Stoodleigh:Dryhill DCNN8879 – A newly installed site demonstrating low standards and biased site selection.

50.9617986 -3.519111 (elevation 230 metres amsl) Met Office CIMO Assessed Class 4 Temperature records from 18/3/2024

Stoodleigh is a small village 6 km/3.6 miles north-north east of Tiverton in Devon – a county already blessed with an unusually high density of Met Office weather stations and home to its headquarters. In my original letter to Peter Kyle (then Government minister with oversight of the Met Office) I cited Stoodleigh as an example of newly opened weather stations being of poor standard and specifically indicating no evidence that the Met Office was seeking to improve on their low standards. This report is to detail these poor standards and further investigate the apparent motives behind such new location “selection” processes.

A reminder of what the Met Office used to publish on their website but is now deleted as, no doubt, it was deemed an “inconvenient truth”.

The first concern regarding this Stoodleigh site lies within its further identifying name – “Dry Hill”. What may appear flat in aerial imagery is often not and the word “hill” is a frequent giveaway. The Ordnance survey sheet for the area indicates this is a very sloping site indeed. The contour lines reflect 10 metre intervals. The CEDA details show the screen is at 230 metres elevation whilst the 150 metre contour line crosses the watercourse to the southwest at 230 metres distant.

In my report on Portglenone (where I felt the site was unreasonably marked down to class 4) I used Elevation Finder to identify a 2.9 metre greater height at 100 metres from the screen – a negligible slope of no material concern. Using the same online tool at Stoodleigh shows a 25 metre fall over 100 metres to the south west and a 10 metre rise at 100 metres north east. This site is immediately, what the Met Office itself terms, an “undesirable site” due to its very significant slope alone.

To assist with bearings I asked AI to overlay compass points on this occasion as below.

To further identify distances I delineated a 10 metre radius circle below centred on the screen. This image is known to date from 26th September 2025. With the shadows running almost due north, it will have been taken around solar midday. From these details together with latitude, the angle of incidence of sunlight can be calculated. Adding in the 12.5 metre distance scaled from the online measuring tool of the Stevenson screen to the hedge allowed an adequately accurate calculation of the height of the hedge and the trees. AI ran these calculations as between 1.2 and 1.5 metres for the hedge and between 3.8 and 4.4 metres for the tree due south of the screen.

The recently maintained hedgeline runs from east southeast to west northwest. The neighbouring buildings and woodland will ensure all winds from east north east around through south to west north west will be significantly calmed. The natural slope uphill to the north east will limit sunshine hours and will also allow colder air to “drain” downhill in still conditions and be entrapped by the notably dense, 2 metre wide hedge potentially creating a frost pocket. Throughout the late autumn, winter and early spring the tree fully shading the screen will be extensive for much of the day.

Below is an image from 31st December 2002 (similarly midday and long predating the screen installation) to demonstrate the length of shadows cast by trees further south. Also note, however, that the modern day trees within the hedging have grown very substantially since.

In addition to all these compromising problems, the 21st century Platinum Resistance Thermometer will be mounted within a Victorian designed screen that despite known problems in low windspeeds will not have WMO recommended artificial ventilation assistance to overcome the problem. The site will not use WMO recommended period averaging of instant readings and, in the event of the site registering a national record a single point reading will be sufficient to verify as at Lingwood.

Then there is the question of observation reliability. Given that all stations are now fitted with PRTs and Data Loggers and hence have a power supply (often a solar panel and battery pack) why are they not all automated? All that is required is a mobile phone link and even for rural areas like this a Starlink or similar connection would provide data transfer if conventional mobile signal was not available. One of the reasons quoted in the past for closure of sites was difficulty in automating, a problem that no longer occurs – technology fixed it. This below is the sort of problem that automation rarely suffers from:

Though not obvious at my first glance, I then noticed that those 2 “NA” entries indicate no readings from 28th August to the 24th September inclusive – 27 straight missing days. Similar gaps have appeared since and bear in mind this site only started readings in March that year. Not the best of records by a long way.

I have gone to great length above to emphasize the point that the Met Office do not seem to be making any efforts at all to improve on either the area representation of their sites (yet another Class 4 at very best but in my view really Class 5 for most of the year) nor overcome known problems causing inaccurate readings. I will in fact go further to point out their overall group of weather station’s quality is severely deteriorating and the representation worsening by design.

I started this review by referring to the high density of sites in Devon and I have extensively covered the changing mix of station locations over time notably London , and also in south west Wales, Dumfries and Galloway, Kent, Derbyshire and Fife. The huge reduction of air temperature monitoring sites from over 630 in 1994 down to 380 in 2024 is further confused by very many more than the difference of 250 closing with many newer and selectively sited ones replacing those closures. These graphs shows the significant variations over time and simultaneously rather call into question dramatic alterations to rainfall data since the 1960s given the order of magnitude rise in number of sampling points over such a short period.

So what determines where new “climate reporting” weather stations are located? One factor that can definitely be ruled out is network improvement – the reverse is the case. It should be deemed totally irresponsible to locate a new site in 2019 in a Victorian walled garden but that is exactly what has happened at Kingston Maurward. Such an inappropriate choice simply cannot be accidental considering the Bisterene walled garden site was chosen in 2026 for an additional site when the gardener simply requested one.

As I have also noted at Bewcastle , North Somercotes and several others it appears that “motivated” individuals requesting a site get one regardless of site quality or any obvious specific meteorological or climatological need. In fact Bewcastle indicated a poor immediate siting was intentional. How does this possibly apply to Stoodleigh?- This may of course be perfectly innocent but the site owner has a professional interest in ecology , attended the University of East Anglia with its associated (in)famous Climate Research Unit and is an employee of Natural England with particular emphasis on Rewilding and the reintroduction of Beavers.

This professional association in itself should not be a cause for exclusion but neither should it be a cause of preference or inclusion. Stoodleigh:Dryhill is under 4 miles from Huntsham, 9 miles from Liscombe with its McDonalds, 14 miles from Lapford (not yet reviewed), Hemyock, Nettlecombe with its parked minibuses and Dunkeswell with its silage heating and 17 miles from Exeter the airport of course. Already having seven weather stations within a seventeen mile radius of Stoodleigh was it such an imperative to have yet another whilst other parts of the UK have vastly larger areas with none at all? Quite a remarkably concentrated cluster as indicated on the current site locations map.

Conversely I am confident that if I were to apply to the Met Office to upgrade and adopt the very important agricultural/viticultural existing and well sited weather station close to my home, they would find reasons to decline. Despite there now being only one official site within a 14 mile radius (Langdon Bay) of my location, one of the most densely populated areas of the UK, representative of a wide inland area and with all current east Kent sites on the coast, I doubt it would be considered “suitable”. There used to be sites at Elmstone, Anvil Green, Wye, Herne Bay, Whitstable, Folkestone and Dover itself all within 14 miles of my location and three of those inland (Elmstone, Anvil Green and Wye ) would certainly merit CIMO Class 1 or 2. Notably though, all three of these sites recorded the rural and natural climate of the area and would not provide the highly modified numbers that now seem so sought after. Here is an extract of my surroundings from a “clickable” map indicating the vast majority of most current and former sites throughout the UK – well worth referencing to see just how things have changed over time. Compare this:

To this – with all points in between “guestimated”:

In conclusion Stoodleigh:Dryhill is a typical example of the low quality weather stations that the Met Office is continuing to adopt even in the mid 21st Century. These are directly replacing former sites of much better provenance and maintenance. The organisation is failing to adopt recognised WMO recommendations for siting, instrumentation sampling and averaging and continues to use outdated naturally ventilated screens it knows are not accurately recording and producing artificially high spikes in readings. Site closures due to the requirements of automation are not being reversed taking advantage of modern technologies. The number and location of sample points are clearly being controlled to create a sampling bias that demonstrates ongoing temperature elevation where none may exist. All of this cottage industry standard is then being statistically tortured by computer modellers who likely have limited understanding of how the raw data is supplied and its shortcomings.

A follow on post is a detailed analysis of the quality of data from the previously reviewed and alleged “Historic Station” of Sutton Bonington to demonstrate just how misleading much Met Office raw data actually is and how the likes of Stoodleigh and others validate such poor data.

Source: https://talkshop.blog/2026/09/06/stoodleighdryhill-dcnn8879-a-newly-installed-site-demonstrating-low-standards-and-biased-site-selection/