Noise and Reproduction Error

In an attempt to quantify the random error ( noise ) and the systematic error ( reproduction error ) with comparable metrics the signal observed overall is 

b determines the interval over which the noise of each measurement is averaged (6 1). Typically b should correspond to a wavenumber region, which is similar to the typical line widths obtained in the spectra. For example, at an observed line width of 10 cm and with a discretisation of the spectra into 

2cm intervals, the choice of 6 = 2 would lead to averaging over 26 + 1 = 5 wavenumber intervals, i.e. over 10 cm .  

---