It’s not 40 data points, it’s 40 years of observations. Something that suggests that a phenomenon has consistently happened for 40 years and it didn’t this year. I think it merits investigation, especially when you have established with other data points that the phenomenon is consequential.

Think about it this way, to be able to say that it takes place annually instead of i.e. biannually or monthly, you need a lot more than one sample per year. You need enough samples to know when it is or isn't occurring.

https://www.earth.com/news/unprecedented-collapse-panamas-oc... mentions a lot of date oriented measurements which suggest they probably have at least 52 samples per year, if not daily samples:

> The 40-year record makes the 2025 failure stand out. Average historical onset around January 20 contrasts with a March 4 threshold crossing in 2025.

> The cool season shrank from roughly nine weeks to less than two weeks. Minimum sea surface temperature (SST) rose from historical lows near 66.2°F to about 73.9°F.

author Mulkey responds to a similar question in the comments:

Aaron O'Dea, told me in an email that the upwelling has been "as predictable as clockwork" for at least 40 years of detailed data used in the study. They have less detailed data showing that it goes back at least 80 years. And while this doesn't mean it never vanished before, he said they can trace the the upwelling's impact on coastal ecology and humans for 11,000 years.

40 records as in "observation" is expectable.

But dont dismis climate science that easily.

> An increasingly popular method to deduce historic sea surface temperatures uses sediment-entombed bodies of marine archaea

https://www.ocean.washington.edu/story/Ancient_Ocean_Tempera...

sometimes it's more than enough