Description: | The 30 minute output of soil moisture sensor #5 at Plevna, Ok, E4 has been
going to a large magnitude negative number for low soil moistures for
several months. The calibration coefficients were inadequate for low soil
moistures. The coefficients have been changed to yield correct soil
moistures at low soil moisture (at least, manipulation of the polynomial
for low soil moistures yields correct results for a fixed temperature
correction - extremes of soil temperature and thus the correction may not
yield as satisfying results). These coefficients were derived from the
original calibration data, but including use of low values not used
originally. This particular probe's calibration data was almost exactly
the same as another probe used at the same site, yet the coefficients
derived by the vendor were very different. This made it particularly
obvious that the original coefficients were not adequately tested
(verification tests are not normally performed with dry soil, so this kind
of problem would not normally show up during verification tests). Several
days of data using the new coefficients yields encouraging results. I will
continue to inspect the data to determine the success of this change.
Data values affected:
30 minute: sm1, ave_shf, g1, c_shf1, cs1, ces1, h, e |