Description: | Apparently, during installation of the present system on 17 July 1996, soil
heat flow sensor #4 was intentionally placed upside down in the soil to
produce the correct sign of flux. The leads from the sensor must have been
connected with the wrong polarity. On 3 December 1996 site operations
corrected some wiring problems and reconnected soil heat flow sensor #4
according to the proper color code; this caused the sign of the soil heat
flow from that sensor to then be opposite of the other four sensors. All
data from sensor #4 is the wrong sign during the period listed above.
On 17 June 1997 site operations removed the sensor and installed it with
the correct side up, fixing the problem.
The result of this problem was a decrease of reported average soil heat flow of
approximately 40%. This translates into a 0 to 20% error in latent and
sensible heat fluxes (soil heat flow is rarely more than 50% of net radiation,
upon which the Bowen ratio calculation depends). Therefore, sensible and
latent heat fluxes were typically overestimated by about 10% during the
daytime. This is of the same magnitude as the EBBR system/technique error.
The following data values were affected for the period listed
above:
15 minute: mv_hft4
30 minute: shf4, c_shf4, g4, ave_shf, e, h
The latent and sensible heat fluxes can be recalculated by using only soil
heat flows 1, 2, 3, and 5 to calculate the average soil heat flow,
ave_shf = (g1 + g2 + g3 + g5)/4
e = -(q + ave_shf)/(1 + bowen)
h = -(e + ave_shf + q). |