Towards LSDCat2.0: An improved algorithm for detecting emission line sources in integral-field spectroscopic datasets

LSDCat is a software for detecting emission-line sources in 3D integral-field spectroscopic (IFS) data. It has been used effectively on MUSE and KCWI cubes and it is an essential part of the MUSE-Wide survey data processing pipeline. Its algorithm is based upon a slightly simplified matched-filter ansatz. It provides optimal detection significances under the assumption of stationary noise. However, the high frequency and high amplitude noise background from the "forest" of telluric OH Meinel bands at longer wavelengths invalidate this assumption. Thus detection significances for emission-line sources falling on or in between sky-lines are biased low. We here present an improved algorithm that remedies this shortcoming. The improvement will be part of an updated LSDCat2.0. It will benefit especially the detection of  \(z \gtrsim 5\) Lyman-alpha emitters in the deepest MUSE data, e.g., in the 140h MUSE eXtreme deep field.