Comments on “Automated Determination of Snow Water Equivalent by Acoustic Reflectometry”

This paper presents the issues in snow water equivalent (SWE) determination using acoustic reflectometry. To determine SWE, the density and the depth of each layer of snow in a snow pack need to be estimated. A noninvasive acoustic reflectometry technique to estimate the SWE is proposed in the mentioned paper, in which it is shown that SWE of the snowpacks with approximately 100 cm depth can be determined using a maximum length sequence (MLS) of length 7 as a probe signal. We performed similar experiments as reported in the mentioned paper to determine SWE using acoustic reflectometry but failed to replicate the given method. The signal processing-related issues such as design of probe signal, direct pickup problem, and so on of the mentioned paper are investigated and presented in this communication. We show that it is not possible to extract the reflection response by using the MLS of length 7 as a probe signal. The amount of sound attenuation and absorption in the snow medium is very high, of the order of 1 dB/cm and greater depending on the frequency, as reported in the literature and measured by us. Hence, the depth of snowpack investigated in the mentioned paper is practically difficult to achieve using conventional loudspeaker and microphone. Also, it has been explained how the cross-correlation peak width and, hence, the layer resolution achieved by the authors of the mentioned paper is not practically feasible by their reported method. We have also demonstrated that Wiener spiking deconvolution-based technique is not able to obtain the sharp peaks as shown in the mentioned paper.