Environment, Coasts, Oceans, & Infrastructure

Hello,

I am a Ph.D. student of Agricultural and Biosystems Engineering at NDSU, and I am working with several instrumentation to measure evapotranspiration, soil moisture and crop characteristics.
I faced with a problem on soil moisture data and I like to have your experience.
We are using Stevens Soil Hydro Probe, and we have these sensors in three fields since 2010. however, in one of the field, they didn't take soil samples at exact time and they didn't calibrate the sensors before installing. Right now, I struggle a lot with collected data and the data range is beyond my expectation. The sensors shows 0.75 moisture and also shows .18 moisture for wet and dry period. The soil is heavy clay fargo soil and according to my lab tests, the FC and PWP is 0.47 and 0.27, respectively. also, the saturation point of soil is about 0.54.
I want to know how we can calibrate or evaluate the data right now.
In one of the field, the sensors removed and no longer available for extra test or calibration.

Thanks,

------------------------------
Ali Rashid Niaghi S.M.ASCE
Fargo ND
------------------------------

Hi Ali:

One approach might be to place carefully calibrated sensors in the fields  that had (Have) uncalibrated sensors, and develop a "calibration curve" between the soil moisture in each of the fields, then use that empirical relationship between field soil moisture to adjust the uncalibrated measurements from the earlier periods of record.  

Seems like this could be treated with as an optimal state estimation problem using, e.g. a kalman filter to optimially separate measurement error from model error.   
Just a thought.

Best,
stu

------------------------------
Stuart Schwartz A.M.ASCE
Sr. rsch Scientist
University of Maryland Baltimore County
Baltimore MD
------------------------------