...
u, v, and tc do not seem to have this issue, which makes me think that it is not an actual misfunctioning of the acoustic part of this instrument. It is non-orthogonal, so any path problem would show up on a horizontal velocity in addition to the vertical. We should call this to the attention of whomever owns it.
A data example, unparsed from the sensor:
2017 05 19 00:00:43.6530 0.04999 44 M:x = 392 y = 65 z = 87 t = 1154\r\n
2017 05 19 00:00:43.7030 0.05 44 M:x = 364 y = 57 z = 94 t = 1176\r\n
2017 05 19 00:00:43.7531 0.05008 44 M:x = 339 y = -33 z = 13593 t = 1169\r\n
2017 05 19 00:00:43.8030 0.04991 44 M:x = 359 y = 55 z = 32 t = 1161\r\n
2017 05 19 00:00:43.8525 0.04949 44 M:x = 387 y = 68 z = 26 t = 1151\r\n
Clearly, the bad value of 13593 came straight from the sonic.