| Description |
All soil levels below 15cm show essentially the same magnitude of soil temperature change on a daily basis and at the same times as the 15cm level, with no delay at greater depths. This does not represent physical reality at these depths.
Furthermore, this behavior began with the very first installation of the SWATS and has been seen in every SWATS system.
However, the magnitude of this problem is much greater at extended facilities E31 through E38.
Comparison of the SWATS with co-located SEBS or EBBR 5cm soil temperatures shows that the diurnal trend of soil temperature at 5cm is similar for the SWATS and the EBBR or SEBS, but usually exhibits a larger maximum and a lower minimum. The SWATS minimum also is reached less smoothly than the EBBR or SEBS. This is likely because the SEBS and EBBR soil temperature measurements are an integration from the surface to 5cm depth.
The SWATS pattern is also seen, to a much more minor extent, in the trise, trise_ref, water content, soil water potential, and fractional water index measurements at 15cm and below; however, the variation in these measurements is well within the uncertainty of the measurements.
It appears that the SWATS probe thermistor lead connections with the datalogger terminal strip were acting as thermocouples. Why the effect was so much greater at E31-38 is not clear, but it probably has something to do with how the lead connections to the datalogger were performed.
This situation particularly affects the measurements during daytime hours, and corresponds with heating of the SWATS enclosure by the Sun.
5cm SWATS soil temperatures are very similar to co-located EBBR or SEBS 1-5 cm integrated temperatures at most extended facilities, and therefore appear to be valid (the solar heating/thermocouple effect may be swamped by the actual soil temperature diurnal change).
15cm SWATS soil temperatures may also be valid at extended facilities other than E31-38, as they show a larger diurnal variation than lower depths and about midway in magnitude between the 5cm and 25 cm measurements.
Since it seems likely that the soil temperature measurements below 15cm were affected by the solar heating of the enclosure at all but E31-38, and at all depths below 5cm at E31-38, individual measurements of soil temperature at these depths during daylight hours, and daily averages of the same, can not be trusted on most (particularly sunny) days. The effect on cloudy days seems smaller.
For further information see the SWATS technical report: https://www.arm.gov/publications/tech_reports/doe-sc-arm-tr-193.pdf |