UK Daily Maximum Temperatures – the influence of ‘microclimate’

Each day the Meteorological Office publishes the previous day’s weather extremes, including the recording station:

UK daily weather extremes – Met Office

which can be compared to the record for that date, with associated recording station, as published by

TORRO | The Tornado and Storm Research Organisation

The earliest daily record is from 16^th June 1858 in Oxford.

Using the daily maximum temperature records up to the end of 2023 a count was taken of the number of records per station. These counts can be ranked and a cumulative % table created for each datum point of % of records and % of stations.

From this a cumulative frequency graph can be constructed plotting the percentage of records (Y) against % of stations (X) for the year and by season.

The factors affecting the maximum daily temperature depend on the altitude of the station,  the time of year and the prevailing weather conditions, especially wind direction and will be influenced by the ‘microclimate’ of the station, which will tend to the regional for WMO Type 1 stations trend to  the site for Type 5.

In an ideal world of all Type 1 stations contributing to the record it may be argued that the daily maximum would be recorded at a wider range of stations than where Type 5 stations are used.

A further complication is the change in recording of temperature from Liquid in Glass Thermometers (LIGT) to Electric Resistance Thermometers (ERT).  With the adoption of the latter from 1990 the data is more prone to ‘microclimate’ effects through the Transient Response Bias of the ERTs (significantly affecting maxima), which amplify the limitations of the Stevenson Screen through Louvre Bias (enhancing maxima and reducing minima) and the limitations imposed by poor ventilation at low airflow.

In an ideal world the daily records should be spread across a large number of stations, with 1 station to 1 record. In reality this is unlikely but the number of records held by a station should be relatively small.

A Station Record Index is calculated to explore the deviation from the ideal. For each station, ranked by number of records expressed as % of total records  is subtracted from the % of that station’s rank in terms of number of observations. These values are summed and the mean calculated . If all stations have an equal number of records ie the mean that value would =0, while if all records were held by one station it would tend to 50.

In reality there is a significant difference in the number of records, with a few stations having a large number of records, suggesting that siting is a dominant factor.

In the daily records to 2023 there are 191 stations holding the 366 daily records, an average of 1.91 record per station. The top 10 are:

Rank	Station	Records
1	Camden Square	21
2	Aber	19
3	Kew Gardens	9
4	Cambridge-Observatory	8
4	Heathrow	8
6	Llandudno	7
6	Colwyn Bay	7
6	Cheltenham	7
9	Hawarden Bridge	6
10	Prestatyn	5
10	Southampton Mayflower Park	5

The data can be subdivided by season

Records	Observations	Stations	Mean Record per Station	Std Deviation	Station Record Index
Spring	92	70	1.31	1.14	10.56
Summer	92	60	1.53	1.44	15.31
Autumn	91	67	1.36	0.97	11.50
Winter	91	55	1.65	1.67	16.58
Annual	366	191	1.92	2.34	19.47

The Station Record Index shows that there is a siting effect in the UK data, which is largest in Winter and Summer.

Separating data between pre and post 1990 (the adoption of ERTs) we find the following

Records	<1990	>=1990	% >=1990
Spring	63	29	31.52%
Summer	54	38	41.30%
Autumn	58	33	36.26%
Winter	54	37	40.66%
Annual	229	137	37.43%

And while not significant, there is a relationship between the % of post 1990 stations in the data and the Station Record Index.

If we look at the daily maxima since June 2022 the pattern is even clearer with 144 stations holding 683 daily records, an average of 4.74 daily maxima per station.

June 2022 to Date	Observations	Stations	Mean Record per Station	Std Deviation	Station Record Index
Spring	137	64	2.14	1.58	18.01
Summer	184	61	3.02	4.18	25.81
Autumn	182	71	2.56	2.14	20.66
Winter	180	67	2.69	3.42	23.17
Total	683	144	4.74	6.00	28.27

The top 10 are

Rank	Station	Records
1	Heathrow	31
1	Scilly St Mary’s	31
3	Cavendish	25
4	Wiggonholt	24
5	Pershore	20
6	Santon Downham	19
6	Plymouth	19
8	Gosport	18
9	Cardiff	17
10	Kew Gardens	16

Summer records are now the most influenced by siting followed by Winter, and the effects are more pronounced than in the Daily Records shown above.

This raises questions about the quality of the UK data, especially when considering Climates’ Change (not Climate Change as there are 25 regional climates across the earth’s surface) and show the continued impact of using ERTs rather than LIGTs.