L'Ombre de l'Olivier

The Shadow of the Olive Tree

being the maunderings of an Englishman on the Côte d'Azur

23 October 2008 Blog Home : October 2008 : Permalink

Fun With Recent Climate Trends

At various climate skeptical blogs such as Climate Audit, Watts Up With That, Rank Exploits, No Consensus and of course at the more orthdox Real Climate and Tamino much has been made of the recent trends in climate and their agreement or not with model predictions and so on. All of these get stuck into statistics with various parties claiming that other parties are incorrectly handling errors, significance, consistency and so on and even sometimes the definitions of these words.

I admit to being but a studious amateur in the field and while I can just about follow what's going on the chances of me making a worthwhile contribution is limited. What I can do though is use the very excellent "Wood For The Trees" interactive plot pages to create my own graphs. I can also use perl to automate the more tedious bits of the graph creation and thus put lots and lots of lines on a single graph.

So what I've done is create a few graphs showing recent trends in the record. Here, for example, are the recent trends for the WoodForTrees Temperature Index which is the WFT average of the four most commonly quoted global temperatures. What I've done is overlay the temperature from 1979 to present day (and note this graphs WILL auto-update so potentially things may change from when I first created them - 23 Oct 2008) along with the OLS trend from start of each year to current date for the years 1990 to 2005.

In this case we can see that up until 1997 the trend has been positive (i.e. things are heating up) although the trend flattens over time, 1998 is negative, 1999 is positive and as is 2000 and from 2001 the trends are increasingly negative. To make it clearer at the cusp here are the 4 years 1997 to 2000 inclusive) only with temps from 1995.
If we look at the same sort of timeframes from the Hadley CRU we get to break the world down in to Northern and Southern Hemispheres and we see very clearly that the more oceanic South has had a cooling trend in every year from 1997 whereas the North took until 2002 to start cooling

HADCRUT3 unadjusted NH mean

HADCRUT3 unadjusted SH mean

When you put this together it turns out that the Southern cooling just fails to balance the Northern warming for 2000 but thereafter the trends become increasingly negative.

Finally it may be worth putting some of these things into a longer term historical perspective. Here is the HADCRUT global from 1930 with 10 year trends every 5 years beginning in 1933 and ending in 2008.

It is worth noting that the 2007/2008 winter minimum temperature is clearly one of the cyclic colder snaps, just as 1998 was one of the warming peaks, and that average temperatures can drop fast. Also when we look at the Southern Hemisphere we see that the climate has not increased nearly as much as it has in the north.

HADCRUT3 unadjusted SH mean

In the south 1998 is barely hotter than previous peaks in the 1940s and the 1970s. If it turns out that the adjustments made to compensate for urban heat islands and other effects (most of which occur in the North) are wrong then this global warming thing definitely looks rather less of a problem.

On the other hand when you look at the whacking great drop in the late 1940s which then reversed itself in the 1970s it could be that the current negative trend is just a blip...

For those that may care the perl code for the graph generation is based on this:
use CGI ':standard'; #code for html bits
my %data = (

'wti'=>'WoodForTrees Temperature Index',
'hadcrut3vgl'=>'HADCRUT3 variance-adjusted global mean',
'hadcrut3vnh'=>'HADCRUT3 variance-adjusted NH mean',
'hadcrut3vsh'=>'HADCRUT3 variance-adjusted SH mean',
'hadcrut3gl'=>'HADCRUT3 unadjusted global mean',
'hadcrut3nh'=>'HADCRUT3 unadjusted NH mean',
'hadcrut3sh'=>'HADCRUT3 unadjusted SH mean',
'gistemp'=>'GISTEMP land-ocean global mean',
'uah'=>'UAH NSSTC lower trop. global mean',
'rss'=>'RSS MSU lower trop. global mean');

for $d (sort keys %data) {
   my $img = "http://www.woodfortrees.org/graph/$d/from:1995";
  $img .= "/plot/$d/from:$_/trend" for (1997..2005);
  print p("$data{$d}".br.img({-src=>$img}));