Bend It Like NASA (part 1)

The mainstream media are attempting to convince the public that July 2023 was the hottest month on record but is this true?

TBH the mainstream media are only following this story created by NASA…

…get that kettle on, open that big cake tin (the one containing the walnuts and coffee icing jobby) and let me reveal just how low NASA and others are prepared to go in order to score goals for Globalistas United.

We shall start with football.

It is Sunday afternoon and, with a flask of hot tea in your hand and pockets crammed with cheese sandwiches, you are down at the ground supporting your local team – Nempnett Thrubwell¹ Wanderers (affectionately known as the ‘thumpers’). Today is the big needle match against arch rivals, Butcombe² Rangers (affectionately known as the ‘badgers’). At half time the thumpers are leading 2 – 0 and you are ecstatic. The cheese sandwiches never tasted so good even though you forgot the pickle.

Players file out of their respective creosote sheds for the second half and you clock team changes for the badgers. Out trot a few famous faces and you spot Harry Kane, Lionel Messi, Gareth Bale and Cristiano Ronaldo, amongst others. Badgers trash thumpers 23 – 2. Suffering inconsolable despair you collar the referee and ask how this travesty is allowed. The referee points out that it is within the rules for teams to make changes at half time, and that includes swapping local lads for the world’s finest.

“But that’s not fair! That’s no longer the badgers!” you exclaim in total anguish. And you’d be right. The team that came out at half time was no longer the badgers per se, and a win was inevitable. It is exactly the same with the land temperature record in the surreal world of climate science; forget bending it like Beckham, we’re going to bend it like NASA!

Diego & Paradise

Let us suppose that there is a weather station called Diego Sands deep in the Sonoran desert. At Diego the thermometer sits on the top of a corrugated iron roof on a brick building set into an expanse of concrete. Let us also suppose that Diego was established in 1978. Aside from Diego there is one other weather station in the region in a leafy glade next to a babbling brook where the thermometer sits on an expanse of mown grass surrounded by cypress and rhododendron. Appropriately called Paradise Springs, records have been carefully kept there since 1824. What would happen if we combined the data from Diego Sands with Paradise Springs?

Precisely. We would introduce observation bias so big you could throw a saddle on it and ride it to town! Now what if the same travesty that beset our local game of footie is being foisted on the global temperature record as established by the big names such as NASA, NOAA, Berkeley and CRU/Hadley Centre?

Nasty business, I shall grant, but this is exactly what is happening behind the scenes and under the bonnet; and because tens of thousands of stations, ships and buoys are involved we have no chance of determining just how much bias is being introduced by pooling big data in a big, bold way.

How Many Thermometers?

Q: how many land stations that are still operative in 2023 were also operative in 1880?

We can attempt to answer this by using NOAA’s global historical climatology network – daily database (GHCNd) in conjunction with KNMI’s Climate Explorer. Bring up the daily average temperature field in that sassy package, dial in 143 years as the minimum record length, expand the region to the globe and we find not 200,000 stations, not 20,000 stations, not 2,000 stations and not 200 stations. In fact we find just 21 super-long series global stations, with the following national breakdown: Australia (1); Austria (3); Belgium (1); Canada (1); Germany (6); Ireland (1); Sri Lanka (1), United Kingdom (3); United States (4). Oh dear.

Now take a look at the land regions on NASA’s global map that are coloured in various angry shades – just where has all that data come from? We don’t have 21 solitary blobs marking the 21 GHCNd stations that have been recording temperature since 1880, we have a continuum that implies millions of stations across the land masses of the Earth. Do these millions of stations even exist? Nope!

If we search NOAA’s GHCNd database for all stations that have been open at least one year we are informed that there are just 40,136 global stations, the majority of which are in the United States (55.3%). Thus, we discover that NASA’s global map cannot represent actual observations of land surface temperature and is based on what we call a gridded product. A gridded product e.g. GISTEMP takes the data records from one weather station at a time and smears them over a wide area to fill in all the holes.

How wide is anyone’s guess but I did once discover stations in Northern Australia being used to fill the holes in the Burmese jungle data record some 4,000km distant. That’s a lot of smearing! Not only that but since most weather stations are located in larger towns and cities, then NASA and others are smearing the urban heat island effect all over the globe and are hoping folk won’t notice. Hence a perpetually pink map; in effect you are looking at the temperature of tarmac in a few places being passed off as a somewhat smudged ‘global’ temperature.

In a nutshell we are looking at a work of highly processed and modelled fiction being passed off as direct observation. Not only that but NASA are also trying to convince us that this fiction can be traced all the way back to 1880.

Going Back to 1880

Let us take those 21 land stations and see what their combined daily temperature record looks like for the months of July dating back to 1880. There are three ways to do this:

1. We can derive the mean daily temperature (average).

2. We can derive the mean daily maximum temperature (my fave).

3. We can derive the absolute maximum daily temperature (hottest potato).

And here are the corresponding slides:

Before we cogitate deeply let us remind ourselves where these data come from. Top of the goal-scoring table is Germany with 6 stations. Next to this is Austria with 3 stations, so we have 9 / 21 (42.9%) stations sitting in large cities in central Europe, thus we shouldn’t be at all surprised if these plots are replete with urban heat island which, judging by the look of things, suggests accelerated urban development after 1970.

Some may argue that this is climate change and it may well be, but how are we going to separate global warming proper from urban warming improper? The answer is we can’t because we’ve stuck our thermometers in the wrong place!

How Did July Do?

So how did July 2023 fare in all this? Well, in terms of the average daily temperature, it fetches up in second place at 21.63°C, with the gold going to 2006 (21.77°C) and the bronze going to 1994 (21.01°C). I can tell you now that the difference between these temperatures is not going to be statistically significant, but the main point here is that NASA is bending the truth by using a huge sample of stations and permitting all manner of changes to the team at half time.

In terms of the mean maximum daily temperature – arguably a more sensible measure of the heat that people actually feel - July 2023 fetches up in second place at 26.53°C, with the gold going to 2006 (27.40°C) and the bronze going to 2018 (26.32°C). In terms of the absolute maximum daily temperature amongst these 21 super-long series stations July 2023 fetches up in 16^th place at 42.80°C, with the gold going to 1925 (45.6°C), the silver going to jointly to 1972 and 2021 (45.0°C), and the bronze going jointly to 1988 and 2002 (44.4°C). But, hey, let’s go bend it like NASA and use as many stations as we want without a care in the world for analytical integrity!

Why 1880?

Why has NASA started the clock in 1880 when they are sitting on data stretching back to 1850 and before? Three of the stations in the super-long GHCNd series sample (Oxford, UK - 1825; Jena Sternwarte, Germany - 1824; Uccle, Belgium - 1833) provide ancient data so let’s have a look at the mean daily July temperature slide again:

Aha! It looks like they enjoyed some warm weather back in 1820 – 1860 that I presume expertistas like NASA would prefer we not discuss. Alarmists will claim folk couldn’t measure temperatures properly back then for they cannot permit anything that would spoil their nice, clean, and unashamedly apocalyptic narrative. That being said there is indeed a fly buzzing in the ointment so I am reaching for the swatter…

Reaching For The Swatter

Despite me insisting that we only consider stations that were operative throughout the period 1880 – 2023 we’ve still got the issue of missing daily records to contend with, and we may think of this as players being sent off the pitch. In our game there are 21 players who’ve been playing without rest for just over 143 years; at least that is the assumption. If we squint at the daily records for the period 1/1/1880 – 31/7/2023 we find some days when only 6 players were on the pitch, with the average being 20.51 players. Let us take a look at the mean number of players each year to see how this has varied:

We expect some difficulty with data collection during WWII and this is what we observe. We also observe periods of difficulty in the early 1880s and 1920s, but what is standing out like a sore thumb is the substantial loss of daily records beyond 1980 when climate was supposed to be of grave concern. How bonkers is that? Compare that big bunch of sorry looking rates for 2005 – 2022 with the war years and you’ll see that meteorologists were doing a better job when being bombed! I don’t know about you but this smells a bit fishy: could we be looking at cherry-picking at the coal face?

By accidentally missing records during cold snaps you generate fake warming; by accidentally missing records during heatwaves you generate cooling. Accidentally miss the latter, and then accidentally miss the former and you can conjure a fake warming trend. I guess we better have a look see if we can detect any signs of foul play…

Dodgy Day Count

I shall start by going back to the daily records and flagging days when the station count dropped below 20 - I’m happy to forgive one station having trouble at a time but that’s about it! This flag gives me what I am calling a dodgy day count, and here’s what this looks like for the period 1880 – 2022:

Ouch. Big holes during WWII we can forgive, but big holes during the last two decades of climate shenanigans? Not on your Nelly! Why collect less meteorological data than at any other point in history when we should be collecting more? Do I smell a cherry pie baking in the oven?

If we calculate the mean rate for the dodgy day count by month and slap on some 95% confidence intervals then we arrive at this error bar plot:

There is no getting away from what is tantamount to a serious amount of bias, with the dodgy day rate blossoming during the winter months. N.B. There is also a modest dollop of bias by day of week, with rates blossoming at weekends as may be expected, but that isn’t of concern. What is of concern is that winter temperatures among the 21 super-long series land stations are going missing and the last two decades have been blatantly dodgy.

Is there an increasing tendency toward missing out more and more winter records over time? If there is, then we’re looking at a smoking gun, so I better do something fancy for part 2 - stay tuned!

Kettle On!

1. A genuine and rather pretty village in Somerset, England.

2. Another genuine and rather pretty village in Somerset, England.