Correlation of velocity and density contributions to spectroscopic channel maps: Reality check on Kalberla et.al (2022)

Abstract

The existence of magnetized turbulence in the interstellar HI is well accepted. A number of techniques to obtain turbulence spectrum and magnetic field direction and strength have been developed and successfully applied to HI spectroscopic data. To better separate the imprints of density and velocity fluctuations to the channel maps, a new theory-based technique, the Velocity Decomposition Algorithm (VDA,Yuen et.al 2021), has been created. The technique demonstrates that the intensity fluctuations are separated into a component pv that mostly arises from velocity fluctuations and pd that mostly arise from density fluctuations. The VDA helps to clarify the nature of the filamentary structure observed in channel maps. A recent publication (Kalberla et.al 2022,K22) claims that the application of VDA to HI4PI data provides negative correlation of pv and pd,which according to the authors invalidates the technique since it requires that pv and pd have zero correlation. However, the quantities pv and pd given by VDA are naturally orthogonal which can be trivially checked analytically or numerically. That means the correct application of the VDA to any data must provide zero correlation. This is the point that we clarify in this paper and search for the cause of the mistake in the application of the VDA in K22 that resulted in the erroneous conclusion. We prove analytically that by construction pv and pd are not correlated. We identify the likely mistake in the VDA expression that K22 used and reproduce their figures with the incorrect expression. We find that 14 out of 15 figures in K22 are invalid, and thus their criticism of the VDA is ill-founded and arises from their use of incorrect expressions. We conclude that the detrimental mistake that K22 made at their analysis completely invalidate their scientific claim that Y21 isnt compatible to observation.