Uncertainty evaluations on the measurement of 'carbon' in the research on karst carbon sink effect - A case for dissolved inorganic carbon

摘要

In order to set up an evaluation method for the uncertainty in determining dissolved inorganic carbon (DIC) in karst river basin by Non-dispersive Infrared (NDIR) absorption detection technique, the continuous propagation model of uncertainty is used to evaluating the uncertainties from DIC measurement in two typical karst groundwater samples. The main steps are as follows: to fit the calibration curve by means of double-error regression firstly; and then to quantify each uncertainty component in the evaluation process; lastly to obtain the synthetic uncertainty model for DIC determination results. Calculations through experimental results show that: (1) the main sources of measurement uncertainty derive from the sub-uncertainties of calibration solutions, calibration curve fitting and measurements process; (2) the lower the DIC content in groundwater samples, the greater the relative uncertainty of measurement results, and the sub-uncertainty from the fitting of calibration curve gives one major contribution to the total uncertainty.