State of the UK Climate 2023

This report provides a summary of the UK's weather and climate through the calendar year 2023, alongside the historical context for a number of essential climate variables. This is the tenth in a series of annual 'State of the UK Climate' publications, published in the International Journal of Climatology (IJC) since 2017, and an update to the 2022 report (Kendon et al., 2023). It provides an accessible, authoritative and up-to-date assessment of UK climate trends, variations and extremes based on the most up to date observational datasets of climate quality. Success is by no means a necessary result of efforts to secure it. Although, therefore, with each consecutive year I have adopted additional precautions against any errors either of observation or computation creeping into this work, it by no means follows that I shall continue to be as successful in keeping them out as I have hitherto been; in fact, in spite of every care, it is hardly reasonable to hope for a continuance of such remarkably small tables of errata as there have been of late years. This report provides a summary of the UK's weather and climate through the calendar year 2023, alongside the historical context for a number of essential climate variables. This is the tenth in a series of annual 'State of the UK Climate' publications, published in the International Journal of Climatology (IJC) since 2017, and an update to the 2022 report (Kendon et al., 2023). It provides an accessible, authoritative and up-to-date assessment of UK climate trends, variations and extremes based on the most up to date observational datasets of climate quality. The majority of this report is based on observations of temperature, precipitation, sunshine and wind speed from the UK land weather station network as managed by the Met Office and a number of key partners and co-operating volunteers. The observations are carefully managed so that they conform to current best-practice observational standards as defined by the World Meteorological Organization (WMO). The observations also pass through a range of quality assurance procedures at the Met Office before application for climate monitoring. Time series of near-coast sea-surface temperature and sea-level are also presented, and in addition there is a short section on phenology that provides dates of 'first leaf' and 'bare tree' indicators for four common shrub or tree species plus several other indicators. The reliance of this report on these observations highlights the ongoing need to maintain the observation networks, in particular the UK land weather station network, into the future. This is vital if the UK's climate monitoring capability is to be continued. National and regional statistics in this report are from the HadUK-Grid dataset which is the principal source of data (Hollis et al., 2019). County-level statistics are also provided in some sections. Temperature and rainfall series from this dataset extend back to 1884 and 1836, respectively. Details of the datasets used throughout this report and how the various series which are presented are derived are provided in the Appendices A–C. The report presents summary statistics for the most recent year 2023 and the most recent decade 2014–2023 against the 30-year standard climate normal period 1991–2020 and the baseline period 1961–1990, following World Meteorological Organization climatological best practice (WMO, 2017). The baseline reference period 1961–1990 provides a consistent reference period used throughout the series of State of UK Climate reports and more widely for historical comparison, climate change monitoring and climate modelling. The full series provides longer-term context, while a comparison is also made to centennial averages for the Central England Temperature (CET) series. The decade 2014–2023 provides a 10-year 'snapshot' of the most recent experience of the UK's climate and how it compares to historical records. Differences between 2014 and 2023, and the 30-year reference periods may reflect shorter-term decadal variations as well as long-term trends. For this annual publication, the most recent decade (currently 2014–2023) changes every year, while the most recent 30-year reference period (currently 1991–2020) changes every decade. The reference period 1991–2020 does not overlap with 1961–1990. Throughout the report's text the terms 'above normal' and 'above average' and so on refer to the 1991–2020 reference period unless otherwise stated. The majority of maps in this report show the year 2023 relative to 1991–2020—that is, they are anomaly maps that show the spatial variation in this difference from average. Some anomaly maps relative to 1961–1990 are also included. Maps of actual values are in most cases not displayed because these are dominated by the underlying climatology, which is strongly influenced, for example, by elevation and distance from the coast. For this report, this is of a lesser interest than the year-to-year variability. These data are presented to show what has happened in recent years, not necessarily what is expected to happen in a changing climate. However, two figures showing UK Climate Projections (UKCP) for annual mean temperature and rainfall are included to provide future context to 2100. Values quoted in tables throughout this report are rounded, but where the difference between two such values is quoted in the text (e.g., comparing the most recent decade with 1991–2020), this difference is calculated from the original unrounded values. The executive summary and main text in the report stated: 'The most recent decade 2013–2022 has had 4%/7% fewer days of air and ground frost per year compared with 1991–2020 and 15%/23% fewer than 1961–1990'. These were incorrectly reported as percentages instead of actual days. The corrected sentence is 'the most recent decade 2013–2022 has had 4/7 fewer days of air and ground frost per year compared with 1991–2020 and 15/23 fewer than 1961–1990'. The main text stated '… representing substantial changes in the UK's climate with the number of ground frosts across Wales, Scotland and Northern Ireland decreasing by a quarter or more (Figure 21)'. The corrected text is '… representing substantial changes in the UK's climate with the number of air and ground frosts decreasing by around 2 and 3 weeks, respectively (Figure 21)'. The correct percentage values (not reported) were 7%/7% and 23%/20%. Figure 21 and caption were correct as presented and the main report conclusions are unaffected. This correction is published alongside the State of UK Climate 2022 report (references; Anon 2024). We welcome any suggestions for future publications of this report. Please send any feedback to the Met Office at email protected. This State of the UK Climate report was supported by the Met Office Hadley Centre Climate Programme funded by DSIT. Figure 1 shows monthly mean sea-level pressure anomalies for the 12 months of 2023 relative to the 1991–2020 average, using the ERA5 reanalysis (Hersbach et al., 2020). This provides an indication of atmospheric circulation patterns for each month overall. Pressure anomalies on the charts are scaled equally across all 12 months for consistency; however, winter months typically have larger pressure anomalies than summer months. Figure 2 shows HadUK-Grid UK monthly mean sea-level pressure anomalies relative to 1991–2020 for the most recent decade 2014–2023 (120 months). These charts and the time series illustrate the characteristic large variability in atmospheric circulation patterns in the UK's climate, which tend to be masked out on a seasonal scale—although inevitably the monthly charts will also mask out daily variability too, since changes in weather type do not neatly coincide with calendar months and each month will usually comprise a mixture of weather types. A large high-pressure anomaly was centred over the UK in February and the UK monthly mean sea level pressure anomaly (12.6 hPa above the 1991–2020 February average) was the highest for any calendar month since February 2012. Pressure was also higher than normal in April, May and June with a high pressure anomaly extending from eastern Europe and Scandinavia to the UK. In February, May and June in particular the weather was often settled and dry. Pressure was lower than normal for the other months of the year, most notably in March, July and October to December, each of these months seeing a low-pressure anomaly extending from eastern Europe to the UK, and in March including much of the north Atlantic. During these months, the jet stream was often stronger and shifted further south than normal, bringing unsettled weather and stormy at times. In summary, the second half of the year was much more unsettled, with fewer spells of fine-settled weather than the first half. The lowest pressure anomalies for any month of the most recent decade 2014–2023 were, by far, January and February 2014 (−15.7 and −21.7 hPa, respectively). These 2 months had the lowest average UK pressure anomalies of any month in the series from 1961; very much lower than any month in 2023 (Figure 2). These 2 months formed part of the exceptionally stormy and wet winter of 2013–2014 (Kendon and McCarthy 2015)—this being the UK's wettest winter on record in the series from 1836. Figure 3 shows the winter North Atlantic Oscillation (WNAO) index from 1850 to 2023 inclusive (Appendix A1 provides details of the WNAO index). (Note here and throughout the report winter refers to the year in which January and February fall.) This index is a measure of the large-scale surface pressure gradient in the North Atlantic between the Azores and Iceland, which determines the strength of westerly winds across the Atlantic, and is the principal mode of spatial variability of atmospheric patterns in this region. When the pressure difference is large, the WNAO is positive and westerly winds dominate with stronger and more frequent storms. When the pressure difference is small, the WNAO is negative with an increased tendency for blocked weather patterns, reducing the influence of Atlantic weather systems. The WNAO index for 2023 was negative, although not markedly so (−0.6), one of only two WNAO negative winters of the most recent decade 2014–2023, the other being 2021 (−1.0). Other earlier winters in the series have been much more negative, especially winter 2010 (−3.1). A WNAO negative winter would tend to be associated with a cold, dry winter. Winter 2023 was slightly milder than average (anomaly +0.2°C) and drier than average (anomaly 85%). February 2023 was very dry in the south of the UK with blocked weather patterns (Figure 1). The UK has experienced a run of mild, wet winters in the most recent decade, consistent with a positive phase of the WNAO, including the very wet winters of 2014, 2016 and 2020—but not 2023 (Figure 39). Overall, the WNAO index shows a large annual variability but also decadal variability with periods of mainly positive phase (e.g., the 1910s–1920s, 1990s and 2010s) and negative phase (e.g., the 1960s) which are represented by the smoothed trend line in Figure 3. Figure 4 shows the summer North Atlantic Oscillation (SNAO) index from 1850 to 2023 inclusive (Appendix A1 provides details of the SNAO index). Similar to the WNAO index, this is a measure of large-scale climate variability in the North Atlantic based on the surface pressure gradient, but based on a more northerly location and smaller spatial scale than the winter counterpart, reflecting the more northerly location of the Atlantic storm track in summer. The SNAO index for 2023 was also marginally negative, (−0.7). A SNAO-negative summer would tend to be associated with lower temperatures and higher rainfall. Summer 2023 was warmer than average (anomaly 0.8°C) and wetter than average (anomaly 113%). June was fine and settled but July and August generally unsettled. Recent notably negative summers include 2015 (−1.5) and 2019 (−1.8) and positive summers 2018 (2.1) and 2021 (1.6). As with its winter counterpart, the SNAO shows periods of mainly positive phase (e.g., the 1970s to 1990s) and negative phase (e.g., the 1880s and 1890s), with a run of wet SNAO negative summers from 2007 to 2012. The recent large fluctuations in the SNAO reflect some markedly contrasting summers—illustrating the UK's large annual variability in the weather and atmospheric circulation patterns. Importantly, neither the WNAO or SNAO can fully explain the variability of UK winters or summers because of the complexity of weather types and associated temperature and rainfall patterns through the season across the UK's relatively small spatial scale. The UK mean temperature (Tmean) for 2023 was 9.97°C, which is 0.83°C above the 1991–2020 long-term average. 2023 was the second warmest year on record in the UK series from 1884 behind only 2022 (10.03°C). The highest anomalies, exceeding 1.0°C relative to 1991–2020, were across Northern Ireland and western England and Wales. The UK mean temperature was 1.66°C above the 1961–1990 baseline long-term average (Figure 5). This was the warmest year on record for both Wales and Northern Ireland, the second warmest for England and third warmest for Scotland. The Republic of Ireland also had its warmest year in a national series from 1900 (Met Éireann, 2023). While many western counties had their warmest year on record, further east for the majority of England and Scotland the warmest year remains 2022, with other record years 2003, 2006 and 2014 confined to the far north and west (Figure 6a,b). Appendix A8 describes these county areas. The UK annual mean daily maximum temperature (Tmax) for 2023 was also the second highest on record, behind 2022, whereas it was the UK's warmest year on record for annual mean daily minimum temperature (Tmin) (6.32°C, 0.79°C above the 1991–2020 average, next warmest 2014, 6.26°C). Annual Tmax and Tmin temperature anomalies across the UK were similar, but with slightly lower anomalies across Scotland for Tmin (Figure 7). Figures 8 and 9 show seasonal and monthly Tmean anomalies for the UK for 2023. Table 1 shows monthly, seasonal and annual actual and anomaly values and ranks for the UK and countries for 2023. Figure 10 shows UK monthly mean temperature anomalies for the most recent decade 2014–2023. Eight of the 12 months of the year were warmer than average, with anomalies in June and September (+2.5°C and +2.2°C) the highest for the UK since December 2015. Four months of the year were in the top-ten warmest for the UK overall (for Tmean) in series from 1884—February (ranked 5th), May (8th), June (1st) and September (equal-1st)—although this is also true for 2006 and 2017, and compares to 6 months in 2022. All countries of the UK had their warmest June, while England and Wales also had their warmest September. Unusually, June was the warmest month of the year; the last time this happened was in 1966 (in 1970 June and August were equal-warmest). 1890 is the only year in the UK series where September was the warmest month. Three months of the year were slightly cooler than average, with the most negative anomaly a modest −0.3°C in July. All other years in the most recent decade have had at least 1 month with a significantly more negative anomaly than this (Figure 10). While winter and spring temperatures were near average overall, summer and autumn were notably warm (ranked 8th and 6th warmest, respectively), although for both seasons 2022 was warmer overall. Autumn anomalies were particularly high in the south, exceeding 1.5°C. Table 2 shows monthly, seasonal and annual Tmean anomaly values for the UK and countries for the most recent decade 2014–2023 against both 1961–1990 and 1991–2020. The most recent decade 2014–2023 has been 0.42°C warmer than 1991–2020 and 1.25°C warmer than 1961–1990, with the most warming across England, then Wales, and slightly less across Scotland and Northern Ireland. Comparing the most recent decade 2014–2023 to 1961–1990, all months have warmed for the UK by between +0.9°C for October and +1.7°C for February, with the greatest warming in England for February (+1.9°C) and the least in Scotland for October (+0.5°C). These statistics reflect annual and decadal variability in the UK's climate in addition to the ongoing warming due to climate change. Figure 11 shows a time series of annual Tmean anomalies for the UK from 1884 to 2023 inclusive, and shows that the main period of warming for the UK has been from the 1980s onward at a rate of ~0.25°C per decade (1°C in 40 years, i.e., two grid lines along for every one grid line up on the graph.) The two warmest years in the series are 2022 and 2023. Six of the 10 years 2014 to 2023 have been within the top-ten warmest for the UK overall (2014, 2017, 2018, 2020, 2022 and 2023), and this is the warmest 10-year period in the UK series. All top 10 warmest years in the UK Tmean series have occurred in the 21st century; none of the top 10 coldest years has occurred in this century, the most recent of these being 1963. The coldest year this century (2010) is ranked 22nd coldest in the UK series; every other year this century falls in the top third warmest years in the series. Thirty years ago, the warmest year in the UK series from 1884 to 1993 was 1990. Half of the years since then (i.e., for the period 1994 to 2023) have subsequently been warmer than 1990. Figure 12 shows annual mean maximum and minimum temperatures for the UK from 1884 to 2023 as anomalies relative to 1991–2020. These series are highly correlated (R2 0.83). Warming is slightly higher for Tmax than Tmin with the most recent decade (2014–2023) 1.41°C warmer than 1961–1990 for Tmax and 1.09°C for Tmin. The UK average diurnal temperature range (DTR, Tmax − Tmin) is ~7°C. There has been a small recent increase in the average DTR but to levels similar to those observed before the mid-20th Century (Figure 13). Figure 14 shows UK seasonal mean temperature for all four seasons. As with the annual series, the seasonal series show large inter-annual variability and some decadal variability, with a marked increase in temperature across all four seasons from the 1970s or 1980s onward. The most recent decade 2014–2023 has seen 16 seasons out of 40 in the top-ten warmest in their seasonal series (more than one in three): five in winter (2014, 2016, 2019, 2020, and 2022); four in spring (2014, 2017, 2020, and 2022); three in summer (2018, 2022, and 2023); four in autumn (2014, 2021, 2022, and 2023)—whereas none have fallen in the top-ten coldest. The uncertainty in these statistics is principally a function of the number and distribution of stations in the observing network which varies through time. For monthly, seasonal and annual averages the standard error is <0.1°C and consequently the uncertainty is much smaller than the year-to-year variability. In this report monthly and seasonal temperature data are presented in tables to the nearest 0.1°C and annual temperature data to 0.01°C. More information relating to the uncertainties and how they are estimated is provided in Appendix B2. A key illustration of the implications of the UK's warming climate is the increased number of high-temperature values which have occurred in monthly, seasonal and annual temperature series in recent years. This question was examined a decade ago (Kendon 2014) with a similar analysis repeated here. Monthly, seasonal and annual temperature values are often reported where they fall within the top-ten of the series at country, national, regional or county level. Figure 15a examines the frequency of occurrence of these values through each decade of the monthly, seasonal and annual mean temperature time series across 97 county areas of the UK. These county areas are described in Appendix A8 and shown in Figure A6. Although these areas vary in size, they are useful for monitoring at this spatial scale. The figure counts the total number of such values across all counties by decade throughout the monthly, seasonal and annual series, converted to a percentage of all monthly, seasonal and annual top-ten values (which is the same as the percentage of the total number of records for each decade). The average across the 14 decades of the 140-year series is 7.1% per decade (100/14, that is, what would be expected on average if these values were evenly spaced). However, as a consequence of the UK's warming climate, the occurrence of top-ten warmest values in the series is dominated by the last three decades, which together account for 52% of all top-ten warmest monthly values, 70% of all top-ten warmest seasonal values and 93% of all top-ten warmest annual values. The most recent decade alone accounts for 25%/33%/50% monthly/seasonal/annual values. So, overall, a quarter of all months, a third of all seasons and half of all years of the most recent decade 2014–2023 have been in the top-ten warmest in their respective monthly, seasonal or annual series (e.g., in the top-ten warmest Januaries, etc.), or an alternative framing, based on the observational series, is that the chance of any year of the most-recent decade 2014–2023 falling within the top-ten warmest in the series is one in four for each month, one in three for each season and one in two for each year. The UK's climate is dominated by natural variability, with the warming trend becoming increasingly apparent as the data are averaged over time, due to the 'signal-to-noise' ratio—warming trend versus variability—increasing. This is because each season comprises three constituent months, and each year four constituent seasons. Although the separate months will not be completely independent, the variability will tend to reduce and the climate change signal will be retained when averaging over these 3-month and 12-month periods. The signal-to-noise ratio is, therefore, increased because the noise (i.e., the variability) is reduced. In statistics this effect is known as the Central Limit Theorem (Fischer 2011). The increased proportion of top-ten annual values compared with seasonal and monthly values in the most recent decades is a direct consequence of this effect. Figure 15b similarly explores the frequency of the top-ten coldest values in the county mean temperature series, with the last three decades together accounting for <3% of monthly and seasonal values and <0.5% of annual values, and the occurrence of these values much more prominent in the early decades of the series. No seasons or years of the most recent decade 2014–2023 have been in the top-ten coldest in their respective series for any county area of the UK, and virtually no months (4 values out of 11,640 records 0.03%). Taken together, these figures show the dramatic consequence of the UK's changing climate on occurrences of both warmest and coldest monthly, seasonal and annual top-ten temperature values. Figure 16 shows annual Tmean for the CET series from 1659 and for England from 1884 to 2023. CET represents a region bounded by Hertfordshire, Worcestershire and Lancashire. The temperature trends in the HadUK-Grid dataset shown in Figure 11 are confirmed by the close consistency with the CET series. 2023 was the second warmest year in the CET series with an annual Tmean of 11.13°C, 0.89°C warmer than the 1991–2020 average with only 2022 warmer (11.18°C). Six of the 10 years in the most recent decade, 2014–2023, in the CET series have been in the top-ten warmest: 2014, 2017, 2018, 2020, 2022 and 2023—the same years as the UK series. 2023 was the second warmest year in the CET Tmax series also behind 2022, but the equal-warmest year (with 2006) in the CET Tmin series with an annual Tmin of 7.30°C (both series from 1878). The CET series provides evidence that the 21st century so far has overall been warmer than any period of equivalent length in the previous three centuries, and that all seasons have also been warmer (Figure 17). When comparing the early 21st century (2001–2023) to previous centennial averages, the annual Tmean difference is +0.9°C compared with 1901–2000, +1.3°C compared with 1801–1900 and 1701–1800, and +1.8°C compared with 1659–1700—with some seasonal variations (Table 3). The most recent decade (2014–2023) has been the warmest 10-year period in the CET series, and 2023 was the first year in the series in which every calendar month was warmer than the 1961–1990 long-term average. The CET and England series are very highly correlated (R2 value of 0.99 for the period of overlap) and have a root-mean-square difference of 0.1°C which is comparable to the estimated series uncertainty as described in Appendix B2. The CET series could effectively be considered a proxy for an England series from 1659, although because these are different datasets produced in different ways, some differences are inevitable. The England series has warmed slightly more than the CET series, which means that in Figure 16 the England series anomalies are slightly lower than the CET series before the 1991–2020 period. The slightly greater warming for England compared with CET warrants further investigation since the cause of this difference cannot be confidently attributed at present. However, this difference is small compared with the overall warming trend common to both series. 2023 was also the second warmest year on record at the Oxford Radcliffe Meteorological station in a series from 1814, with 11.85°C, behind 2022 (12.14°C). The next warmest year was 2014 (11.79°C). Figure 18 plots annual Tmean for the UK from 1884 to 2023 from HadUK-Grid alongside global mean surface temperature (land surface air temperature and sea surface temperature) based on the 'best estimate' time-series from the HadCRUT5 dataset (Morice et al., 2021) and the global land surface only from the CRUTEM5 dataset (Osborn et al. 2021). All three series are plotted as anomalies relative to the baseline reference period 1961–1990. The annual variability in UK Tmean is very much larger than HadCRUT5 and CRUTEM5, as the UK covers only a small fraction (~1/2000) of the Earth's surface. Globally, 2023 was the warmest year in the HadCRUT5 series from 1850, exceeding the previous warmest year, 2016, by 0.17°C. It was also the warmest year for the global land surface only. Based on HadCRUT5, 2023 was 1.10°C warmer than 1961–1990. Global average surface temperatures are commonly reported with reference to a 'pre-industrial' baseline period of 1850–1900. 2023 was 1.46°C warmer than 1850–1900. This 'pre-industrial' baseline is not available for the UK since the monthly series start in 1884. Globally, June to December 2023 were each the warmest such month on record in the HadCRUT5 series. On top of long-term warming, this was in part due to a transition into El Niño conditions in the latter part of the year, further elevating temperatures. El Niño is part of a pattern of climate variability in the tropical Pacific that imparts warmth to the global atmosphere, temporarily adding up to 0.2°C to the temperature of an individual year. This stands in contrast to the reverse pattern of climate variability, La Niña, which suppressed global average temperatures in 2021 and 2022. El Niño/La Niña is just one example of a mode of variability affecting the global climate. Globally, warming is greater across high latitudes compared with the equator, and over land compared with the ocean (IPCC, 2021, Blunden and Boyer, 2022). The most recent decade 2014–2023 has been 1.25°C warmer than 1961–1990 for the UK, compared with 0.85°C for global mean surface temperature and 1.15°C for global land only. The UK's climate is subject to natural multi-annual to multi-decadal modes of variability which will super-impose on any longer-term trend. Taking this factor into consideration, the underlying warming observed for the UK is consistent with that observed globally over land. However, the details of any comparison will also depend on the choice of 1961–1990 as the baseline. Figure 19a,b shows daily maximum and minimum temperature anomaly maps relative to the 1991–2020 monthly averages for each day of 2023. In the UK's climate, daily maximum and minimum temperature anomalies are mostly within ±8°C of the monthly average (encompassing the full-colour scale of these charts) with anomalies generally only exceeding these values on a few days of the year—often across only a relatively small area. Figure 20a,b shows the UK area average daily maximum (Tmax) and minimum (Tmin) temperatures through the year. In common with most recent years, in 2023 the number of days where the UK average daily maximum and minimum temperatures were above the 1991–2020 monthly averages exceeded the number of days below: 2023 comprised 230/135 days above/below for maximum and 218/147 days above/ below for minimum temperature. The most recent year where the number of days below average exceeded days above was 2016 for Tmax and 2015 for Tmin. By far the most significant spells of above-average temperature in 2023 were in June and early September (these events are described in Sections 9.1 and 9.2), while much of December was also very mild. There were significant cold spells in mid-January, early March and late November to early December, although despite this no months of the year saw temperature anomalies well below the 1991–2020 average (Figure 10, Table 1). Figure 21a,b shows the UK's highest daily maximum and lowest daily minimum temperatures for each calendar day of 2023 based on the HadUK-Grid dataset. These are point values—the location of which will vary on a daily basis depending on where in the UK that daily extreme happens to be located—so they differ from Figure 20a,b which is UK mean (i.e., area average) daily maximum and minimum. The highest and lowest values, therefore, represent the absolute temperature ranges of point values across the UK through the calendar year in 2023 and from 1960 t