Hacker News new | ask | show | jobs
by zaps 23 days ago
Stupid question: how do they know with such fidelity what the water temp was in the Pacific in 1877?

It’s not like solid land where there are strata and whatnot that leave a geological record? Or is it?

3 comments

TLDR: Data sources include ships at the time that made records of temperatures as they traveled; coastal weather-stations kept pressure records indicating how strong the big weather pattern was; biological indicators from tree-ring growth, coral formations, and changes in agricultural output.

_________

I found a NOAA paper about the 1877 El Nino [0] and here are two quotes with lengthy references omitted:

> The existence of the strong 1877/78 El Niño event is supported by in-dependent data sources other than [Sea Surface Temperature], such as the Southern Oscillation index derived from sea level pressure, the drought indices derived from tree rings and corals, and records of famine or food production around the world.

> [...] However, there were few in situ [Sea Surface Temperature] observations at that time. [...] These sparse [Sea Surface Temperature] records in the east-central Pacific were measured during 14–31 December 1877 and provided by Deutsche Seewarte Marine (ship ID 120) and Met Office Marine Data Bank (ship IDs 4238 and 4270).

IANAClimatologist and I can't guarantee that's exactly the research that went into the Washington Post article cited by the substack post... however I think it's "good enough" for your question about methods.

P.S.: While I'm usually a "data is toxic" person when it comes to private data and surveillance, this kind of data is the polar opposite: It's amazing when humdrum daily information becomes useful to someone decades later, and I can only hope we continue "paying it forward" with similar gifts to those who will come after us. They are literally irreplaceable, without some form of time-travel.

[0] https://repository.library.noaa.gov/view/noaa/46581

By playing with model inputs, to find a set of input conditions which will create outputs which that historic records.

It's one of those things that is much more precise than accurate, which means that the data on the 1877 map is all plus or minus a range that most of the data falls into. It's like election polls that show candidates a percent or two ahead or behind, but the data is all plus or minus a larger percentage. It gives a general idea of the state of things, but isn't useful for specifics, so it could tell which types of candidates are winning, but is useless at guessing whether any individual candidate will win or lose.

It wasn't that long ago. Maybe someone stuck a thermometer down there.
Over enough of the Pacific to be able to say this with confidence? They might have... but I doubt it.
Oceanic science has has been happening for quite a long time.

You may wish to have a look at this: https://en.wikipedia.org/wiki/European_and_American_voyages_...

cntrl-f temperature: zero matches.

cntrl-f therm: zero matches.

I know voyages have been done for a long time. There weren't many undiscovered islands by then. There were a bunch of species found (my ancestor did a fair amount of that).

But how good was the temperature map?

How do they know what the water temperature was, across the whole Pacific, in 1877? At what resolution do they know it?

dlcarrier's answer (at https://news.ycombinator.com/item?id=48202959) says that they are using the data that they do have to derive the temperature, with "more precision than accuracy". They don't know that accurately, and what they do know, they don't know from "someone sticking a thermometer down there".