13 March 2017

New improved Australia RSS TLT satellite temperature

In my first attempt at this I got the longitude wrong (details here). Having corrected that, in my second attempt I defined a smaller rectangle to avoid any ocean areas as shown in the graph that follows.

...which yields the graph (16-point moving average in red):

In the end I decided the ocean areas didn't affect things that much because the larger area was a better match to Steve Goddard's graph of 25 randomly chosen points.  So I prefer the one below. Here is the map area measured:

And here is my derived anomaly with 12-point moving average to match Goddard's with Matlab code:

a1 =
AusNew = squeeze(tsnansum(a1(118:131,:,:)))
AusNew2 = squeeze(tsnansum(AusNew(23:30,:)))
Aus4 = Aus3/(8*14)

Here is the Steven Goddard graph:

Here is the official up-tweaked BoM version which note extends further back in time (from here):

No comments:

Post a Comment