I have been
on Woodfortrees comparing sea surface temperatures (hadsst3) alongside global mean surface temperatures 1880 - 2016.
You misuse the line - drawing tool
at woodfortrees in exactly the way he warns against.
Woodfortrees now includes the BEST data so people can easily graph it themselves.
# 471 — Amusingly enough, Dan's
Woodfortrees graph confirms what I thought he had done — simply fitted a linear trend to the entire timespan.
(Note for those unfamiliar with the «
Woodfortrees» site: the data are direct from the relevant research organizations, and the various display parameters are standard stuff.
I went to
WoodForTrees and every single graph I ran for the time in question was either up or flat.
Reflecting on that graph a bit, I used
woodfortrees to do a rough - and - ready trend analysis, using successive 5 - year linear trends:
You can check this out on
woodfortrees or download the HadCRUT record directly.
Tamino, and others — if you'll allow a brief aside from the serious statistics — I've found showing amateurs
the woodfortrees site, including the site's caution about fooling yourself with trends, does encourage people to poke at this themselves.