Detection statistic (lazuli_exo.detect)
See also
The template statistic — what this statistic does and does not mean. Read it before using these values.
Spectral-template S/N of a noiseless binned model spectrum.
The statistic is a matched-template amplitude S/N under a diagonal covariance: snr = sqrt(sum w (d - dbar)^2), w = 1/sigma_d^2, dbar the weighted mean depth. It is the expected S/N of fitting the (noiseless) model spectrum as a template against the best-fit flat line, assuming independent Gaussian errors and no nuisance parameters beyond the mean. It is NOT a calibrated Gaussian-equivalent detection significance: there is no defined null/alternative likelihood ratio, no covariance, and no false-positive calibration.
Per-molecule values under “band_snr” evaluate the same sum restricted to that molecule’s wavelength band mask (relative to the same global dbar). They are band-restricted template norms, not molecule detections: they include absorption from every other species in the mask, haze/cloud/ continuum structure, and the offset of the masked region from the global mean; bands of different molecules overlap (H2O and CH4 near 1.15 um). The rigorous upgrade path is a likelihood ratio against leave-one-species- out model tables. CO and NH3 are present in the models but have no band masks in the 0.4-1.7 µm range, so they are never reported — their absence from the output is not a non-detection.