How do you handle systematic errors when calculating total measurement error?

Systematic errors are not handled by the root-sum-of-squares method. They represent a consistent bias in measurements. Systematic errors are typically determined through calibration or comparison to known standards and are added *algebraically* to the random error (calculated using the root-sum-of-squares) after the random error is determined. They are not combined in quadrature.