The Arctic science diving season is in full swing (late May). Starting in September and October, the Austral spring will reach Antarctica and science diving will resume there as well.
Virtually all polar diving is done by open-circuit diving, usually with the use of scuba. 
As has often been reported, regulator free flow and freeze up is an operational hazard for polar divers. However, even locations in the Great Lakes and Canada, reachable by recreational, police and public safety divers, can reach excruciatingly cold temperatures in both salt and fresh water on the bottom.
Decades ago a reputed Canadian study measured temperatures in a scuba regulator, and found that as long as water temperature was 38° F or above, temperatures within the second stage remained above zero.
Recent measurements made on modern high-flow regulators at the U.S. Navy Experimental Diving Unit show that the thermal picture of cold-water diving is far more complex than was understood from the earlier studies.
NEDU instrumented a Sherwood Maximus regulator first and second stage with fast time response thermistors. The regulators were then submerged in 42°, 38°, and 34° F fresh water, and 29° F salt water, and ventilated at a heavy breathing rate (62.5 liters per minute), simulating a hard working diver.
In the following traces, the white traces are temperatures measured within the first stage regulator after depressurization from bottle pressure to intermediate pressure. That site produces the lowest temperatures due to adiabatic expansion. The red tracing was obtained at the inlet to the second stage regulator. The blue tracing was from a thermistor placed at the outlet of the “barrel” valve within the second stage regulator box. Theoretically, that site is exposed to the lowest temperatures within the second stage due to adiabatic expansion from intermediate pressure to ambient or mouth pressure.
Regulators were dived to 198 ft (60.4 meters) and breathed with warm humidified air for 30-minutes at the 62.5 L/min ventilation rate. The regulator was then brought to the surface at a normal ascent rate.
To make the breathing wave forms more distinct, only one minute of the 30-minute bottom time is shown in the following traces, starting at ten minutes.
The first two tracings were at a water temperature of 42° F. In the first tracing, bottle pressure was 2500 psi, and in the second, bottle pressure was 1500 psi. (For all of these photos, click the photo for a larger view.) 
Color coding of thermistor locations, all internal to the regulator.
When bottle pressure was reduced from 2500 psi to 1500 psi, all measured temperatures increased. The temperature at the entrance to the second stage oscillated between 0° and 1°C. At 2500 psi that same location had -1 to -2°C temperature readings.
The next two tracings were taken in 29° F salt water. The coldest temperatures of the test series were in 29° F water with 2500 psi bottle pressure.
As a reminder, 32°F is 0°C, -22° C is equal to -7.6° F, and -11°C is 12.2°F. At a bottle pressure of 2500 psi, the temperature inside the second stage (blue tracing) never came close to 0° C. So we’re talking serious cold here. No wonder regulators can freeze.
This material was presented in condensed form at TekDiveUSA 2014, Miami. (#TekDiveUSA)