C. L. Pérez Díaz , T. Lakhankar , P. Romanov , J. Muñoz , R. Khanbilvardi, and Y. Yu
who wrote and published:
Near–surface air temperature and snowskin temperature comparison fromCREST-SAFE station data with MODISland surface temperature data
~Although not looking for a Skin snow temperature (T***) vs Surface air (Ts) relation, a very significant paper measured it with great precision using different instrumentation. At Caribou, Maine USA (46◦ 520 5900 N, 68◦ 010 0700 W)
~ Instruments : "An Apogee Infrared Radiometer is used to measure snow skin temperature directly by converting thermal energy radiated from the surface in its field-of-view (FOV) to an electrical signal with a response time of less than 1 s (Muñoz, 2014). This process is automated at every 3 min to an accuracy of 0.2 ◦C. The air temperature is measured directly by a Vaisala Temperature/RH Probe through an automated process; also at a 3 min sampling interval with the same accuracy"
The results from this effort are very important to study:
|Caribou Station equipment extraordinary capacity to measure snow surface skin temperatures with accuracy even with the presence of the sun probably offers the proper way to re-equip mass balance sea ice buoys.|
Here are a few very important observations and conclusions from the authors (in Italics):
~"Results indicate that near-surface air temperature correlates better than snow skin temperature with MODIS LST data"
I have found that so, in particular if NOAA daily climate composites uses MODIS as their data source. I established that we can detect a satellite error by using said simple formula T***<=Ts.
~"This leads to the suspicion that maybe ground-measured LSTs in high-latitude regions covered in snow might not display congruent behavior with satellite readings. Because if the snow temperature satellite readings are far from the real values, this can lead to confusion when trying to predict the occurrence of avalanches or spring floods."
Suspicion confirmed, particularly in the Arctic, this was frequently observed while comparing NOAA daily composites, while they had skin temperature option available, a significant problem here, recognized by the authors, are irregular surface features, either not covered by snow completely or affected by high vegetation, trees for instance.
~Near-surface air temperature tends to affect the snow skin temperature directly, although the latter’s fluctuations are not as drastic (Walsh et al., 1985). The record shows that the winter of 2013 was the coldest of the two (hourly lows of −26 and −36 ◦C in late January for T -air and T -skin, respectively). However, it cannot be ruled out that it is possible for the near surface air temperature to be colder than the snow skin temperature at particular times throughout some winter days, but not common on a daily average basis."
While using much simpler instrumentation and a different technique altogether, the latter assumption: "However, it cannot be ruled out that it is possible for the near surface air temperature to be colder than the snow skin temperature at particular times throughout some winter days",
has never been measured with more primitive method, unless the ground surface has a mix configuration of snow and exposed land, similar to sea ice mix with open water, which gives a different horizon height."Near-surface air temperature tends to affect the snow skin temperature directly, although the latter’s fluctuations are not as drastic" , this has not been observed here in the High Arctic, surface temperatures and skin temperatures vary in tandem almost if not instantly, sometimes skin temperatures vary independently while surface temperatures do not and vice-versa. If the authors rather implied that very near the skin of snow air temperatures may be colder than top of snow, I do not believe so, but the temperatures can be equal.
Self published related articles:
WD January 9, 2018