.

Sunday, August 11, 2013

WUWT is below average in the Arctic

Sou | 1:52 PM One comment so far. Add a comment

Anthony Watts does come out with some beauties from time to time.  That's being generous.  He comes up with doozies almost every time he puts fingers to keyboard.

Today he's blogged an article telling everyone that the climate predictions and projections for 2100 haven't yet all come to pass, and it's already 2013.  I mean, there's only 87 years to go.  All those scientists must be wrong.

I won't bother going into detail about everything he's written.  I'll just mention a few things in passing.  For example, I notice he switched to the USA part way through and ignored the rest of the world.  He put up statistics on hurricanes that were still hurricanes by the time they made landfall in the USA.  That way he could avoid Sandy as well as all the other tropical cyclones in the world. And why he put up tornadoes I don't know, because the science isn't in on what will happen with tornadoes in a warmer world.  He put up winter snowfall in the USA but failed to mention the record melt from April to May this year.  He most certainly avoided mentioning Australia's Angry Summer and all the other signs of global warming in recent years.  It would have spoilt the story he's spinning and upset the punters no end.

What you won't have noticed (because what self-respecting person reads Anthony Watts' blog, bar we blog war people) is that Anthony doesn't seem to know what an average is.


The Arctic is getting hotter


Anthony seems to be trying to tell his readers that the Arctic hasn't warmed, despite the fact that sea ice is disappearing at a faster rate than projected.

Here is a plot of temperature for the Arctic from GISTemp.  It shows the anomalies from the 1951 to 1980 mean for latitudes 64N to 90N.  The average annual temperature for the Arctic has risen by more than 3 degrees Celsius since the late 19th century, by almost one degree since the Arctic warm spell in the 1930s and by almost two degrees since the early 1970s.

Data Source: NASA

Here's a chart showing the amount of sea ice in the Arctic (blue) and Antarctica (dark red) in the month of their respective annual minima for the period since 1979, as well as the total sea ice (green).  Of course the total isn't the total at a real point in time because the months of the minimum are different north and south.

Data Source: NSIDC


What is the "climatic normal" at WUWT?

Getting back to why I said that Anthony is trying to kid people that the Arctic isn't warming.  He put up an animated version of the chart below, showing each year from 1958 onwards.  The chart is described by DMI as the daily mean temperatures for the Arctic area north of the 80th northern parallel, plotted with daily climate values calculated from the period 1958-2002.  (Note that this is a smaller area that the GISTemp chart above, which is north of the 64th northern parallel.)  Anyway, Anthony claimed:
The DMI plot of Arctic temperature for 2013 (at the end pause of this animation) hasn’t gone above the climatic normals since this dataset began in 1958.

Source: Danish Meteorological Institute

Really?  It's a very mixed up sentence to start with.   I assume by "climatic normal" he means the average as calculated by DMI, which is the mean of the period 1958-2002 (the solid green line).  If Anthony means that no year has gone above the "climatic normal" then it doesn't make sense.  The temperature has to have gone above normal at least half the time between 1958 and 2002 - right? Or if not as much as "half the time", considerably above normal some years.

If he means that 2013 hasn't gone above the mean, then he's blind as a bat.

If he expects the Arctic Basin north of 80N to get much warmer than freezing in summer time, then he doesn't know much about the Arctic.  Remember, we're talking above the 80th parallel here.  If your geography is a bit rusty, here is the region we're referring to.  It's the bit inside the circle marked "80".  It's not even all of the Arctic Ocean, and very little land:

Source: Wikipedia


Does Anthony think his chart is an actual physical mean temperature of the Arctic?

I wonder does he also realise that the data is from a model.  It's data assimilation and reanalysis - as described here.  DMI warns readers to not use the charts as a measure of actual physical mean temperature but to use them to compare years.  Here's what DMI say (my bold italics):
The temperature graphs are made from numerical weather prediction (NWP) "analysis" data. Analyses are the model fields used to start NWP models. They represent the statistically most likely state of the atmosphere, given the information available to make the analysis. Since the data are gridded, it is straight forward to deduce the average temperature North of 80 degree North.
However, since the model is gridded in a regular 0.5 degree grid, the mean temperature values are strongly biased towards the temperature in the most northern part of the Arctic! Therefore, do NOT use this measure as an actual physical mean temperature of the arctic. The 'plus 80 North mean temperature' graphs can be used for comparing one year to an other....
The process of making the analysis is called "data assimilation". In an NWP data assimilation system many, very different types of observations and other information are combined in a statistical manner. In practice the assimilation is done via adjusting a recent NWP forecast, a so-called first guess. Because the data assimilation system knows about interrelations of different model variables, assimilation of for example a pressure observation, will adjust not only the pressure, but also wind and temperature. Precisely how much weight to give different types of observations, and how far to distribute their effect in the first guess field, is deduced statistically. The analysis is the maximum likelihood estimate of the state of the atmosphere, provided the statistical information is correct.
An NWP analysis is based on vastly more information than available from any single observing system. Data from ground, aircraft, bouys, ship, satellites, radiosondes, etc. are all combined to adjust the first guess field. As a consequence the quality of an analysis is much better than what can be obtained from gridding, or treating in other ways, data from a single or a few observing systems.
In the plot, the red curve is based on the average 2 m temperatures north of 80 degree North, from the twice daily ECWMF analyses. These are gradually becoming better and more detailed, as the NWP model system at ECMWF is improved with time. That is why the name shift with time (e.g. from T799 to T1279 in year 2010).
... The green curve is based on ERA40 data for the period 1958 to 2002. ERA40 data are in fact analyses, made in the same way as above, but done as a hind-cast, using a fixed version of the NWP model, and spending time on carefully validating and eventually correct or remove all observations found to be in error, before the data assimilation. These, so-called "re-analysis", data represent our best estimate of the properties of the atmosphere for the period they cover.

I think that's about enough on Anthony's misdirection for one blog article, don't you?

1 comment:

Victor Venema said...

Maybe one should add that analysis data is very inhomogeneous (contains non-climatic changes) as the model and the input data used change a lot in time.

Also reanalysis data is inhomogeneous. They do keep the model version fixed and have more data as a near real-time analysis, but still the amount and type of data used as input changes in time.

Especially the introduction of new satellites often produce jumps and especially so in data sparse regions such as the Arctic.

Reanalysis data is very interesting, not only because it can be more accurate as any individual measurement. You can also see global patterns and relationships between various variables. Furthermore, it is computationally very handy to have data everywhere.

However, you should be aware of the limitations of reanalysis data. Especially if you "only" want to look at surface surface temperature, which is directly measured, and you want to look at trends, I would say that climate station data is much more reliable, especially more homogeneous.