Starting at the East Coast over between the Arabian peninsula and Madagascar; inland from the Seychelles (where we left off in the last posting about islands). We will then cross the inland central areas and finish on the West Coast.
A small rise, but almost all of it from the end section where we’ve taking the typical 1990 pivot but this time accompanied by a lot of data dropouts and then it ends in 1999. Not much to work with here.
Right next door to Dijibouti. Substantially the same climate zone.
Whoa! Dropping like a large boulder. Almost 2 C in a century. Better drop that one on the floor Real Quick… no sense keeping it in after the 1990 re-jiggering.
A bit inland but fronting on both of the first two. (Also fronting on Somalia that follows).
Nice dropping trend coming in to the 1991 Pivot. We get a low value in “the baseline” used by GISS and CRU, but we’re still below the 1954 period. Notice the dramatic compression of the “low going” peaks of the monthly anomalies in recent years.
Mighty darned flat up to the 1980 point. Then it just halts. Pirates, dontcha know ;-)
Now here we have something we can work with. Homogenize this in ot the surrounding “flat and falling” countries, especially after the ‘cut off’ points, and we’ll have a nice rising trend with gradual onset.
But looked at up close and personal, it’s a very questionable start on life, then a flat middle with a small rise, followed by the usual “hockey sticking” at the end. Loads of “splice” artifacts, IMHO. Along with whatever that change of process was in 1990.
Right next door to Kenya, and substantially identical climate influences. Same coastal waters. Same inland reaches. Even share the same large inland lake. These two ought to be identical twins.
Or not… We have a ‘step up’ on thermometer count increases in about 1925, then it looks pretty flat to me. Not at all like Kenya. Lets look at it divided by segments and see if we get what looks like a couple of flat segments spliced together in the record.
Uh, yeah. I’d call that two flat segments spliced. Though we do get a nice cold spot in the “baseline” used by the major climate codes like GIStemp and the Climategate CRU. While I’d count Tanzania as “not warming” they will find it to be “warming” by 1/2 C based on that “Dip”.
How about just a bit more inland?
Steady dropping into 1978 then dropped on the floor. No worries, we can fill it in from Kenya…
Too short to really use for anything, but I note that it was rising into the ’70s while the others were dropping.
Zaire ( Democratic Republic of Congo )
A very nice rising trend, but look at the “thin” nature of that middle part, around 1978 to 1999. Almost no ‘cold going’ peaks or excursions. Lots of wobble in that Count line as well. Plenty of opportunities to splice bits and pieces of different series into a “rising trend”. It looks like Zaire is going to need a “real temperatures” hair graph, perhaps done by subsets of thermometers.
Congo (on the coast)
Right next door is “the other Congo”. The “People’s Republic of the Congo”.
Wow! Nice steady drop right into the data dropout and “1990 Pivot” event. We managed to prune enough thermometers and create enough splice artifacts to recover a bit of warming, all the way back to the zero line…
Central African Republic
How about just north of those two? Does it do any better?
Oh Dear! It’s dropping too! The best that the 1990 Pivot and data dropouts can do is to “Hide the Decline”.
Gabon is on the coast, but with The P.Rep. of the Congo wrapped around it. If anything defines “dead flat” this is it. Oh Look! It gets cut off in 1990… Wonder why?…
No Data in GHCN
Just north of the P.Republic of Congo and just next to the Central African Republic. This is what I would have expected to find world wide. A LOG fit to the rising temperature curve (as CO2 has a log profile to the CO2 blocking of IR radiation). I think this makes 2 of these we’ve found so far where a LOG fit “looks right”.
How about when looked at “by segments”?
I find the symmetry of the Thermometer Count line and the “dT” Anomaly Running Total line to be a striking feature. Too bad we can’t smoothly grade thermometers in and out in all the countries. Guess we will have to use “fill in” and “homogenizing” to get this kind of curve in other places…. (Or just admit that instrument change creates the anomalies, not climate change.)
Right next to Cameroon, ought to have a very similar profile. Equatorial with shared coastal area / ocean. Similar “reach” inland. Nigeria has a few more mountains, but since GHCN has largely abandoned mountain thermometers, they will not contribute much to the recent record.
Wow! Again! Not at all like the others. Magnificent warming at the start of time, then dead flat. Notice that the 1980 point has massive data dropouts. Many many years gone. Other codes will fill them in from somewhere else. The dT/dt system will use the fragments that do show up and only compare the same months in the same place to each other. And we find a very low volatility, but no net warming. Substantially a ‘dead zone’.
So I’d be very interested in recent temperature data from Nigeria. I’d guess that the oil companies probably gather some at their locations. Probably an airport or two has data as well. If GIStemp shows a warming Nigeria, and the present data say’s “NOPE!”, we’ve got a pretty good “smoking gun” that the warming is Bogus.
Oh, look at that. They find 1 to 2 C of “warming” in Nigeria… Who knew?
GISS is broken. The baseline is a “cherry pick” and the baseline DOES matter.
The data are seriously compromised and “fill in” with “homogenizing” makes it worse, not better. It hides what is really happening and smoothes out divergent and disjoint obviously incongruent series to blend them into a fictional common warming that does not exist.
Averaging a bunch of broken things together does not make them better. No matter how fancy the averaging process nor how elegant the splice.
A splice is still a splice, and splicing data is a bad idea.