The Bottom Temperature of Snow cover (BTS) is defined as the temperature measured at the snow/ground interface (Haeberli 1973). This is different to Ground Surface Temperature (GST), which is measured in the ground (soil or rock) slightly below the surface. BTS measurement probes do not penetrate the ground and measure therefore the temperature of the lowermost snow rather than that of the ground.
The scope of BTS measurements is to assess the Winter Equilibrium Temperature (WEqT). The WEqT will depend i) on the presence/absence of permafrost, and ii) on the history of the snow pack on the measurement location. In presence of permafrost, the negative heat flux coming up from the cold frozen subsurface will lead to strongly negative WEqT (typically less than -2 °C), whereas on non-frozen soil, the WEqT will be close to 0 °C or moderately negative (Haeberli 1973). Thus the WEqT can be a good indicator of permafrost occurrence and can help to discriminate permafrost from non-permafrost areas, provided that the snow cover developed early in the winter and remained sufficient to isolate the soil surface from atmospheric influence.
From September 2013 – September 2014 I measured BTS on Suldenferner, using rugged HOBO water temperature Pro v2 combined sensor and dataloggers produced by Onset. Each sensor was placed under one flat rock to shield it from the sun, while true BST would be measured on top of the debris.
The figure below shows where the sensors were laid out on the ground, overlain on the surface lowering in the area measured by repeat airborne laser scans between Autumn 2013 and Autumn 2016, the glacier outline mapped from surface lowering and optical imagery for 2005. The numbered HOBO locations are those that I could recover in September 2014 – the others had gone missing from the glacier surface. The transparent yellow circles are an indication of the debris thickness found by digging to the glacier surface in August 2015, scaled between 3-67cm thick:
HOBO sensors 1, 2, maybe 3, and 9 are on areas binned as having between 0 and +5m of surface change. In the case of sensors 1-3, this zero or potentially slightly positive change is because these locations are higher up the glacier, towards, or within the accumulation zone over this interval. In the case of sensor 9 though, this is right at the terminus position of the glacier in 2005.
The temperature data from March, which is a good approximation of WEqT, for all sensors is plotted below:
All the sensors except 9 indicate that there is ground ice beneath, but at 9 it seems that either there is no glacier ice beneath this location or the debris cover is so thick that the BST cannot ‘feel’ the influence of the ice beneath the debris cover.The two closest thicknesses to this site, just to the west (10cm) and to the SW (37cm) are not the thickest on the glacier but 37 cm is certainly thicker debris than is found at any of the other recovered HOBO sensors. See this other post for more information on the debris thickness on Suldenferner: http://lindseynicholson.org/2015/08/suldenferner-debris-thickness/)
How can we interpret this? I think the temperature data indicates that the site of sdf9 can no longer be considered part of the glacier than is undergoing surface lowering by ablation, and this is backed up by the surface lowering measured by airborne laser scanning.
Aside from this I have yet to find an interesting story to draw out of these surface temperature data. They show that the snowpack over Suldenferner becomes fully temperature (BST reaches 0°C) at sites 3-13 by about 22 May, and additionally at site 1-2 by 06 June. Snowcover persists across the glacier for a further month, and the HOBO sites begin to show signs of strong diurnal cycles in temperature from 07 July, although snow cover is not lost from sdf1 until 30 July.
The maximum temperatures reached during summer days are above 35°C, which is not unreasonable for surface debris, but I would not trust these data as the sensors are housed in black rubber and likely to experience solar heating if the debris moves at all. The night-time temperatures on the other hand might be able to tell us something, so I looked at just those from the first 5 days after I installed the sensors – when I hope the sensors have not moved much from where I put them. Below I have plotted the mean temp from UTM 21:00-01:00 against the elevation of the HOBO sensor:
The character of the elevation relationship remains similar across all nights, and the positive values at 2600m are from sdf9. I wondered if I might be able to draw out a relationship between night-time temperature and elevation, but in fact I now think (as I should have before) that this is really a better indication of the debris thickness at each site, with thicker debris showing the higher nocturnal temperatures, as a thick debris layer has more stored heat from the day time. For the other HOBO sites though the debris thickness is small. The vertical lapse rates in temperature on, for example the 30th September, is about -0.5° over 100m elevation gain, which is in the ball park of typical standard atmospheric lapse rates.