scope:

This report describes the effect of instrumental bandpass and measurement wavelength interval on spectral quantities and provides recommendations on suitable methods to minimize the effect of instrumental bandpass functions. The report concentrates on a correction method of the form commonly used, that of a weighted mean over surrounding values applied to the data and based on assumption of the form of the spectrum across the bandpass. The report also provides information relating to the determination of the bandpass function itself for both monochromators and spectrometers.

As this report was being compiled, a true deconvolution method, not dependent on assumptions about the spectrum, was being developed (Eichstaedt et al. 2013). This method offers the promise of better spectral correction than the weighted-mean methods described in this report, but further study is required and a detailed description is not included here. Weighted-mean methods are simple to apply and give good results for spectral integrals (e.g. as in colorimetry), which are a common application of bandpass correction.

This report also describes the effect of instrumental bandpass on spectrally integrated quantities, but it does not cover the propagation of errors and uncertainty from spectral measurement to spectrally integrated quantities.

Other related effects in spectrometer characterization include stray-light correction and wavelength accuracy. These are included only in as far as they affect the determination of the bandpass function. Full determination and correction of stray-light and wavelength errors are not described within this report.