Kew Gardens Revisited – “For some reason the screen was not at the standard height – well below the standard 1.25 m above ground level”

Fellow researcher into the questionable practices of the Met Office, Dr Eric Huxter, recently forwarded me his latest work on the Kew Gardens weather station which can be viewed on his excellent blog here.

https://frayedendsblog.wordpress.com/2026/05/03/kew-gardens-mayday-again/

Eric has devised a “bench-marking” system using the Class 1 Rothamsted site to monitor the likely inaccuracy of daily extremes temperatures recorded at lower CIMO rated sites. This seminal work is a must read and highlights the highly significant changes in readings since the implementation of Platinum Resistance Thermometers. Simultaneous to receiving this I was given online access (paid for) to many more detailed historic aerial images on which I experimentally browsed the Kew site. My findings were quite a revelation – Kew most certainly does not offer a reliable dataset of temperature readings.

The Kew Gardens site is often held up as a “flagship” location by the Met Office and indeed they often justify alleged CIMO Class 3 Heathrow readings by referring to similarity with Kew. –

“If you compare Heathrow to nearby Kew – which is eight miles away – there is hardly any temperature difference between the two.“

Eric’s research suggests that Heathrow’s variations from bench mark indicate the Heathrow site is actually more realistically a widely inaccurate Class 5. My own personal assessment is that climate reference readings should not be used for this wholly unnatural site. Eric goes on to show that at particular times (notably May Days) Kew has an uncanny ability to produce extremes almost “on cue”.

So to the astonishing evidence my research uncovered. Firstly that headline quote is taken from an investigation by internationally recognised meteorologists, the late Philip Eden (then principal of the Royal Meteorological Society) and Dr. Stephen Burt into the UK 2003 heatwave when the national record high temperature was awarded to the obviously unacceptable Faversham site.

“Kew, Royal Botanic Gardens,Greater London

The station at Kew is maintained by the Royal Botanic Gardens. A photograph of the site taken by one of the authors (SB) on 17 August 2003 is shown in Fig. 7. The site is within a new enclosure area, approximately 11 m × 5 m; a sturdy wooden slatted fence
has replaced a dense low-level hedge, often criticised in the past. A Campbell Scientific AWS was added to the enclosure in 2002, although this has not yet been brought into operation (Dr V. Sarasan, Kew, Royal Botanic Gardens, personal communication, 24 February 2004). The mercury-in-glass thermometers were housed in a large Stevenson screen, although for some reason the screen was not at the standard height – the thermometers were well below the standard 1.25 m above ground level. (A new screen at the standard height has since been installed.)
The site is in an open position in a public area within Kew Gardens. By the time of the heatwave the grassland surrounding the
site had become parched owing to the prevailing dry conditions and higher than usual public access in the hot weather. This in addition to the low screen height may have led to the reported maximum of 38.1 °C on 10 August being slightly in excess of a hypothetical ‘perfect site’ in the same location, but the value is well corroborated by other readings in west London of between 37.5
and 37.9 °C. {Ed note: what evidence is there that any of these other sites were any more reliable?}
The Met Office inspected this site on 12 September 2003. The maximum thermometer was removed for calibration checks and
was found to be reading accurately. The Met Office’s official conclusion was that there was no reason to doubt the validity of the
Kew reading.” {my bold}

I find this quite astonishing. The screen was noted as “well below the standard 1.25 m above ground level. ” Although the exact height was not indicated, “well below” does not suggest a few millimetres but rather several centimetres at least. Would this make a difference? Who on earth could get such a simple issue so wrong? Since when was it the wrong height? Was the site never inspected and if so why could Stephen Burt notice it but seemingly not the official Met office inspectorate?

As I am guaranteed to say such readings must be ruled out on simple basic “rules breaking” alone, I asked AI for the impartial reasons quoted in meteorological terms -, after all if a “100” metre running track was only 95 metres long there would be no world records awarded there!

Main Problems with Lower Height

Here are the key issues:

• Greater influence from ground heat/radiation (especially daytime maxima): The ground (soil, grass, pavement, etc.) absorbs and re-radiates solar heat. Closer to the surface, sensors pick up more of this upward long-wave radiation and warmer air in the shallow boundary layer. This leads to overestimated daytime high temperatures (potentially by several degrees in strong sunshine/low wind). Studies and observations show larger diurnal ranges at lower heights. scispace.com
• More extreme minimum temperatures at night: Radiative cooling is stronger near the ground, especially on clear, calm nights, leading to colder readings (stronger surface inversions). This exaggerates nightly lows. scispace.com
• Increased sensitivity to surface type and micro-environment: Readings become more affected by the immediate ground cover (e.g., wet grass, snow, bare soil, asphalt). This reduces comparability across sites and violates WMO/Met Office standardization for climate records and forecasts. Snow depth or vegetation changes would also have a bigger impact. weather.gov
• Altered ventilation and response times: Lower placement can change airflow patterns around/into the screen (e.g., more obstruction from ground or low vegetation), potentially worsening radiation errors in calm conditions or increasing lag in responding to temperature changes.
• Reduced data comparability and homogeneity: Historical and international records use the standard height. Lower readings would introduce biases, making it hard to compare with other stations, detect climate trends, or feed into models. Diurnal ranges are systematically larger at lower elevations. scispace.com

Other Practical Issues

• Increased risk of damage or contamination — from animals, mowing, splashing, dust, or debris.
• In urban or varied terrain — effects are amplified by heat islands, paving, or structures.

In extreme cases (e.g., very close to hot surfaces), differences of several °C have been noted compared to standard height. The 1.25 m standard minimizes these “purely local influences” to provide a more consistent measure of air temperature at screen level.

For precise work, always follow WMO or national guidelines on siting, exposure, and height. Lower heights might be used for specialized applications (e.g., frost monitoring or pedestrian-level studies), but they are not suitable for standard meteorological records.” {my bold}

So is it possible to ascertain what the low height actually was? I guess that is largely the sort of verification that the Surface Stations Project is all about – so here goes. This is that image taken by Dr. Stephen Burt on 17/8/2003.

Note the base of the wider old screen is lower than the top cross rail of the perimeter fencing. Now here below is an image subsequent to the screen change (summer 2006) with the new narrower screen notably higher than the same rail of the unchanged fencing.

My best estimate is coming in at 250mm / 10 inches too low and very definitely causing significant differences in recorded temperature especially on high solar insolation, calm days such as 19/8/2003.

How more damning could this be? And yet the Met Office has verified all the readings from Kew – is it just me or does anyone else see the obvious comparisons with the Volgwagen emissions scandal? Known inaccurate and almost certainly elevated readings have been willfully archived and are on file to this present day. If the Met office moved into the used car sales business, who would trust buying a car off them?

The above revelation and Dr Huxter’s May Day analysis reminded me of my own investigations into May Day 2025 and the invaluable and diligent work of X poster John Dodders. John passes through St James’s Park going to work on a regular basis and noted obvious modifications to the area for VE Day celebrations. By publishing detailed imagery and pressing the Met office directly he managed to get them to admit to the obvious corruptions to readings and have them suspended – I very much doubt the Met Office would have taken action had it not been for John’s public naming and shaming. He supplied me with images as detailed in my report at the time. John’s online comment below rather nails it! And yes that really is the Stevenson screen behind the portaloo and surrounded by hoarding.

Thus I wondered if Kew Gardens were similarly frequently compromised by “Special Events” or other significant site modifications over time. My new and more extensive aerial images access proved multiple exceptionally major alterations to the immediate environment and surroundings of the screen.

My headline image dates from 19th July 2013. The surrounding structures are almost equally as bad as those compromising St James Park last May Day but were readings suspended? Despite the temperature readings in the period being of no more than comedy value they most certainly were archived. Column “I” for maximums and “J” for minimums.

The subsequent image of the 28th August 2013 shows the structure still there and proves a long term readings corruption. In fairness the Met Office does appear to have suspended readings for a period from 2nd August to 1st September 2013 but readings predating August still stand thus for example how genuine was the 33.2°C on 27/7/2013? I certainly would not accept its provenance.

Are such site modifications rare? – This from 13th July 2018

Some form of major staging to the north and similar development to the south with access ways in between alongside the enclosure. Whatever event was on will unquestionably have resulted in numerous people within almost immediate touching distance of the screen and no readings were suspended.

Even the surrounding trees and shrub beds have been the subject of extensive changes over time as demonstrated in this image from 2006.

I do not wish to overload our wordpress hosts with too many images stored thus I recommend any one who can to use Google Earth Professional to view the historic images from this site to view the numerous and quite drastic changes this site has experienced to confirm for themselves.

These changes explain the varying CIMO classifications I noted in my original report. At times this site has been rated Class 4 but the last one I noted from my CIMO listing obtained under Freedom of Information request was shown as a fully reliable Class 2. However, a more recent FOI for Class 1 and 2 rated stations only, as shown on my Syerston report, did not include Kew thus indicating it has yet again been (almost certainly correctly) down-rated to relative junk status of Class 3 or lower.

A final alarming aspect was the revelation that Kew is NOT a long term site. Whilst there used to be the separate Kew Obsevatory site which I noted in my earlier review, there is still the suggestion that the Kew Gardens site is long term one with readings going back to 1910. This is not the case.

The archives only show a few days of 1948 (1st July to 30th September) then abruptly stopping before resuming in 1981.

I could find nothing in the manual archives pre-digitalisation and the black and white 1960s aerial image of the area shows no screen at all at the present location nor anywhere in the immediate vicinity.

In summary I feel Kew highlights multiple unacceptable failings on the part of the tax payer funded Met Office. A screen set too low that despite the thermometers being checked was only revealed by independent inspection by the Royal Meteorological Society. Had they not intervened the screen could seemingly be set too low to this date with no inspectors noticing what was obviously apparent. Readings are regularly being corrupted by seemingly unnoticed extraneous heat sources and events rendering observations useless that are still archived and quoted as accurate. Worst of all even when their failings have been clearly pointed out by expert peers, the Met Office (being a law unto itself) simply dismisses any criticisms and declares false readings as almost Gospel truth.

It seems hardly surprising the Met Office quotes Kew in support of Heathrow – they are as bad as each other, though probably not quite as bad as St James’s Park! I will shortly be doing a London-wide review of sites from 1960 to 2020 (as I did for Dumfries and Galloway) to demonstrate how the data is being systematically manipulated not only by low grade sites but also the systematic alteration of sample points.

I challenge the Met Office to justify such malpractices demonstrated at Kew.

Source: https://tallbloke.wordpress.com/2026/05/10/kew-gardens-revisited-for-some-reason-the-screen-was-not-at-the-standard-height-well-below-the-standard-1-25-m-above-ground-level/