh sample at the other end.
Means of moving and supporting the cylinder are not shown for clarity. The mechanical power of rapid cylinder motion could easily induce artificial heating as well as greater turbulence within the sample chamber, so each stroke should be slow enough to make a significant difference in the temperature (etc.) measured.
The entire assembly is to be mounted in a protective structure like a Stevenson screen to avoid heating from direct sunlight, disruption from strong winds and wind-borne debris.
It’s important to keep in mind that the measured temperature will be different to the original temperature of the air sampled; unless the temperature of the sensing element is exactly the same as that of the air as it was sampled. It is safe to assume that there will always be some error.