East Anglia – Home of the Zombies? A critical look at Met Office gridding and their interesting semantics.

This article follows on from my analysis of Devon sites and their changing locations and numbers over the last 50 years. In this report I will be attempting to identify the issues and inherent problems surrounding the Met Office and CRU systems of climate modelling and how readily such systems can be played. Further aspects of personal attacks on me will be considered in context.

The above map is AI derived from the Met Office Location Specific Long Term Averages web page. I opened out the markers and asked AI to name them for clarity so this is NOT how it appears under normal interrogation.

Taking Walton-on-the-Naze (where temperature readings stopped 31/8/1973) as a starting point and running north and east there are:

Levington – Temperature readings from 1/10/1977 Ended 1/12/1996

Scole – Temperature readings from 1/1/197 Ended 31/7/1980

Lowestoft – Long term claimed “Historic Station Data” Temperature readings Ended 1/11/2010 Online viewable data post closure DELETED by the now resigned government Minister.

Morley St Botolph – Temperature readings from 1/6/1968 Ended 1/4/2005

Hemsby Former major station and Radiosonde launch site from 1948. Temperature readings ended 2/4/2001

Coltishall WW2 aviation site. Temperature readings Ended 4/4/2006

Weybourne is the first currently operational “climate averages” site in that section of almost 2,000 square miles of Essex, Suffolk and Norfolk. However Weybourne did not start temperature readings until 1/1/1991. This station recorded precisely ZERO readings for the period now shown below to the second decimal place.

Station: Weybourne

Climate period: 1961-1990

Month	Maximum temperature (°C)	Minimum temperature (°C)	Days of air frost (days)	Sunshine (hours)	Rainfall (mm)	Days of rainfall ≥1 mm (days)	Monthly mean wind speed at 10 m (knots)
January	6.30	1.57	8.48	–	51.98	11.20	14.19
February	6.15	1.53	7.65	–	34.47	9.60	13.66
March	8.49	2.94	4.01	–	39.91	10.97	14.44
April	11.07	4.98	1.16	–	42.98	9.60	12.58
May	13.97	7.61	0.15	–	50.72	10.00	11.64
June	17.48	10.49	0.03	–	47.11	7.10	11.22
July	19.65	12.44	0.00	–	51.69	8.30	10.66
August	19.39	12.73	0.01	–	44.42	8.73	11.25
September	17.65	11.23	0.01	–	49.77	8.17	13.00
October	14.32	8.58	0.07	–	53.02	9.63	12.24
November	9.54	4.66	2.15	–	68.58	12.17	14.25
December	7.48	2.85	5.62	–	60.93	11.10	14.48
Annual	12.66	6.83	29.35	–	595.58	116.57	–

So how is it that numbers are produced with such faux accuracy long before (30 years in the case of Weybourne) and after (53 years in the case of Walton) these stations actually existed? The answer is simply that they are the output of a “computer model” that is implying what the conditions would be in these immediate localities (“Gridded Squares”) IF there were a weather station there. Rather than me explain it further in overview, here is a Youtube Clip from the Met Office themselves explaining it.

I shall leave it readers to draw their own conclusions from this official “explainer”. For my part I find the laungauge used rather deliberatelly vague but intended to sound authoritative. The word “Robust” is doing a lot of heavy lifting in its multiple appearances. “Complex statisitical calculations” sounds impressive until one discovers that the public are seemingly not allowed to openly know any of them. “Science backed estimates” and “Peer reviewed” all sound impressive especially when “all brought together from real observations” and thus allowing them to “not be reliant on opening and closure”. All sounds impressive but is it just a smokescreen? After all the presentation is specially “modelled” as a computer game analogy.

My personal opinion is that the presentation uses no more than a sales brochure style and is “dumbed down” for a wider audience acceptance that is not informed enough to question the basic concepts implied. I have particular concern with the term “understanding the weather of the past”. “Past weather” is a matter of historical record – understanding the details that caused it is obviously important but the actual climate (not weather) data is a matter of record. The suggestion of “fewer observations in earlier decades” is a very misleading statement. There certainly were fewer recording points in the latter nineteenth century but similarly there were vastly more surface stations in the latter 20th century than there are today.

As an important aside – I find the above presentational style quite amusing – ” ’cause you know sometimes words have two meanings“. I was personally heavily criticised for using the precise term “Fabrication” regarding Met Office data used for closed weather stations.

A definition of the term “Fabrication” is “the act of constructing an object from raw or semi-finished materials,“. Surely “all brought together from real observations” meets this definition of Fabrication.

Fabrication can also mean “the creation of a deliberate falsehood” – The Met Office taking such exception to my terminology seems more indicative of a guilty conscience than my correct use of the term – they can fabricate “data” but I am not allowed to say they do. Do Steel Fabricators tell lies about steel?

With apologies to readers for offering such a simple interpretation of this “gridding” system, I give this extremely basic explanation of the principles of how their system works minus the ” Complex statisitical calculations”. Firstly a small blank grid

I asked AI “If Point E shows a value of 10. H is E -1, I is H + 0.5, F is I – 0.5, C is E – 1.5, B is C – 0.5, A is B -1, G is equal to H and D = I +0.5. Can you populate those numbers to the above as answers to those conditions?” AI duly fulfilled a task similar to those “coffee time” puzzles in daily pnewspapaers.

So I then asked AI “Using the same relationships as above can you start the grid again from scratch and derive the values for each point starting with only knowing that Point A is 5” AI duly obliged with this below.

The point I am demonstrating is that once any set of “rules” are established it is possible to propagate numbers throughout the grid from as little as just any one single point. It does not matter how large or elaborate the grid area is, once the game rules are established all that is required are inputs from anywhere to complete the full grid.

Obviously as the Met Office states in their glorified online sales pitch, many more factors are used and the realtionships will be much more complex. However, it still all comes down to firstly, how the “model” is designed and secondly the quality and variable choice of the inputs (weather stations) used. Varying the model parameters and the end result will immediately change but also adjusting the quality and locations of the inputs will similarly change the output.

The construction of a historic temperature record of the UK for the last 150 years to indicate any temperature changes over time DOES NOT require any computer modelling at all – it is just recording historical fact. As I stipulated at Wye, providing the stations used remain unchanged, are in consistent quality sites and major instrument changes avoided then it really is a simple plot of readings. The only difficulty should relate to meeting historic provence of these requirements. A small number of geographically representative stations will be suffiecient to record change over time. The Central England Temperature series currently only operates using CIMO Class 1 Rothamsted, Class 4 Stoneyhurst and Class 4 Pershore College – just 3 locations. 302 climate averages stations of which 103 do not exist are not required to quantify change – 20 good quality sites should suffice and be both transparent and accurate. For the purpose of recording climate changes, hundreds of sites of any quality are not required and indeed can, and do, produce distorted readings such as UHI, transient heat blips recorded, and all the many problems I have reported multiple times.

Putting this back into an East Anglian context, how many currently operating weather stations are in that same area represented by SEVEN zombie grid points with no real prime data being recorded at six of them? This is where we enter the realm of real surreality – an almost virtual reality typified by computer game culture. There are more climate averages stations noted than currently functioning weather stations!

As of 2026 there are SIX sites being Weybourne, Cromer, Tibbenham, Charsfield, Neatishead and the appaling bad but record breaking Lingwood. Furthermore Neatishead was only installed in 2023 meaning that as at the end of the climate “simulation” period up to 2020 there were only FIVE. ……………Hold this thought – the undead outnumber the living.

“Weather stations come and go” was the mantra in that presentation. So how many were there at the start of that climate monitoring period way back 65 years ago in 1961? Cromer, West Raynham, Hemsby, Gorleston, Coltishall, Sprowston, Burlingham, Lowestoft, Felixstowe, Dovercourt and Walton-on-the-Naze. A total of ELEVEN – which does not really fully support the quoted line of “fewer observations in earlier decades”.

A further point (obvious to anyone familiar with East Anglian geography and weather) is that only Cromer and Weybourne (in close proximity to each other) are now on the coast which is frequently prone to North Sea cooling effects particularly at certain times in the summer. In my report on Kent I noted how Margate recorded a mean maximum over the 16 day 1976 very hot period that was 8.3°C lower than that of Goudhurst inland. This is how Lowestoft’s historic data showed that overall summer or 1976.

A summer mean maximum of just 20.1°C is not that exceptional at all. Obviously though as the Met Office has now deleted their ongoing production of synthetic (or rather removed from public view) data since site closure in 2010, {following my challenges to how it was actually being derived} it is no longer possible to do a modern day comparison……….but then again it is a climate averages site to 2020 so where are the figures used to construct those last 10 of the 30 year climate averages? How really confident can anyone be that these now deleted figures for the first two months of summer 2025 in Lowestoft (screen captured prior to deletion) are anything other than questionable fabrications. Were these two months really so much warmer or was it just a feature of a computer model deriving data from long distant inland and markedly dissimilar points?

It all becomes a very dangerously woven web.

In conclusion (and again apologies for any potential off topic rambling) Met Office climate data appears remarkably inconsistent and highly contrived. Zombie undead sites can actually dominate reality readings. The examples here of East Anglia and previously of Kent show how drifts in recording points and the production of fabricated readings from multiples of other assembly parts via computer models can be potentially producing wholly unrealistic comparisons. Sadly none of this controversy needs to exist and the potential to eliminate it is very simple – build a climate network based on the US CRN model. If the Met office will not do it perhaps others will.

Source: https://talkshop.blog/2026/10/22/east-anglia-home-of-the-zombies-a-critical-look-at-met-office-gridding-and-their-interesting-semantics/