We need to be careful focussing upon «trends» — it can lead to serious errors of context — and this underlies the entire «global warming» thesis which relies upon computer
models with entirely false (i.e. non-natural) notions of an equilibrium starting point and
calculations of trend — this conveniently ignores cycles, and it has to because a) there are several non-orbital cycles in motion (8 - 10 yr, 11, 22, 60, 70, 80, 400 and 1000 - 1500) depending on ocean basic, hemisphere and global view — all interacting via «teleconnection» of those ocean basins, some clearly timed by solar cycles, some peaking
together; b) because the cycles are not exact, you can not tell in any one decade where you are in the longer cycles.