Okay, but before everyone freaks out, 2 of the 5 samples for this year are record lows for a data set of 8 years. Also they seem to have taken only about 2/3 as many measurements this year as previous years, because the low points for previous years are between a number of the samples for this year.

The trend for the final two samples is well into new territory, for sure, but the overall tone is... shrill.