Calibration and quality control of the Coshocton weighing lysimeters

Many weighing lysimeters, including the Coshocton lysimeters, use load cell s to determine evapotranspiration (ET). Because measured ET is sometimes sm all (less than 0.5 mm for hourly ET), these instruments must be very accura te mass measurement devices. To achieve acceptable accuracy, weighing lysim eters require calibration. Electronic changes were found to affect the perf ormance of the Coshocton weighing lysimeters, therefore, a quality control plan was needed. The objectives of this article are: to present a weighing lysimeter calibration method; to provide a statistically valid quality cont rol plan (using control charts); and to describe the performance of the Cos hocton lysimeters operating under this calibration and quality control plan . Control charts were used to monitor the performance of the weighing mecha nism because they can: be relatively easy to user indicate when re-calibrat ion is necessary; identify measurement drift; indicate instrument problems; indicate human mistakes; provide confidence in measurement results; and be adapted to most measurement processes. A calibration procedure was develop ed that: could be performed by one person; could be performed without cover ing the lysimeter; and could be used in conjunction with the quality contro l (QC) procedure. The results show that the Coshocton lysimeters, operating under the described quality control and calibration procedure, have an err or less than 0.025 mm (water equivalent mass) for a single mass measurement . The resulting error in lysimeter-computed daily ET was about 0.03 mm indi cating that the lysimeters are capable of measuring low ET (e.g., hourly ET ). The mass measurement process was generally statistically 'in control' wh ich provides confidence in the lysimeter-computed ET. Under normal conditio ns, two of the three weighing lysimeters were found to be significantly aff ected by wind, likely due to the steep topography (13-26% slopes). However even under somewhat windy conditions (up to 4.5 m/s), the lysimeters were v ery stable and any wind induced error can be considered negligible for most applications.