Evaluation of Surface Temperature Measurements during BROMEX 2012

During the BRomine, Ozone, and Mercury EXperiment (BROMEX) field study, thermochrons (small, battery-powered temperature sensors) were deployed on the tundra snow surface and the sea ice near Barrow, Alaska, in March and April 2012. Moderate-resolution Imaging Spectroradiometer (MODIS) surface temperatures from both the Terra and Aqua satellites were derived from standard products, air temperature was measured at experimental and permanent meteorological stations and temperatures were acquired from drifting buoys http://iabp.apl.washington.edu/ that were deployed for BROMEX.

The objective of BROMEX is to investigate impacts of Arctic sea ice reduction on bromine, ozone, and mercury chemical processes, transport, and distribution, from sea ice surfaces and near leads in the Arctic Ocean, and atmospheric transport of these chemicals to high mountains on land http://seaice.apl.washington.edu/AirChemistry/ . Since the rates of chemical reactions are affected by air and surface temperature, it is imperative to obtain accurate temperature measurements. To assess the accuracy of the temperature measurements made during BROMEX, a comparison of thermochron, satellite, drifting buoy and aircraft-derived temperatures will be undertaken. For the present study, we compare simultaneously-acquired thermochron and MODIS surface temperatures at a tundra site and at a sea ice site during BROMEX.

For a ~25-hour period beginning on 7 March 2012 at approximately 7:00 GMT Alaska Standard Time (AST), six thermochrons were placed on top of the snow cover near Barrow for cross-calibration. Results show that the temperature measurements were consistent in terms of the patterns of rising and falling temperature, and that five of the six sensors provided results during the entire period that were within ~1°C of each other. The other sensor (#9) was within ~2°C for the first six hours and then matched more closely – within 1.5°C – for the rest of the period.

MODIS-derived surface temperatures from the MOD10_L2 (Terra) and MYD10_L2 (Aqua) land-surface temperature (LST) swath products were compared to the thermochron-derived surface temperatures. For the preliminary work, the center point of the MODIS pixel was within 1 km of the location of the thermochrons. For the ~25-hr period beginning on 7 March, the MODIS temperatures were generally within 2°C of the thermochron temperatures at the tundra site. Within the 1 km X 1 km MODIS pixel, there is significant heterogeneity both at the tundra and sea ice sites, and this could explain all or most of the difference between the MODIS and thermochron temperatures.

At the tundra study site (71°16' 30.45”, 156° 38' 25.00”) thermochrons #9 and #10 were placed together from 9 March through 4 April 2012, but #9 was in the sun during the daytime and #10 was shaded using a white folder to cover it. The daytime temperatures derived from thermochron #9 were up to ~14°C higher than the temperatures recorded by the shaded thermochron (#10). During darkness, the sensors agreed within ~0.5°C. Interestingly, at the beginning of the study period thermochron #9 was inadvertently shaded by its handle according to a photograph taken on 9 March. After 13 March the handle no longer shaded the sensor, which was repositioned, and it was in full sunlight during the daytime, and daytime temperatures were dramatically higher as compared to the shaded thermochron, #10. This demonstrates the extreme sensitivity that the thermochrons (and by extension other temperature sensors) have toward local environmental conditions. Moreover, we will compare the thermochron and MODIS temperatures from the lagoon site and the sea ice site.