If a dataset has uncertainties that vary dramatically with wavelength (for example due to falling stellar photon flux across the bandpass), when binning in wavelength it may be helpful to have a better estimate of the effective wavelength center of a bin. Since we apply inverse variance weighting when binning datapoints together, it would be good to have a point estimate for each bin of what's the most representative wavelength that should be used for comparing to models. If a bin spans 5-10 microns, the center of the bin is 7.5 microns but if it is heavily weighted by the inverse variances to shorter wavelengths, then an effective wavelength should be biased away from the exact midpoint.
If a dataset has uncertainties that vary dramatically with wavelength (for example due to falling stellar photon flux across the bandpass), when binning in wavelength it may be helpful to have a better estimate of the effective wavelength center of a bin. Since we apply inverse variance weighting when binning datapoints together, it would be good to have a point estimate for each bin of what's the most representative wavelength that should be used for comparing to models. If a bin spans 5-10 microns, the center of the bin is 7.5 microns but if it is heavily weighted by the inverse variances to shorter wavelengths, then an effective wavelength should be biased away from the exact midpoint.