Closed jradavenport closed 11 months ago
Answer is... WAY smaller than the spectral resolution. For example with KOSMOS (computed here):
Red 1.18-ctr median delta lambda (A) -1.0350731162379816 median GP var 0.0009277686476707458 stddev of arclines 0.09873894440402012
In the case of real Kosmos data, using the GP to estimate uncertainty, what is the typical error? How does it compare to the dispersion and/or expected velocity resolution?