Weather and Climate Extremes 44 (2024) 100668 Available online 8 April 2024 2212-0947/© 2024 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC license (http://creativecommons.org/licenses/by- nc/4.0/). Dynamics of an extreme low temperature event over South Africa amid a warming climate Hector Chikoore a,*, Innocent L. Mbokodo b, Mukovhe V. Singo c, Tumelo Mohomi a, Rendani B. Munyai d, Henno Havenga e, Dawn D. Mahlobo b, Francois A. Engelbrecht f, Mary-Jane M. Bopape g, Thando Ndarana h a Department of Geography and Environmental Studies, University of Limpopo, Sovenga, South Africa b Climate Services, South African Weather Service, Centurion, South Africa c Institute for Soil Climate and Water, Agricultural Research Council, Arcadia, South Africa d Department of Economic Management Education and Social Sciences Education, University of Limpopo, Sovenga, South Africa e Unit for Environmental Sciences and Management, North-West University, Potchefstroom, South Africa f Global Change Institute, University of the Witwatersrand, Johannesburg, South Africa g South African Environmental Observation Network, National Research Foundation, Colbyn, South Africa h Department of Geography, Geo-Informatics and Meteorology, University of Pretoria, Hatfield, South Africa A R T I C L E I N F O Keywords: Extreme low temperature events Snow falls Ground frost Cut-off low Wave-breaking event Intense cold front Ridging anticyclone Jet streak A B S T R A C T Despite robust warming trends in surface air temperatures over southern Africa, extreme low temperature (ELT) events still occur from time to time. A recent ELT event affected South Africa resulting in disruptions in socio- economic activities amid a coronavirus pandemic. At least 27 long-term low temperature records were broken during 22–24 July 2021, with snow falls observed mostly over high ground in subtropical districts. This study employs weather station data and European Centre for Medium-Range Weather Forecasts (ECMWF)’s ERA5 and ERA5-Land reanalyses to investigate dynamics of the ELT event focusing on the South African Highveld. Our approach employs multiscale analysis, with long term trends and climatologies of surface air temperatures, snow events and ground frost days presented as background to understanding the observed extreme weather anom- alies. We found consistent and statistically significant warming trends in daytime and overnight temperatures, with corresponding decreases in ground frosts. The July 2021 ELT event resulted from a combination of complex circulation anomalies which included an intense offshore cut-off low (COL) that extended to the surface (and associated wave breaking), a cold front and a Type-S ridging anticyclone, all intensifying surface cold air advection from the Southern Ocean. A most significant finding is that COLs do not need to enter South Africa to cause severe weather over the country. Our study contributes to understanding the occurrence and dynamics of cold extremes in subtropical regions, against a robust warming trend. 1. Introduction In Africa, a dry-wet seasonality is more pronounced over large areas of the continent compared to a cold-hot seasonality as rainfall is the most important meteorological parameter (Tyson 1986; Nicholson 2000). Winter seasons in Africa’s tropical region are considered mild, rarely falling below 15 ◦C (Leroux 2001; Ongoma and Chen 2017) compared to boreal winters. Farther south - South Africa lies predominantly in the subtropics, extending from about 22◦S to 34◦S, and the traditional summer, autumn, winter and spring seasons are more applicable (van der Walt and Fitchett 2020a; Mbokodo et al., 2020; Kruger and Nxumalo 2017a). The winter spans the period from mid-May to mid-August, sometimes extending into early September in the south (van der Walt and Fitchett 2020a). Dominant weather systems affecting southern Africa during the austral winter months are the subtropical anticyclones (highs) which migrate equatorward coupled with semi-permanent upper air high pressure systems (Tyson and Preston-Whyte 2000; Ndarana et al., 2021), causing dry and stable weather over much of the region from May to August. The predominantly anticyclonic circulation produces * Corresponding author. E-mail address: hector.chikoore@gmail.com (H. Chikoore). Contents lists available at ScienceDirect Weather and Climate Extremes journal homepage: www.elsevier.com/locate/wace https://doi.org/10.1016/j.wace.2024.100668 Received 5 November 2023; Received in revised form 5 March 2024; Accepted 3 April 2024 mailto:hector.chikoore@gmail.com www.sciencedirect.com/science/journal/22120947 https://www.elsevier.com/locate/wace https://doi.org/10.1016/j.wace.2024.100668 https://doi.org/10.1016/j.wace.2024.100668 https://doi.org/10.1016/j.wace.2024.100668 http://crossmark.crossref.org/dialog/?doi=10.1016/j.wace.2024.100668&domain=pdf http://creativecommons.org/licenses/by-nc/4.0/ http://creativecommons.org/licenses/by-nc/4.0/ Weather and Climate Extremes 44 (2024) 100668 2 subsidence inversions which trap pollutants causing haze over industrial regions during this time (Jury and Freiman, 2002; Wallace and Hobbs, 2006; Abiodun et al., 2014; Jury 2017). Early morning fogs (van Schalkwyk 2011; van Schalkwyk and Dyson 2013) due to radiation losses, advection or mountain-valley circulations are common during winter whilst ground frosts (Moeletsi et al., 2016) are also observed over susceptible and exposed surfaces. Whilst most of the interior plateau over southern Africa is experi- encing the dry season in winter, cold fronts, cut-off lows (COLs) and ridging anticyclones produce thunderstorms and rainfall over the southwestern Cape and southern coastal areas of South Africa during this time (Engelbrecht et al., 2015a; Barnes et al., 2021a; Omar and Abiodun 2020). Thus, the winter months in South Africa are important to the southwestern Cape region as they coincide with the main rainy season there. The frequency and establishment of rainfall-producing systems and thus how wet (dry) or cold (cool) the winter season may be linked to remote phenomena such as the Southern Annular Mode (SAM; Limpa- suvan and Hartmann, 2000). The El Niño Southern Oscillation (ENSO) is more important for the summer rainfall regions with minimal impact on winter rainfall regions (Mahlalela et al., 2019). Whereas the SAM is a dominant mode of variability in the higher latitudes of the Southern Hemisphere (SH) and is known to influence rainfall and temperature variability in parts of the hemisphere (Gillett et al. 2006; Abram et al., 2014; Thompson et al., 2011; Fogt and Marshall 2020). The SAM, at times referred to as the Antarctic Oscillation, is defined as the zonal mean sea level pressure difference between the 40◦S and 65◦S (Lee et al., 2019). In South African context, the negative phase of SAM has been associated with anomalously wet conditions over the southwestern Cape (Reason and Rouault 2005; Mahlalela et al., 2019) consistent with Mediterranean-type climate regions in south Australia and southern South America due to the northward shift of storm tracks (Gillett et al. 2006). Conversely, the poleward migration of mid-latitude cyclones and associated cold fronts lead to a decline in rainfall over these regions (Sousa et al., 2018; Picas and Grab, 2020) and these systems are pro- jected to further retreat poleward in future due to climate change (Catto et al., 2019). The passage of a westerly wave bringing winter rainfall is usually followed by intrusion of cold air from the south advected by ridging anticyclones (Ndarana et al., 2021). Thus, extreme low temperature (ELT) events are typically associated with drizzly/rainy days and in some cases snow over high ground or mountain peaks. Due to South Africa’s location, the occurrence of snow over the country is rare, except over mountain tops in the south (Stander et al., 2016). When it does occur, snow causes disruptions to economic activities, property and may lead to loss of lives of the homeless (Stander et al., 2016) and liveli- hoods. As the country experiences limited snowfall, snowmelt in its headwater catchments may contribute to increased soil moisture, groundwater, and river flows (Wu and Xu 2005; Stander et al., 2016). Recently, an anomalous COL centered over the far northeast of the country resulted in freezing temperatures and rare but widespread snow extending to the commercial capital of Gauteng on the Highveld in July 2023 (The New York Times, 2023).1 Depending on the timing of ELT, a variety of negative losses can occur for different agricultural commodities. Impacts of ELT events in South Africa include severe frosts on winter crops (e.g. wheat, barley, oats), loss of young and vulnerable livestock and economic losses (Archer et al. 2021; Farmer’s Weekly, 2023). An ELT event that affected South Africa in 2010 resulted in a 20% loss in vegetable production in the northern Limpopo Province, and losses of 124 000 tons of sugarcane in KwaZulu-Natal on the east coast (Farmer’s Weekly, 2010). Often, severe ELT events result in elevated energy demand for agriculture, domestic heating and industrial production with implications on in- terruptions of electricity supply (Mail and Guardian, 2023), as was the case during 22–24 July 2021 (Eskom, 2021). It was also reported that water reticulation pipes were frozen interrupting water supply on July 22, 2021 in the central Free State Province (Timeslive, 2021). The main weather systems associated with ELT conditions over South Africa are ridging anticyclones from the south Atlantic Ocean (Ndarana et al., 2021) coupled with mid-tropospheric COL pressure systems aloft. Cold fronts, which usually precede ridging anticyclones are responsible for most of the cold air advection from the higher colder latitudes as the South Atlantic High ridges over the south of Africa (Ndarana et al., 2021). The orientation of ridging anticyclones de- termines the length of fetch of cold air. Type-N (centered north of 40◦S) ridging anticyclones result in deeper intrusions of cold air onto the southern Africa subcontinent compared to their Type-S (centered south of 40◦S) counterparts whose influence mostly affects South African lat- itudes (Ndarana et al., 2022). In addition to cold air advection, they contribute to nearly half (~46%) of rainfall over the all-year rainfall region of the Cape South Coast (Engelbrecht et al., 2015a). The inter- action of cold air advection with the sharp escarpment near the coasts results in orographic lifting and rain or drizzle. The frequency and duration of ridging events was also found to be related to drought in the southwest Cape during the 2015 winter (Burls et al., 2019). COLs are closed lows that occur in the upper air over subtropical latitudes and are associated with ELT conditions, heavy rainfall and snowfalls (Singleton and Reason 2007; Favre et al., 2013; Muofhe et al., 2020; Xulu et al., 2023). They occur when a mid-tropospheric low be- comes detached from the westerlies (Palmen and Newton, 1969) and are often coupled with a ridging anticyclone at the surface (Muofhe et al., 2020, Ndarana et al., 2023) and may, on occasion, extend to the surface (Barnes et al., 2021b). Mid-tropospheric low-pressure systems imply low mean air temperature values of the layer from the surface to the upper air. The ridging anticyclone at the surface advects low level moisture and cold air from the Southern Ocean. COLs are also responsible for the most number of flood events, human fatalities and damage to infra- structure in the southern provinces of South Africa (Bopape et al., 2024). More recently, floods from COLs that affected the eastern KwaZulu-Natal region resulted in more than 440 deaths in April 2022, destruction of 4 000 homes, and displacement of 40 000 people (SAWS 2023). Several studies have found consistent and statistically significant warming trends of surface minimum and maximum air temperatures in most areas of South Africa (Karl et al., 1993; Engelbrecht et al., 2015b; Kruger and Nxumalo 2017b; Jury 2018; van der Walt and Fitchett 2020b; Mateyisi et al., 2021; McBride et al., 2021; IPCC et al., 2021). Tshiala et al. (2011) investigated trends in surface air temperatures over Limpopo Province over a 50-year period (1950–1999) and found sta- tistically significant positive trends (0.12 ◦C per decade at α = 0.01 significance level) for most stations, with winter warming faster than the summertime. Despite the observed and robust warming trends in South Africa, extreme temperature events continue to occur. However, general trend of decreasing frequency of ELTs events, cold waves and their duration were found in most stations across South Africa including on the Highveld (van der Walt and Fitchett 2020b). Summer heat wave events are frequent and often occur as compound events with meteo- rological droughts (Mbokodo et al. 2020, 2023), linked to sea-surface temperature variability of the tropical Indian and Pacific oceans (Klopper et al., 1998; Mbokodo et al., 2023). Whilst several environ- mental systems and human communities are suited to the mean climate of an area and moderate variability (Salinger and Griffiths 2001), extreme deviations from the mean may have significant impacts. Several episodes of anomalous cold weather affected the country during the 2021 austral winter season. An ELT weather event from 22–24 July 2021 broke 27 long-term records of lowest temperatures on the South African Weather Service (SAWS) database (SAWS 2021). The event resulted in disruptions of socio-economic activities and threatened 1 https://www.nytimes.com/2023/07/12/world/africa/south-africa-snow- johannesburg.html Accessed on 20 July 2023. H. Chikoore et al. https://www.nytimes.com/2023/07/12/world/africa/south-africa-snow-johannesburg.html https://www.nytimes.com/2023/07/12/world/africa/south-africa-snow-johannesburg.html Weather and Climate Extremes 44 (2024) 100668 3 human lives amid a raging coronavirus pandemic. Our study is moti- vated by these impacts and the record-breaking low temperatures against a significant warming trend. Unlike heatwave events, little work has been done focusing on ELT events in South Africa, and what causes them remains unknown. Thus, the aim of this study is to analyse the nature and causes of anomalous ELT events during the 2021 austral winter season over South Africa, with a large part of the work focusing on the Highveld region. We use a larger domain that includes the surrounding oceans to understand the driving dynamics. The rest of the paper is organized as follows: the methodology is presented next; results are presented and discussed starting with climatologies of variables related to cold weather, then a focus on the 22–24 July 2021 extreme event. A conclusion and recom- mendations are offered at the end. 2. Data and methods 2.1. The South African Highveld South Africa is characterised by complex topographic gradients with narrow coasts, sharp escarpments, and high peaks in the east, falling to the Limpopo River in the northeast (Fig. 1a). Some of the long term mean spatial patterns of minimum (Myburg, 1974) and maximum temperatures and the occurrence of snowfalls may be best explained using topographic arguments/parameters (Maisha, 2014). Local circu- lations due to topographic effects also affect cloudiness, fogs, and rain- fall patterns (Singleton and Reason 2006; Omar and Abiodun 2021). In addition, the distance from the sea, surface ocean currents, latitude, aspect, cloudiness, and prevailing winds also contribute to observed spatial, seasonal, and diurnal variability of surface air temperatures over the country. Much of the analyses focus on the South African Highveld which is defined as that region that lies between 1 400 and 1800 m above mean sea level and found on the eastern part of the country but north of Lesotho2 (Fig. 1a). The Highveld covers large areas of the North West Province, Free State Province and Mpumalanga Province. Higher ele- vations lie southeast of the Highveld, toward the mountain kingdom of Lesotho, with elevations mostly above 2000 m (Fig. 1a). The South Af- rican Highveld includes the major population (~15 m people, STATSSA, 2023) and industrial centers of the Gauteng Province, the economic capital of South Africa. Even though the Highveld region is not a large proportion of South Africa’s land area, extreme weather and climate events cause significant impacts there on transportation, economic ac- tivities, human lives, and livelihoods. This motivates our interest in this region. Besides, the winter season on the Highveld lasts longer compared to the lower altitude regions where the warmer temperatures return early (van der Walt and Fitchett 2020a). Despite the focus on the Highveld, we consider temperature patterns over the rest of the country, and circulation anomalies extending seaward to understand the extreme weather event. 2.2. Analysis of observed and reanalysis data The spatial analysis and interpolation of surface air temperatures in this study is largely based on quality-controlled data from the extensive station network of the SAWS (Fig. 1b). The spatial distribution of sta- tions used in this study is representative of the thermal regions (Kruger and Sekele 2013; Mbokodo et al., 2023) across the country. The tem- peratures were interpolated on spatial maps using the Inverse Distance Weighted (IDW) technique in an ArcGIS environment. The IDW assumes decreasing influence of a mapped variable with increasing distance from a point of sampling (Watson and Philip, 1985). The IDW method of interpolation is straightforward to understand and implement, which renders it less prone to errors in the interpretation of results. It also performs well with high spatial points density, as in this study (Harman et al., 2016, Fig. 1b). We sought to find evidence of climate change over the Highveld based on a long (1901–2020) time series of maximum temperature (Tmax), minimum temperature (Tmin), and ground frost days from the CRU TS4.04 (Harris et al., 2020). The Highveld is defined by a polygon bounded by latitudes 26◦ to 29◦S and longitudes 26◦ to 30.5◦E. In winter, the coldest overnight temperatures (− 2 to 13 ◦C) in South Africa occur typically between 03:00 UTC and 04:00 UTC whilst day- time maxima (15–27 ◦C) are often observed between 12:00 UTC and 13:00 UTC (Mbokodo 2017). These times would vary during cloudy and rainy days for example. Ground frost occurs when the ground/soil temperature reaches and falls below 0 ◦C and is often concerning to farmers whose focus is on winter crops. Frosts were traditionally measured from minimum thermometers placed on the ground and measured at 06 UTC in southern Africa. The CRU TS4.04 uses the gridded absolute Tmin variable to calculate ground frost days via an empirically determined function (Harris et al., 2020). Commercial farmers often use greenhouses and overhead sprinkler irrigation to prevent crop failure due to ground frosts. As irrigation water freezes to ice in cold weather, the latent heat of fusion released replaces the heat lost by crops to the environment (Perry 1998; Yongguang et al., 2016). We analysed key aspects of a time series including the seasonality, and long-term trends in anomalies of Tmin and Tmax and frequency of ground frost days. Anomalies were calculated based on 1991–2020 long term climatologies whilst trends were tested for statistical significance using the Mann-Kendall test at the 95% confidence level. Long-term trends provide context of the 2021 winter event, to demonstrate that it was one of the most extreme events, hence our interest in it. The winter season is defined in this study as spanning the months June, July and August (JJA), as in van der Walt and Fitchett (2020a, b). For completeness, the long-term (1991–2020) mean spatial patterns for the austral spring (September, October and November; SON), summer (December, January and February; DJF) and autumn (March, April and May; MAM) seasons are also analysed. Trends in cold nights (TN10P), cool days (TX10P), warm nights (TN90P), and hot days (TX90P) from 1991 to 2020 were calculated using the RClimDex version 1.0 software (Zhang and Yang, 2004). The RClimDex is an R-based library that conducts quality control for climate data and also computes indices of climate extremes as defined by the Expert Team on Climate Change Detection and Indices (ETCCDI). This tool was developed and is maintained by the Climate Research Branch of the Meteorological Service of Canada and is freely available on the ETCCDI website.3 We calculated the RClimDex trends for Johannes- burg’s Oliver Reginald Tambo International Airport station which is the only station on the Highveld whose daily records span at least 30 years (1991–2020) from the SAWS database. This RClimDex has been used in several studies (e.g., Kruger et al., 2019; Mbokodo et al., 2020; Mbokodo et al., 2023) to analyse historical trends in climate extremes over South Africa. We analysed the occurrence (frequency and location) of snow events obtained from the SAWS Caelum publication which documents infor- mation on weather extremes and their impacts in South Africa dating back to the 1500s (SAWS 2023). The Caelum publication is maintained by the SAWS and provides a continuously updated history of notable weather-related disasters in South Africa (Viljoen, 1991; Caelum, 2022; Mashao et al., 2023). This publication provides information such as dates of extreme weather events, resulting socio-economic impacts and also the regions affected. Two snow datasets were used in this study. While the snow data from Caelum provided a 1981–2020 climatology 2 https://www.oneearth.org/ecoregions/highveld-grasslands (Accessed on 22 June 2023). 3 http://etccdi.pacificclimate.org/software.shtml (Accessed on 22 June 2023). H. Chikoore et al. https://www.oneearth.org/ecoregions/highveld-grasslands/ http://etccdi.pacificclimate.org/software.shtml Weather and Climate Extremes 44 (2024) 100668 4 Fig. 1(a). The Highveld of South-Africa (demarcated in blue) lies at a mean altitude of 1 500 m. The area covers most of central South Africa and is also home to various industries. Johannesburg and Pretoria are two major metropolitans in the region, and the economic hub of Africa. The locations of Automatic Weather Stations (AWS) that broke temperature records are shown as asterisks. The locations of Oliver R Tambo Airport and Cape Town Airport are pinned in Gauteng. Fig. 1 (b) shows the location of weather stations (WS) (represented by black points) used for the interpolation of temperatures across South Africa by using the inverse distance weighted (IDW) method. H. Chikoore et al. Weather and Climate Extremes 44 (2024) 100668 5 over the country, the ERA5-Land (Munoz-Sebater et al., 2021) snow dataset was used for event scale analyses during 22–25 July 2021. The ERA5-Land snow data was detailed in Munoz-Sebater et al., 2021 and evaluated against two sets of observations including station data and from the Global Historical Climatology Network-Daily (GHCN-daily; Menne et al., 2012). Radiosonde soundings made at Cape Town Inter- national Airport and Bloemfontein were obtained from the University of Wyoming are used to investigate the vertical profiles of environmental air during 22 July at 00Z and 12Z respectively. The vertical profiles provide information on freezing levels, upper winds and vertical wind shear. We employed ECMWF ERA5 reanalyses at 0.25◦ resolution (Hers- bach et al., 2020) to analyse anomalies in the atmospheric circulation focusing on geopotential heights (m), wind vectors and isotachs (m/s) in the lower (850 hPa) and middle (500 hPa) troposphere at the event timescale to understand dynamical processes. We isolated and shaded streamlines greater than 30 m/s and identified the jet streak (region of very strong winds above 30 m/s in the upper air) on the plots. The location and interaction of jet streaks with surface weather systems may result in low-level advection of colder/warmer air into a region (Chang and Lau 1980; Singleton and Reason 2006). In addition, the 1 000-500 hPa thickness (m) is analysed as an important indicator for the likeli- hood of frozen precipitation and hence snowfalls. Thickness is a measure of the distance between two pressure levels and depends on the mean temperature and moisture. The 540 dam (decameters) contour is a critical isopleth of constant thickness below which snow falls are most likely (Grumm, 1998). 3. Results and discussion 3.1. Climatologies 3.1.1. Seasonal mean temperatures We present spatial climatologies (Figs. 2 and 3) of seasonal mean surface air temperatures (Tmin and Tmax) for all seasons from DJF, MAM, JJA, to SON from the SAWS station data. The winter season over South Africa is, as expected, the coldest season across the country (Fig. 2c), but low temperatures are also experienced on occasion during autumn and spring especially over high ground (Fig. 2b–d). Tmin averaged below 6 ◦C during JJA over the interior plateau, with slightly warmer temperatures near the east coast and the northeast Lowveld. The coldest overnight temperatures occur over the region south of the Highveld, extending into the mountains of the Eastern Cape Province (Fig. 2c) and Lesotho (inferred, not shown). During JJA, minimum temperatures drop below freezing/0 ◦C, particularly in the south of the Highveld, which includes parts of the Free State, the Eastern Cape, and Lesotho’s Maluti Mountains. Tmin is generally mild during DJF with mean values between 12 and 15 ◦C over the eastern Highveld and higher in the west (Fig. 2a). The rest of the country is much warmer overnight during the summertime, with mean temperatures above 18 ◦C over the far northeast Lowveld and the arid Kalahari in the Northern Cape Province. During the transition seasons (MAM and SON), Tmin are generally cool, although the spring appears about 3 ◦C warmer partic- ularly in the south of the Highveld (Fig. 2b–d). The SON spring season (Fig. 2d) also coincides with the peak of the fire season (Singo et al., Fig. 2. Seasonal mean spatial patterns of minimum temperatures during (a) DJF, (b) MAM, (c) JJA, and (d) SON over South Africa based on a 1981–2020 climatology from the SAWS station observations The purple polygon indicates the Highveld, while Lesotho and Eswatini have been masked out. H. Chikoore et al. Weather and Climate Extremes 44 (2024) 100668 6 2023). Mean daytime temperatures (Tmax) over the Highveld are consid- ered cool during JJA, mostly ranging between 18 and 21 ◦C (Fig. 3c). The lowest Tmax temperatures during winter occur in a region extending from southeast of the Highveld, south through Lesotho and zonally along about 32◦S (Mountain range). Across the country, most temperatures during this time are generally below 21 ◦C, with the Lowveld of Limpopo and narrow coastal areas of KwaZulu-Natal being slightly warmer. The spring season (SON) is characterized by rising temperatures particularly over the northern districts, whilst the south remains relatively cool (Fig. 3d). This is a significant finding, supporting the argument that the winter season lasts longer in the south, including eastern parts of the Highveld. The highest seasonal mean Tmax are observed during DJF with maxima (exceeding 33 ◦C) in the Northern Cape bordering Botswana and Namibia (Kalahari) and (Fig. 3a). The very high temperatures observed over the Northern Cape coincide with a surface heat low that becomes semi-permanent during this time (Howard and Washington 2018). Following the summer, the transition MAM season is often considered the most pleasant time of the year (Torrance 1981), when the rainy season has ended and as temperatures are mild (Fig. 3b). Overall, the mean Tmin and Tmax across all seasons are warmer in the west of the South African Highveld (Figs. 2 and 3). As shown in Fig. 1, the Highveld has a stepper topography in the east, hence the cooler temperatures. Both altitude and latitude play a role as we move farther south. As expected, the narrow coasts around the country are generally warmer compared to the interior plateau. 3.1.2. Seasonal cycles and interannual variability of Tmin, Tmax and ground frosts The annual cycle of Tmin over the Highveld shows the coldest con- ditions occurring during June and July (Fig. 4a). July is slightly colder with a mean Tmin of 2.42 ◦C while June has a mean of 2.76 ◦C. This coldest period also corresponds with the highest frequency (13–14 days) of ground frost days (Fig. 4b). Tmin peaks during DJF at mean values of 15–16 ◦C. Whilst intrusions of cold air onto the interior plateau may result in cold days even during summertime, ground frost occurs only during April to September (Fig. 4b). Frost free days (maize growing season) occur from September/October to April/May/June southwest of the Highveld (Free State Province, Moeletsi et al., 2016). The annual cycle of Tmax closely resembles that of Tmin, though with June recording the lowest temperatures (18.4 ◦C). The highest average Tmax (28.48 ◦C) occurs during January over the Highveld, consistent with large areas of the country (Mbokodo et al., 2023). Anomalies or Tmin over the period 1901 to 2021 show random fluctuations but with a general upward trend suggesting warming overnight temperatures over the Highveld (Fig. 4c). The Mann-Kendall trend and Sen’s slope test shows a statistically significant (cf. appen- dix a) and increasing trend in Tmin (Fig. 4c) and Tmax (Fig. 4d) over the 121-year period. Anomalies in the frequency of ground frost days show gradual and statistically significant declines over the area (Fig. 4e), consistent with a warming trend in mean surface air temperatures. Similarly, Moeletsi et al. (2016) found increasing trends of frost-free periods over South Africa’s Free State Province focusing on the period 1960–2015. Fig. 3. Seasonal mean spatial patterns of maximum temperatures (◦C) during (a) DJF, (b) MAM, (c) JJA, and (d) SON over South Africa based on a 1981–2020 climatology from the SAWS station observations The purple polygon indicates the Highveld, while Lesotho and Eswatini have been masked out. H. Chikoore et al. Weather and Climate Extremes 44 (2024) 100668 7 3.1.3. Trends in temperature extremes over Johannesburg airport Johannesburg’s OR Tambo International Airport is located on the Highveld at about 1700 m above mean sea level (Fig. 1a). We ran RClimDex for Tmin and Tmax to calculate trends at the airport and found declining trends in cold nights and cool days over the period 1991–2020 (Fig. 5a–c). Warm nights and hot days exhibit increasing trends in Johannesburg (Fig. 5b–d), consistent with country (McBride et al., 2021) and regional trends (Engelbrecht et al., 2015b; Singo et al. 2023). All trends are statistically significant at the 95% confidence level (cf. appendix b). While the trend for cold nights and cool days is clearly declining, very cold days still occur regularly, occasionally breaking long-term records (SAWS 2021; McBride et al., 2021). Fig. 4. Mean annual cycles of CRU TS4 (a) minimum and maximum temperature (Tmin) (b) ground frost days, and (c) long-term (1901–2021) interannual anomalies of Tmin, and (d) ground frost days. The anomalies are calculated based on a 1991–2020 climatology. H. Chikoore et al. Weather and Climate Extremes 44 (2024) 100668 8 Fig. 5. RClimDex trends from 1991 to 2020 (dotted lines) in (a) COLD NIGHTS and (c) COOL DAYS and (b) WARM NIGHTS and (d) HOT DAYS over the Johan- nesburg Intl Airport station on the South African Highveld, one of the stations that broke records during July 2021. Fig. 6. (a) Total snow events and seasonal mean spatial patterns during (b) MAM, (c) JJA, and (d) SON over South Africa based on a 1981–2020 climatology from the Caelum database. The Highveld area is shown as the shaded polygon while Lesotho and Eswatini have been masked out. H. Chikoore et al. Weather and Climate Extremes 44 (2024) 100668 9 3.1.4. Snow occurrence over South Africa A total of 476 snow events were recorded in South Africa on the SAWS Caelum between 1981 and 2020 (Fig. 6a). From this total number, no event was reported during the mid-summer (DJF) season (not shown), with only 38 events (8%) during the autumn (MAM) season (Fig. 6b). As expected, most (92%) snow events occurred during the winter (JJA; Fig. 6c) and spring (SON; Fig. 6d) seasons with a winter peak of 332 events (70%). Spatially, most snow events have occurred on high-lying areas, extending from the Western Cape, Eastern Cape, through Lesotho mountains (inferred), the eastern Highveld and along the north-eastern escarpment of the country (Fig. 6a). The tropical far north of the country (<25◦S), which includes the entire Limpopo Province, did not experience snow falls during the study period. The propagation of COLs from the west over the arid/semi-arid Northern Cape may account for the snow events observed over the region. 3.2. Case study: ELT weather during 22–25 July 2021 3.2.1. Temperature observations and anomalies (a) Long-term records broken Twenty-seven (27) extreme low temperature records were broken at 26 stations during July 2021 over South Africa, with more records during periods before 22–24 July. Most broken records had been set more than twenty (20) years prior, with 61-year records broken at Johannesburg International Airport, Kimberley, and Warmbad Tow- oomba (Table 1). Several other lowest Tmin or lowest Tmax were broken during 12, 14, 15, 16 and July 20, 2021. The geographical distribution of broken records was mainly over the eastern high ground (>1 000 m) over the interior plateau, and along the coast of the Eastern Cape, Table 1 Twenty-seven (27) temperature records that were broken at weather stations in South Africa during 22–24 July 2021. The asterisk (*) indicates stations over the Highveld. (Data Source: SAWS). Station Co-ordinates Elevation (M) Nature of records Old records New records Date Length of records Warmbad Towoomba − 24.8992, 1 143 Lowest Tmin − 5.5 ◦C − 5.6 ◦C 2021-07-23 61 years 28.3239 Johannesburg Int WO* − 26.143, 1 695 Lowest Tmin − 6.3 ◦C − 7.0 ◦C 2021-07-23 61 years 28.2346 Kimberley WO − 28.8061, 1 196 Lowest Tmin − 8.1 ◦C − 9.9 ◦C 2021-07-23 61 years 24.7698 Estcourt − 29.0091, 1 159 Lowest Tmin − 4.6 ◦C − 4.8 2021-07-22 48 years 29.8627 Kroonstad* − 27.6665, 1 440 Lowest Tmin − 7.7 ◦C − 8.0 ◦C 2021-07-23 43 years 27.3136 Royal National Park − 28.6858, 1 392 Lowest Tmin − 7.6 ◦C − 7.9 ◦C 2021-07-24 42 years 28.9542 Lydenburg − 25.1119, 30.4766 1 439 Lowest Tmin − 3.2 ◦C − 4.3 ◦C 2021-07-23 34 years Cape town - Royal Yacht Club − 33.9206, 11 Lowest Tmin 4.3 ◦C 2.5 ◦C 2021-07-24 32 years 18.443 Warden – Heritage* − 27.9611, 1765 Lowest Tmin − 6.3 ◦C − 6.7 ◦C 2021-07-23 32 years 29.0594 Patensie − 33.7653, 95 Lowest Tmax 10.2 ◦C 9.0 ◦C 2021-07-22 32 years 24.8233 Grahamstown − 33.2907, 642 Lowest Tmax 7.2 ◦C 6.8 ◦C 2021-07-22 29 years 26.5026 Mtunzini − 28.9474, 41 Lowest Tmin 1.8 ◦C 1 ◦C 2021-07-23 28 years 31.7079 Ermelo WO* − 26.4977, 1774 Lowest Tmin − 4.7 ◦C − 5.1 ◦C 2021-07-23 28 years 29.9838 Uitenhage − 33.7141, 156 Lowest Tmax 10.6 ◦C 10.2 ◦C 2021-07-22 28 years 25.4352 Postmasburg − 28.3454, 1 321 Lowest Tmin − 7.2 ◦C − 7.5 ◦C 2021-07-22 28 years 23.079 Orania AWS − 29.8185, 1 122 Lowest Tmin − 7 ◦C − 7.4 ◦C 2021-07-23 27 years 24.4112 Wepener* − 29.9157, 1 414 Lowest Tmin − 9.8 ◦C − 10.2 ◦C 2021-07-24 22 years 26.847 Cape Town - Royal Yacht Club − 33.9206, 11 Lowest Tmax 12.4 ◦C 11.6 ◦C 2021-07-23 20 years 18.443 Port Alfred - Airport − 33.5595, 84 Lowest Tmax 11.7 ◦C 8.1 ◦C 2021-07-22 20 years 26.8809 Ngqura - Coega − 33.8051, 47 Lowest Tmax 11.1 ◦C 10 ◦C 2021-07-22 18 years 25.6695 Kokstad − 30.5028, 1 451 Lowest Tmin − 6.1 ◦C − 8.6 ◦C 2021-07-24 15 years 29.3942 Tzaneen-Westfalia Estate − 23.7367, 840 Lowest Tmin 5.4 ◦C 4.2 ◦C 2021-07-24 14 years 30.1127 Mossel Bay − 34.1894, 135 Lowest Tmin 5.7 ◦C 5.2 ◦C 7/22/2021 13 years 22.1318 Bronkhorstspruit AWS* − 25.8087, 1 394 Lowest Tmin − 7.1 ◦C − 7.4 ◦C 7/23/2021 12 years 28.7386 Port Elizabeth AWS − 33.9827, 60 Lowest Tmax 12 ◦C 10.9 ◦C 7/22/2021 12 years 25.6138 King Shaka AWOS − 29.6019, 91 Lowest Tmin 2.1 ◦C 0.9 ◦C 7/23/2021 11 years 31.1295 Venetia Mine − 22.4517, 697 Lowest Tmin 0.9 ◦C − 0.1 ◦C 7/23/2021 11 years 29.3315 H. Chikoore et al. Weather and Climate Extremes 44 (2024) 100668 10 KwaZulu-Natal and Western Cape (Fig. 1a). The majority of the stations that broke records were located east of 24◦E, apart from Cape Town - Royal Yacht Club in the Western Cape (Fig. 1a). The longest records were broken on July 23 by Kimberly WO (61 years) in the North West and the two highveld stations, Johannesburg Int (61 years) and Kroon- stand (43 years), and on July 24 at Royal National Park (42 years) in the interior (SAWS 2021). Most of the lowest Tmax during the period occurred along the coast of the Eastern Cape Province. The extreme temperature drops experienced by interior stations are therefore re- flected in the overnight minimum temperature as a result. (Table 1). Stations on the interior plateau dominated the highest records broken on July 22nd, the coastal Eastern Cape stations dominated in extreme cold conditions (Fig. 1a; Table 1). Three of the top five stations, based on their coolest/lowest temperature record on 22–23 July, are located in the Highveld region. (b) Diurnal Temperature Range The mean monthly diurnal temperature range (DTR) is large (15–21 ◦C) during the austral winter in most interior parts of South Africa (Mbokodo 2017), as observed during July 2021. The DTR time series of record-breaking stations above high ground (>1 000 m; Fig. 7a) which are essentially stations within the Highveld and its proximity, are compared with DTR for record-breaking stations near sea level (<200 m; Fig. 7b). Over- night temperatures dipped below freezing levels over the high ground stations whilst the coastal stations remained above 0 ◦C, despite bitter cold conditions (Fig. 7a and b). Overall, stations over high ground exhibited a strong DTR between 18 and 20 ◦C (Fig. 7a) while lower elevation stations near the coast showed a very weak DTR (5–7 ◦C; Fig. 7b) during 22–23 July. This period corresponds with the days when most records were broken (Table 1). It seems elevation accounts for the differences in DTR between the coastal stations and the interior plateau. It must be noted that daily changes of local factors such as soil moisture, ground heat capacity and albedo also influence variations in DTR trends (Tshiala et al., 2011). This may account for the differences between the daily and monthly temperature anomalies. (c) Anomalies Monthly Tmin anomalies for July 2021 were below normal over much of the country including the Highveld, apart from a few stations in the far north and far west (Fig. 8a). Over the Highveld, the mean monthly Tmax was near normal (Fig. 8b). However, when considering temperature anomalies during the 22–25 July, the whole country was colder than normal with Tmin values up to − 7 ◦C below (Fig. 8c). Regarding Tmax anomalies, colder than normal conditions were observed mainly over the east of the country (Fig. 8d). Consistent with Fig. 1a, it appears this ELT event was most intense over the east of the country. (d) Snowfall patterns Frozen precipitation often occurs when the freezing level is at or near ground level. We found low freezing levels for radiosonde profiles for July 22, 2021 falling to below 880 hPa (00 UTC) at Cape Town (Fig. 9a) and 820 hPa (12 UTC) at Bloemfontein (Fig. 9b). As a result, up to 2.5 cm of snow falls were observed mainly over the southern mountains of the country in the Eastern Cape extending to the Western Cape (Fig. 10). Some snowfalls were also observed atop the Tsitsikamma Mountains near the Cape South Coast. However, the entire Highveld area did not experience any snow falls during 22–25 July 2021 despite some snow deposits over the Maluti Mountains in the south of Lesotho (Fig. 10). The distribution of snow cover over the Eastern Cape prevented a more positive flux of sensible heat from the ground into the lower atmosphere compared to the interior, resulting in a greater drop in maximum temperature as the surface and sur- rounding air remained colder over the Eastern Cape. The 1 000- 500 hPa thickness values were typically low across the country due to the extreme low mean temperatures of the lower levels of the atmosphere. Much of the country lies poleward of the critical 540 dam (5 400 m) isopleth, below which snow falls are most likely. Interestingly, some snow fell on the 25th over the Western Cape mountains at thickness values of up to 5 480 m (Fig. 10d). 3.2.2. Circulation dynamics during 22–25 July 2021 We analyse the synoptic dynamical environment during the ELT event that affected southern Africa during 21–24 July 2021. An upper Fig. 7. Hourly dry-bulb temperature (C) data (a) for stations on high ground (above 1 000 m) with those stations outside the South African Highveld in dotted lines, and (b) stations near sea level (<200 m) during the period 20–26 July 2021. The extreme low temperature period 22–23 July is highlighted by gray shading. H. Chikoore et al. Weather and Climate Extremes 44 (2024) 100668 11 Fig. 8. Climatological (1981–2020) anomalies of (a) minimum temperature (Tmin; ◦C) for July 2021; (b) maximum temperature (Tmax; ◦C) for July 2021; (c) minimum temperature for the period 22–25 July 2021; and (d) maximum temperature for the period 22–25 July 2021 over South Africa from the SAWS station observations The purple polygon indicates the Highveld, while Lesotho and Eswatini have been masked out. Fig. 9. Radiosonde sounding profiles for (a) Cape Town Intl Airport and (b) Bloemfontein launched at 00Z and 12Z respectively during July 22, 2021. H. Chikoore et al. Weather and Climate Extremes 44 (2024) 100668 12 trough was observed over the subcontinent propagating slowly east- wards (Fig. 11a). The trough developed into a COL over the ocean during 22–23 July and extended to the surface. Barnes et al. (2021b, see their composite Fig. 10), showed that this category of COLs is climatologically linked to negative surface temperature anomalies that are oriented such that they are located along the trough axis. In the case of the COL observed for this ELT event, the location of low temperatures is over South Africa. Because the system extends to the surface, it may then be categorised as intense (Pinheiro et al., 2021). Behind the upper trough, a high pressure system over the South Atlantic Ocean, combined with the action of the cyclonic flow associated with the surface component of the COL, was pushing and driving divergent southwesterlies into the country (Fig. 11a,b,c,d). The observed COL on 22 July (Fig. 11 b) has an Omega shape with two jet streaks on either side of the trough axis. The upstream jet streak extends to the surface, as the COL resulting in strong southerly flow west of the surface low with intense cold air advection from the south into the country (Fig. 11e and f). This then led to very cold temperatures over South Africa (Fig. 8) thus leading to persistent ELT conditions throughout the country (Fig. 8). As a result, long-term records were broken, with the coldest Tmax observed along the south coast of the Eastern Cape (Fig. 1a). The orientation of the near surface and mid-tropospheric isopleths of constant geopotential is in the same direction on 22 July (Fig. 11a–e), suggesting weak vertical wind shear through the layer. This is supported by the upper sounding profiles which show southwesterlies from near the surface to at least about 200 hPa (Fig. 9a and b). The ELT event followed the passage of an intense cold front which approached Cape Town on the 19th, propagating across South Africa (Fig. 12a), allowing southwesterly air flow to approach the country with a massive fetch on 21–22 July (Fig. 11e and f). The July 19, 2021 EUMETSAT satellite image shows a distinct and well-defined field of cold air cumulus behind the cold front (Fig. 12a), with a cloud pattern depicting cold environmental air. Such a pattern warns bench fore- casters of imminent cold air advection and possibility of an ELT event. The cold front is seen propagating gradually eastwards to lie over the southwest Indian Ocean by the 23rd (Fig. 12b and c). Behind the cold front was an intense Type-S (Ndarana et al., 2022) ridging South Atlantic anticyclone which, in addition to introducing very cold air from the southwest as discussed above, transported moisture into the country. The influence of the ridging process extends well onto the interior plateau as far north as ~20◦S in Zimbabwe (Fig. 12d). Typically, low clouds of a stratiform type (stratocumulus) seen here occur over the Lowveld of South Africa and enter southeast Zimbabwe, associated with ridging anticyclones over the KwaZulu-Natal coast. As the ridging process progresses, the surface winds back from southwesterly to southerly becoming southeasterly over the KwaZulu- Fig. 10. Snow depth (cm, shading) and 1 000-500 hPa thickness (m; thin solid contours) from 00Z–03Z during (a) 22, (b) 23, (c) 24, and (d) July 25, 2021. The dark bold black contour represents the critical 540 dam (5 400 m) thickness isoline while the purple polygon indicates the Highveld. H. Chikoore et al. Weather and Climate Extremes 44 (2024) 100668 13 Fig. 11. Geopotential heights (m) in the mid-troposphere (500 hPa) with wind isotachs (m/s) during (a) 21, (b) 22, (c) 23, and (d) July 24, 2021. The panels (e) to (h) show geopotential heights and winds near the surface (850 hPa) for the corresponding period. H. Chikoore et al. Weather and Climate Extremes 44 (2024) 100668 14 Natal coast (Fig. 11e,f,g,h). The orientation of the isopleths on 23 July as the winds are backing suggest that the geostrophic component of the flow is parallel to the coast (Ndarana et al., 2021) so the component of the flow that is ageostrophic is onshore. This enhances moisture fluxes and cold air advection to penetrate the interior plateau. A strong moisture flux is consistent with the observed Type-S ridging high (Ndarana et al., 2022) and this ridging event is induced by the presence of an anticyclonic wave breaking (Ndarana and Waugh, 2011). The breaking itself is caused by intensification of the extratropical jet streak that is first located just upstream of the Greenwich Meridian on 22 July (Fig. 11b); after which it propagates east and intensifies. As it does, it increases the shearing strain rate on the anticyclonic barotropic shear side of its axis (Nakamura and Plumb, 1994; Ndarana et al., 2023). This caused the wave to break anticyclonically, changing the orientation of the COL, as indicated by the 500 hPa isopleths. The curious issue raised by this case study is that COLs (and their associated wave breaking events) do not have to propagate into South Africa to affect the country. 4. Conclusion Whilst several studies (e.g., Karl et al., 1993; Jury 2018; Kruger and Nxumalo 2017b; van der Walt and Fitchett 2020b; Mateyisi et al., 2021; McBride et al., 2021; IPCC et al., 2021) have found significant warming trends across South Africa consistent with global patterns in the sub- tropics, ELT events remain a key feature of the regional climate. This is despite a reduction in frequency and duration of cold days (Mbokodo et al., 2023) and cold waves in the country (van der Walt and Fitchett 2020b). The frequency of ground frost days over the country has also declined significantly in recent years according to ERA5 data. The coldest period of the year occurs during June and July over the South African Highveld and nearly everywhere else across the country. How- ever, South African winter temperatures are comparatively higher than typical boreal winters in the high latitudes of the Northern Hemisphere. In the Southern Hemisphere, ELTs and cold waves are more common in South America due to its high elevation which reaches 3 000 m over the Andes Mountain Range (e.g., Marengo et al., 2023). We employed station, upper-air, satellite, and model data to analyse an ELT weather event over South Africa that broke 27 long-term records during 22–24 July 2021. We also investigated historical trends in key ETCCDI temperature indices, including the occurrence of ground frost and snow falls with a focus on the South African Highveld. We found that the 2021 austral winter season over South Africa experienced some of the coldest episodes, as long-term records were broken repeatedly throughout July 2021 (including before 22–24 July). The 20–24 July 2021 period was also colder than normal over large areas of southern Africa including neighbouring Botswana and Zimbabwe (not shown), as cold air intrusions penetrated deeper into the interior. Low level guti cloud was shown extending well into the southeast of Zimbabwe as positive sea level pressure anomalies developed over South Africa’s east coast. Snowfalls of up to 2.5 cm were deposited over high ground, mainly over the southern districts (south of 30◦S). Results reveal that both overnight minimum and daytime temperatures were colder than normal across the country, with larger anomalies over the Highveld, hence the focus of this study. For context, the July 2021 ELT event occurred following an anom- alously wet season characterized by positive rainfall and soil moisture anomalies during La Niña over the equatorial Pacific Ocean. Most La Nina conditions are associated with above average seasonal rainfall over large areas of southern Africa resulting in high soil moisture (Chikoore and Jury, 2021). Soil moisture anomalies have been found to affect and modify the atmospheric circulation over an area at seasonal and longer time scales (Shukla and Mintz 1982; Delworth and Manabe, 1988; Qi et al., 2023). We found that circulation anomalies during 22–24 July 2021 were dominated by a combination of a cold front, a Type-S ridging anti- cyclone, and an intense COL aloft. The strong south westerly fetch through a deep layer of the troposphere minimized mixing and resulted in freezing mean temperatures of the 1 000-500 hPa layer. Conse- quently, much of the southern districts of South Africa lay poleward of the critical 540 dam thickness isopleth values, creating conditions conducive for snow falls especially over high ground. A steep 500 hPa trough developed into a COL that extended to the surface over the ocean Fig. 12. EUMETSAT Infrared (IR) satellite images before and during the ELT event on (a) 19, (b) 22, (c) 23 and (d) July 24, 2021. The field of cold air cumulus is prominent behind the cold front. H. Chikoore et al. Weather and Climate Extremes 44 (2024) 100668 15 south of Africa during 22–23 July 2021. COLs that extend to the surface are often intense (Pinheiro et al., 2021) and induce negative tempera- ture anomalies near the surface (Barnes et al., 2021b). A key finding from this case study is that South African weather can be severely affected by COLs (and their associated wave breaking events) located at sea. A ridging anticyclone with a strong meridional fetch from the south/ polar region is often the main weather system associated with ELT events over South Africa. In this event, intensification of the extra- tropical jet streak upstream resulted in anticyclonic wave breaking which, in turn, induced a Type-S ridging event. Jet streaks have been linked to extreme temperature events which may include heatwaves in several regions as they modulate the location and orientation of surface systems (e.g., Chang and Lau 1980; Singleton and Reason 2006). Due to the complex topography characterized by a high plateau and a narrow coastal belt, our results suggest that the impact of cold advection and ELT events is greater over the interior than near coasts. This charac- terization of synoptic circulation anomalies that combined and pro- duced the ELT event of 22–24 July 2021 does not present a generalization for a broader set of ELT events affecting the South African domain. Rather, this paper provides a valuable foundation for future research and more extensive analyses towards development of a typol- ogy of synoptic conditions associated with ELT events. Such a typology may be constructed via Self-Organizing Maps or similar machine learning algorithms. Whilst van der Walt and Fitchett (2020a) proposed an aggregated winter season across South Africa spanning JJA, their study also showed that only the months of June and July are consistently classified as winter across the whole country. They found winter ends early in some districts in the far north and northwest, while persisting for longer in the south. Our study also determined that cool conditions persist into the SON season in those areas. We argue that mid-May to mid-August may be a more realistic definition of the winter season over northern South Africa whilst May to September should apply in the south. Predictability of extreme temperature events at seasonal (e.g., Laz- enby et al., 2014) and medium range weather time scales are critical for early warning systems and natural disaster risk reduction. Understand- ing the role of jet streak dynamics in cold advection and ELT events can lead to better medium range weather forecasts. Whilst observing sys- tems and forecasting capabilities have improved markedly in recent decades, communicating weather alerts and warnings still requires concerted effort. The limitations of the science including implications of uncertainties also need to be communicated effectively. In addition to warming trends, several studies have projected sig- nificant rises in mean surface air temperatures over South (ern) Africa in future (Englebrecht et al. 2015b; Garland et al., 2015; Singo et al. 2023). Future changes in intensity, frequency and return periods of both hot and cold extreme temperature events may become more significant and require further investigation. Periods of ELTs often result in profound impacts on society, winter agriculture and the environment (van der Walt and Fitchett 2020b). Assessments of present and future vulnera- bilities to impacts of ELT hazards are necessary for proactive civil pro- tection and reduction of economic losses. This paper highlights the importance of understanding ELTs, in an era of increasing weather and climate extremes. Government policies should be targeted at increasing societal resilience to a new climate through effective climate change adaptation, also recognizing other extreme events not discussed in this paper. Funding This study received no external funding. CRediT authorship contribution statement Hector Chikoore: Conceptualization, Resources, Writing – original draft, Writing – review & editing, Formal analysis. Innocent L. Mbo- kodo: Formal analysis, Investigation, Writing – original draft, Writing – review & editing. Mukovhe V. Singo: Formal analysis, Investigation, Methodology, Visualization, Writing – original draft, Writing – review & editing. Tumelo Mohomi: Formal analysis, Investigation, Visualization, Writing – original draft, Writing – review & editing. Rendani B. Munyai: Investigation, Writing – original draft, Writing – review & editing. Henno Havenga: Visualization, Writing – original draft, Writing – review & editing. Dawn D. Mahlobo: Writing – original draft, Writing – review & editing. Francois A. Engelbrecht: Writing – original draft, Writing – review & editing. Mary-Jane M. Bopape: Conceptual- ization, Writing – original draft. Thando Ndarana: Conceptualization, Formal analysis, Investigation, Writing – original draft. Declaration of competing interest The authors declare no conflict of interest. Data availability The authors do not have permission to share data. Acknowledgements Daily and monthly temperature and rainfall data, records of ex- tremes and synoptic maps we analysed here were obtained from the South African Weather Service (SAWS). The ERA5 and ERA5-Land reanalyses were obtained from the ECMWF. The University of East Anglia’s CRU Timeseries 4 (TS 4) data were also analysed. APPENDICES. Appendix a. Mann-Kendall and Sen’s slope trend test for Tmin, ground frost days and Tmax Variable p-value Z score Sen’s slope Tmin 2.2 × 10 − 16 22.632 0.0011 ◦C/month Tmax 2.2 × 10 − 16 19.865 0.0014 ◦C/month Ground Frost 2.2 × 10 − 16 − 17.993 − 0.0003 days/month H. Chikoore et al. Weather and Climate Extremes 44 (2024) 100668 16 Appendix b. RClimDex slope trend test (trends computed by linear least square and locally weighted linear regression) for COLD NIGHTS, COOL DAYS, WARM NIGHTS and HOT DAYS over the Johannesburg Intl Airport station on the South African Highveld, one of the stations that broke records during July 2021. The period did not include 2021 which had extreme old temperature anomalies. Indices Start year End year Slope P_Value TX10P 1991 2020 − 0.319 0.021 TX90P 1991 2020 0.377 0.022 TN10P 1991 2020 − 0.236 0.02 TN90P 1991 2020 0.65 0 References Abiodun, B.J., Ojumu, A.M., Jenner, S., Ojumu, T.V., 2014. The transport of atmospheric NOx and HNO3 over Cape Town. Atmos. Chem. Phys. 14 (2), 559–575. Abram, N.J., Mulvaney, R., Vimeux, F., co-authors, 2014. Evolution of the southern annular mode during the past millennium. Nat. Clim. Change 4 (7), 564–569. Archer, E.R.M., Landman, W.A., Malherbe, J., Maluleke, P., Weepener, H., 2021. Managing climate risk in livestock production in South Africa: how might improved tailored forecasting contribute? Climate Risk Management 32, 100312. Barnes, M.A., Turner, K., Ndarana, T., Landman, W.A., 2021a. Cape storm: a dynamical study of a cut-off low and its impact on South Africa. Atmos. Res. 249 https://doi. org/10.1016/j.atmosres.2020.105290. Barnes, M.A., Ndarana, T., Landman, W.A., 2021b. Cut-off lows in the Southern Hemisphere and their extension to the surface. Clim. Dynam. 56 (11–12), 3709–3732. Bopape, M.M., Keebine, G., Ndarana, T., co-authors, 2024. Weather related disasters in South Africa from 1979 to 2022. Nat. Commun. submitted for publication. Burls, N.J., Blamey, R.C., Cash, B.A., co-authors, 2019. The Cape Town “Day Zero” drought and Hadley Cell expansion. npj Climate and Atmospheric Science 2, 27. https://doi.org/10.1038/s41612-019-0084-6. Caelum, 2022. In: Notable Weather and Weather-Related Events in Geo-Coded Format: 1647–2021. South African Weather Service. 2021 Available online: https://www. weathersa.co.za/Documents/AnnualReports/SA_Weather_Services_Annual_Report_2 020_21_Digital_Version_14092021122723.pdf. (Accessed 15 August 2022). Catto, J.L., Ackerley, D., Booth, J.F., Champion, A.J., Colle, B.A., Pfahl, S., Pinto, J.G., Quinting, J.F., Seiler, C., 2019. The future of midlatitude cyclones. Curr. Clim. Change Rep. 5, 407–420. Chang, C.P., Lau, K.M.W., 1980. Northeasterly cold surges and near-equatorial disturbances over the winter MONEX area during December 1974. Part II: Planetary- scale aspects. Mon. Weather Rev. 108 (3), 298–312. Chikoore, H., Jury, M.R., 2021. South African drought, deconstructed. Weather and Climate Extremes 33, 100334. Delworth, T.L., Manabe, S., 1988. The influence of potential evaporation on the variabilities of simulated soil wetness and climate. J. Clim. 1, 523–547. Engelbrecht, C.J., Landman, W.A., Engelbrecht, F.A., 2015a. Synoptic decomposition of rainfall over the Cape South Coast. Clim. Dynam. 44, 2589–2607. Engelbrecht, F., Adegoke, J., Bopape, M.J., co-authors, 2015b. Projections of rapidly rising surface temperatures over Africa under low mitigation. Environ. Res. Lett. 10 (8), 085004. Eskom, 2021. Stage 2 loadshedding will be implemented from 16:00 until 21:00 tonight due to shortage of generation capacity. https://www.eskom.co.za/stage-2-loadsh edding-will-be-implemented-from-1600-until-2100-tonight-due-to-shortage-of-g eneration-capacity/. (Accessed 3 March 2024). Farmer’s Weekly, 2010. Black frost will see vegetable prices rocket. https://www.farme rsweekly.co.za/archive/black-frost-will-see-vegetable-prices-rocket/. (Accessed 3 March 2024). Farmer’s Weekly, 2023. Cold weather hits large parts of South Africa. https://www.far mersweekly.co.za/agri-news/south-africa/cold-weather-hits-large-parts-of-sout h-africa/. (Accessed 3 March 2024). Favre, A., Hewitson, B., Lennard, C., Cerezo-Mota, R., Tadross, M., 2013. Cut-off lows in the South Africa region and their contribution to precipitation. Clim. Dynam. 41, 2331–2351. Fogt, R.L., Marshall, G.J., 2020. The southern annular mode: variability, trends, and climate impacts across the southern hemisphere. Wiley Interdisciplinary Reviews: Clim. Change 11 (4), e652. Garland, R.M., Matooane, M., Engelbrecht, F.A., co-authors, 2015. Regional projections of extreme apparent temperature days in Africa and the related potential risk to human health. Int. J. Environ. Res. Publ. Health 12 (10), 12577–12604. Gillett, N.P., Kell, T.D., Jones, P.D., 2006. Regional climate impacts of the southern annular mode. Geophys. Res. Lett. 33 (23). Grumm, R.H., 1998. Adding or degrading model forecast: anatomy of a poorly forecast winter storm. NOAA US National Weather Service. https://repository.library.noaa. gov/view/noaa/6683/noaa_6683_DS1.pdf. Harman, B.I., Koseoglu, H., Yigit, C., 2016. Performance evaluation of IDW, Kriging, and multiquadric interpolation methods in producing noise mapping: a case study at the city of Isparta, Turkey. Appl. Acoust. 112 (2), 147–157. https://doi.org/10.1016/j. apacoust.2016.05.024. Harris, I., Osborn, T.J., Jones, P., Lister, D., 2020. Version 4 of the CRU TS monthly high- resolution gridded multivariate climate dataset. Sci. Data 7 (1), 109. Hersbach, H., Bell, B., Berrisford, P., co-authors, 2020. The ERA5 global reanalysis. Quarterly Journal of the Royal Meteorolgical Society. https://doi.org/10.1002/ qj.3803. Howard, E., Washington, R., 2018. Characterizing the synoptic expression of the Angola Low. J. Clim. 31 (9), 7147–7166. https://doi.org/10.1175/JCLI-D-18-0017.1. IPCC, 2021. In: Masson-Delmotte, V., Zhai, P., Pirani, A., Connors, S.L., Péan, C., Berger, S., Caud, N., Chen, Y., Goldfarb, L., Gomis, M.I., Huang, M., Leitzell, K., Lonnoy, E., Matthews, J.B.R., Maycock, T.K., Waterfield, T., Yelekçi, O., Yu, R., Zhou, B. (Eds.), Climate Change 2021: the Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA. https://doi.org/10.1017/9781009157896 (in press). Jury, M.R., 2017. Statistics and meteorology of air pollution episodes over the South African Highveld based on satellite–model datasets. J. Appl. Meteorol. Climatol. 56 (6), 1583–1594. Jury, M.R., 2018. Climate trends across South Africa since 1980. WaterSA 44. https:// doi.org/10.4314/wsa.v44i2.15. Jury, M.R., Freiman, T., 2002. The climate of tropical southern Africa during the SAFARI 2000 campaign: research in action. South AfricanJjournal of Science 98 (11), 527–533. Karl, T.R., Jones, P.D., Knight, R., Kukla, G., Plummer, N., Razuvzyev, V., co-authors, 1993. Asymmetric trends of daily maximum and minimum temperature. Bull. Am. Meteorol. Soc. 74 (6), 1007–1023. https://doi.org/10.1175/1520-0477(1933) 074<1007:anporg>2.0.co;2. Klopper, E., Landman, W.A., Van Heerden, J., 1998. The predictability of seasonal 50. maximum temperature in South Africa. Int. J. Climatol. 18 (7), 741–758. Kruger, A.C., Nxumalo, M.P., 2017a. Historical rainfall trends in South Africa: 1921–2015. WaterSA 43 (2), 285–297. Kruger, A.C., Nxumalo, M.P., 2017b. Surface temperature trends in homogenized time series in South Africa: 1931-2015. Int. J. Climatol. 37, 2364–2377. Kruger, A.C., Sekele, S.S., 2013. Trends in extreme temperature indices in South Africa: 1962–2009. Int. J. Climatol. 33 (3), 661–676. Kruger, A.C., Rautenbach, H., Mbatha, S., Ngwenya, S., Makgoale, T.E., 2019. Historical and projected trends in near-surface temperature indices for 22 locations in South Africa. South Afr. J. Sci. 115 (5–6), 1–9. Lazenby, M.J., Landman, W.A., Garland, R.M., Dewitt, D.G., 2014. Seasonal temperature prediction skill over southern Africa and human health. Meteorol. Appl. 21 (4), 963–974. Lee, D.Y., Petersen, M.R., Lin, W., 2019. The southern annular mode and Southern Ocean surface westerly winds in E3SM. Earth Space Sci. 6 (12), 2624–2643. Leroux, M., 2001. The Meteorology and Climate of Tropical Africa. Springer Science & Business Media. Limpasuvan, V., Hartmann, D.L., 2000. Wave-maintained annular modes of climate variability. J. Clim. 13, 4414–4429. Mahlalela, P.T., Blamey, R.C., Reason, C.J.C., 2019. Mechanisms behind early winter rainfall variability in the south western Cape, South Africa. Clim. Dynam. 53, 21–39. Mail and Guardian, 2023. Cold weather puts an end to day-time relief in load-shedding. https://mg.co.za/news/2023-07-10-cold-weather-puts-an-end-to-day-time-relief-in- load-shedding/. (Accessed 3 March 2024). Maisha, T.R., 2014. The influence of topography and model grid resolution on extreme weather forecasts over South Africa. MSc dissertation. University of Pretoria. Marengo, J., Espinoza, J.C., Bettolli, L., Cunha, A.P., Molina-Carpio, J., Skansi, M., Correa, K., Ramos, A.M., Salinas, R., Sierra, J.-P., 2023. A cold wave of winter 2021 in central South America: characteristics and impacts. Clim. Dynam. https://doi.org/ 10.1007/s00382-023-06701-1. Mashao, F.M., Mothapo, M.C., Munyai, R.B., Letsoalo, J.M., Mbokodo, I.L., Muofhe, T.P., Matsane, W., Chikoore, H., 2023. Extreme rainfall and flood risk prediction over the East Coast of South Africa. Water 15 (1), 50. Mateyisi, M.J., Maoela, M.A., Maluleke, A., Moeletsi, M.E., Von Maltitz, G., 2021. Changes in annual extreme temperature and heat indices in Limpopo province: period 1941-2016. Theor. Appl. Climatol. https://doi.org/10.1007/s00704-020- 03511-x. Mbokodo, I.L., 2017. Heatwaves in South Africa: Observed Variability, Structure and Trends. University of Venda. MSc dissertation. Mbokodo, I.L., Bopape, M.M., Chikoore, H., Engelbrecht, F.A., Nethengwe, N.S., 2020. Heatwaves in the future warmer climate of South Africa. Atmosphere 11 (7), 712. Mbokodo, I.L., Bopape, M.M., Ndarana, T., co-authors, 2023. Heatwave variability and Structure in South Africa during summer drought. Climate 11 (2), 38. H. Chikoore et al. http://refhub.elsevier.com/S2212-0947(24)00029-X/sref1 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref1 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref2 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref2 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref3 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref3 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref3 https://doi.org/10.1016/j.atmosres.2020.105290 https://doi.org/10.1016/j.atmosres.2020.105290 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref5 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref5 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref5 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref6 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref6 https://doi.org/10.1038/s41612-019-0084-6 https://www.weathersa.co.za/Documents/AnnualReports/SA_Weather_Services_Annual_Report_2020_21_Digital_Version_14092021122723.pdf https://www.weathersa.co.za/Documents/AnnualReports/SA_Weather_Services_Annual_Report_2020_21_Digital_Version_14092021122723.pdf https://www.weathersa.co.za/Documents/AnnualReports/SA_Weather_Services_Annual_Report_2020_21_Digital_Version_14092021122723.pdf http://refhub.elsevier.com/S2212-0947(24)00029-X/sref9 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref9 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref9 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref10 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref10 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref10 http://refhub.elsevier.com/S2212-0947(24)00029-X/optCyLYdlBRAD http://refhub.elsevier.com/S2212-0947(24)00029-X/optCyLYdlBRAD http://refhub.elsevier.com/S2212-0947(24)00029-X/sref11 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref11 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref12 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref12 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref13 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref13 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref13 https://www.eskom.co.za/stage-2-loadshedding-will-be-implemented-from-1600-until-2100-tonight-due-to-shortage-of-generation-capacity/ https://www.eskom.co.za/stage-2-loadshedding-will-be-implemented-from-1600-until-2100-tonight-due-to-shortage-of-generation-capacity/ https://www.eskom.co.za/stage-2-loadshedding-will-be-implemented-from-1600-until-2100-tonight-due-to-shortage-of-generation-capacity/ https://www.farmersweekly.co.za/archive/black-frost-will-see-vegetable-prices-rocket/ https://www.farmersweekly.co.za/archive/black-frost-will-see-vegetable-prices-rocket/ https://www.farmersweekly.co.za/agri-news/south-africa/cold-weather-hits-large-parts-of-south-africa/ https://www.farmersweekly.co.za/agri-news/south-africa/cold-weather-hits-large-parts-of-south-africa/ https://www.farmersweekly.co.za/agri-news/south-africa/cold-weather-hits-large-parts-of-south-africa/ http://refhub.elsevier.com/S2212-0947(24)00029-X/sref17 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref17 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref17 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref18 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref18 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref18 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref19 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref19 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref19 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref20 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref20 https://repository.library.noaa.gov/view/noaa/6683/noaa_6683_DS1.pdf https://repository.library.noaa.gov/view/noaa/6683/noaa_6683_DS1.pdf https://doi.org/10.1016/j.apacoust.2016.05.024 https://doi.org/10.1016/j.apacoust.2016.05.024 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref23 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref23 https://doi.org/10.1002/qj.3803 https://doi.org/10.1002/qj.3803 https://doi.org/10.1175/JCLI-D-18-0017.1 https://doi.org/10.1017/9781009157896 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref27 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref27 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref27 https://doi.org/10.4314/wsa.v44i2.15 https://doi.org/10.4314/wsa.v44i2.15 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref29 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref29 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref29 https://doi.org/10.1175/1520-0477(1933)074<1007:anporg>2.0.co;2 https://doi.org/10.1175/1520-0477(1933)074<1007:anporg>2.0.co;2 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref31 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref31 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref32 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref32 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref33 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref33 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref34 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref34 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref35 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref35 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref35 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref36 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref36 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref36 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref37 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref37 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref38 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref38 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref39 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref39 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref40 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref40 https://mg.co.za/news/2023-07-10-cold-weather-puts-an-end-to-day-time-relief-in-load-shedding/ https://mg.co.za/news/2023-07-10-cold-weather-puts-an-end-to-day-time-relief-in-load-shedding/ http://refhub.elsevier.com/S2212-0947(24)00029-X/optSkWITl4zXj http://refhub.elsevier.com/S2212-0947(24)00029-X/optSkWITl4zXj https://doi.org/10.1007/s00382-023-06701-1 https://doi.org/10.1007/s00382-023-06701-1 http://refhub.elsevier.com/S2212-0947(24)00029-X/opt7Dr1u0NJmR http://refhub.elsevier.com/S2212-0947(24)00029-X/opt7Dr1u0NJmR http://refhub.elsevier.com/S2212-0947(24)00029-X/opt7Dr1u0NJmR https://doi.org/10.1007/s00704-020-03511-x https://doi.org/10.1007/s00704-020-03511-x http://refhub.elsevier.com/S2212-0947(24)00029-X/sref44 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref44 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref45 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref45 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref46 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref46 Weather and Climate Extremes 44 (2024) 100668 17 McBride, C.M., Kruger, A.C., Dyson, L., 2021. Trends in probabilities of temperature records in the non-stationary climate of South Africa. Int. J. Climatol. https://doi. org/10.1002/joc.7329. Menne, M.J., Durre, I., Vose, R.S., Gleason, B.E., Houston, T.G., 2012. An overview of the global historical climatology network-daily database. J. Atmos. Ocean. Technol. 29 (7), 897–910. Moeletsi, M.E., Tongwane, M., Tsubo, M., 2016. The study of frost occurrence in Free State province of South Africa. Adv. Meteorol. https://doi.org/10.1155/2016/ 9586150. Munoz-Sebater, J., Dutra, E., Agusti-Panareda, A., 2021. ERA5-Land: a state-of-the-art global reanalysis dataset for land applications. Earth Syst. Sci. Data 13 (9), 4349–4383. Muofhe, T.P., Chikoore, H., Bopape, M.M., co-authors, 2020. Forecasting intense cut-off lows in South Africa using the 4.4 km Unified Model. Climate 8, 129. https://doi. org/10.3390/cli8110129. Myburg, J., 1974. An index to relate local topography to mean minimum temperatures. Agrochemophysica 6, 73–78. Nakamura, M., Plumb, R.A., 1994. The effects of flow asymmetry on the direction of Rossby wave breaking. J. Atmos. Sci. 51 (14), 2031–2045. Ndarana, T., Lekoloane, L.E., Rammopo, T.S., Reason, C.J.C., Bopape, M.M., Chikoore, H., Engelbrecht, F.E., 2023. Downstream development during ridging South Atlantic Ocean anticyclones. Clim. Dynam. https://doi.org/10.1007/s00382- 023-06717-7. Ndarana, T., Mpati, S., Bopape, M.M., Engelbrecht, F.A., Chikoore, H., 2021. The flow and moisture fluxes associated with ridging South Atlantic Ocean anticyclones during the subtropical southern African summer. Int. J. Climatol. https://doi.org/ 10.1002/joc.6745. Ndarana, T., Rammopo, T.S., Reason, C.J.C., Bopape, M.M., Engelbrecht, F.A., Chikoore, H., 2022. Two types of ridging South Atlantic Ocean anticyclones over South Africa and the associated dynamical processes. Atmos. Res. 265, 105897. Ndarana, T., Waugh, D.W., 2011. A climatology of Rossby wave breaking on the Southern Hemisphere tropopause. J. Atmos. Sci. 68 (4), 798–811. Nicholson, S.E., 2000. The nature of rainfall variability over Africa on time scales of decades to millennia. Global Planet. Change 26 (1–3), 137–158. Omar, S.A., Abiodun, B.J., 2020. Characteristics of cut-off lows during the 2015–2017 drought in the Western Cape, South Africa. Atmos. Res. 235, 104772. Omar, S.A., Abiodun, B.J., 2021. Simulating the influence of topography on cut-off lows over Southern Africa. Int. J. Climatol. 41, E2231–E2243. Ongoma, V., Chen, H., 2017. Temporal and spatial variability of temperature and precipitation over East Africa from 1951 to 2010. Meteorol. Atmos. Phys. 129, 131–144. Palmén, E., Newton, C.W., 1969. Atmospheric Circulation Systems: Their Structure and Physical Interpretation. Academic press. Perry, K.B., 1998. Basics of frost and freeze protection for horticultural crops. HortTechnology 8. https://doi.org/10.21273/HORTTECH.8.1.10. Picas, J., Grab, S., 2020. Reconstruction of cold front frequency over Cape Town, South Africa, using daily mean sea level pressure values: 1834-1899. Int. J. Climatol. https://doi.org/10.1002/joc.6930. Pinheiro, H., Gan, M., Hodges, K., 2021. Structure and evolution of intense austral cut-off lows. Quarterly Journal of Royal Meteorological Society 147, 1–20. https://doi.org/ 10.1002/qj.3900. Qi, Y., Chen, H., Zhu, S., 2023. Influence of land-atmosphere coupling on low temperature extremes over southern Eurasia. Journal of Geophysical Research Atmospheres. https://doi.org/10.1029/2022JD037252. Reason, C.J.C., Rouault, M., 2005. Links between the Antarctic Oscillation and winter rainfall over western South Africa. Geophys. Res. Lett. 32, 1–4. https://doi.org/ 10.1029/2005GL022419. Salinger, M.J., Griffiths, G.M., 2001. Trends in New Zealand daily temperature and rainfall extremes. Int. J. Climatol. 21 (12), 1437–1452. Singo, M.V., Chikoore, H., Engelbrecht, F.A., Ndarana, T., Muofhe, T.P., Mbokodo, I.L., Murungweni, F.M., Bopape, M.M., 2023. Projections of future fire risk under climate change over the South African savanna. Stochastic Environ. Res. Risk Assessm. 37 (7), 2677–2691. Sousa, P.M., Blamey, R.C., Reason, C.J., Ramos, A.M., Trigo, R.M., 2018. The ‘Day Zero’Cape Town drought and the poleward migration of moisture corridors. Environ. Res. Lett. 13 (12), 124025. SAWS, 2021. Climate Summary of South Africa: July 2021, vol. 32 (7), ISSN 1992-2566. SAWS, 2023. Annual State of the Climate of South Africa 2022. Pretoria. South Africa. Shukla, J., Mintz, Y., 1982. Influence of land-surface evapotranspiration on the earth’s climate. Science 215, 1498–1501. Singleton, A.T., Reason, C.J.C., 2006. Numerical simulations of a severe rainfall event over the Eastern Cape coast of South Africa: sensitivity to sea surface temperature and topography. Tellus Dyn. Meteorol. Oceanogr. 58 (3), 335–367. Singleton, A.T., Reason, C.J.C., 2007. Variability in the characteristics of cut-off low pressure systems over subtropical southern Africa. Int. J. Climatol. 27, 295–310. Stander, J.H., Dyson, L.L., Engelbrecht, C.J., 2016. A snow forecasting decision tree for significant snowfall over the interior of South Africa. South Afr. J. Sci. 112 https:// doi.org/10.17159/sajs.2016/20150221. STATSSA, 2023. Census 2022. https://census.statssa.gov.za/#/province/9/2. (Accessed 3 March 2024). The New York Times, 2023. Johannesburg sees first snowfall in over a decade. Residents flock to the flakes. https://www.nytimes.com/2023/07/12/world/africa/south-afr ica-snow-johannesburg.html. Thompson, D.W., Solomon, S., Kushner, P.J., co-authors, 2011. Signatures of the Antarctic ozone hole in Southern Hemisphere surface climate change. Nat. Geosci. 4 (11), 741–749. Timeslive, 2021. Some relief as SA moves from cold to ‘cool’ this week. https://www. timeslive.co.za/news/south-africa/2021-07-26-some-relief-as-sa-to-ease -from-cold-to-cool-this-week/#google_vignette. (Accessed 3 March 2024). Tshiala, F.M., Olwoch, J.M., Engelbrecht, F.A., 2011. Analysis of temperature trends over Limpopo province, South Africa. J. Geogr. Geol. 3 (1), 13–21. https://doi.org/ 10.5539/jgg.v3n1p13. Torrance, J.D., 1981. Climate Handbook of Zimbabwe. Dept of Meteorological Services, Harare, p. 221. Tyson, P.D., 1986. Climatic Change and Variability over Southern Africa. Oxford University Press, Cape Town. Tyson, P.D., Preston-Whyte, R.A., 2000. Weather and Climate of Southern Africa. Oxford University Press, p. 396. Van Der Walt, A., Fitchett, J.M., 2020a. Statistical classification of South African seasonal divisions on the basis of daily temperature data. South Afr. J. Sci. 116 https://doi. org/10.17159/sajs.2020/7614. Van Der Walt, A., Fitchett, J.M., 2020b. Trend analysis of cold extremes in South Africa: 1960-2016. Int. J. Climatol. https://doi.org/10.1002/joc.6947. Van Schalkwyk, L., 2011. Fog Forecasting at Cape Town International Airport: a Climatological Approach. University of Pretoria. MSc dissertation. Van Schalkwyk, L., Dyson, L.L., 2013. Climatological characteristics of fog at Cape Town International airport. Weather Forecast. 28 (3), 631–646. Viljoen, F., 1991. Caelum: A History of Notable Weather Events in South Africa: 1500- 1990. Department of Environment Affairs. Wallace, J.M., Hobbs, P.V., 2006. Atmospheric science: an introductory survey, 2nd edition. Academic Press, pp. 504–pp. Watson, D.F., Philip, G.M., 1985. A refinement of inverse distance weighted interpolation. Geo Process. 2, 315–327. Wu, Y., Xu, Y., 2005. Snow impact on groundwater recharge in Table Mountain Group aquifer systems with a case study of the Kommissiekraal River catchment, South Africa. WaterSA 31, 275–282. Xulu, N.G., Chikoore, H., Bopape, M.M., Ndarana, T., Muofhe, T.P., Mbokodo, I.L., Munyai, R.B., Singo, M.V., Mohomi, T., Mbatha, S.M., Mdoka, M.L., 2023. Cut-Off Lows over South Africa: A Review. Climate 11 (3), 59. Yongguang, H., Chen, Z., Pengfei, L., Amoah, A.E., Pingping, L., 2016. Sprinkler irrigation system for tea frost protection and the application effect. Int. J. Agric. Biol. Eng. 9 (5), 17–23. Zhang, X., Yang, F., 2004. RClimDex (1.0) user manual. Climate Research Branch Environment Canada 22, 13–14. H. Chikoore et al. https://doi.org/10.1002/joc.7329 https://doi.org/10.1002/joc.7329 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref48 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref48 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref48 https://doi.org/10.1155/2016/9586150 https://doi.org/10.1155/2016/9586150 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref50 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref50 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref50 https://doi.org/10.3390/cli8110129 https://doi.org/10.3390/cli8110129 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref53 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref53 http://refhub.elsevier.com/S2212-0947(24)00029-X/optfdsgzDmlwM http://refhub.elsevier.com/S2212-0947(24)00029-X/optfdsgzDmlwM https://doi.org/10.1007/s00382-023-06717-7 https://doi.org/10.1007/s00382-023-06717-7 https://doi.org/10.1002/joc.6745 https://doi.org/10.1002/joc.6745 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref55 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref55 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref55 http://refhub.elsevier.com/S2212-0947(24)00029-X/optrQ2P6LwNyn http://refhub.elsevier.com/S2212-0947(24)00029-X/optrQ2P6LwNyn http://refhub.elsevier.com/S2212-0947(24)00029-X/sref56 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref56 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref57 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref57 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref58 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref58 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref59 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref59 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref59 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref60 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref60 https://doi.org/10.21273/HORTTECH.8.1.10 https://doi.org/10.1002/joc.6930 https://doi.org/10.1002/qj.3900 https://doi.org/10.1002/qj.3900 https://doi.org/10.1029/2022JD037252 https://doi.org/10.1029/2005GL022419 https://doi.org/10.1029/2005GL022419 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref66 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref66 http://refhub.elsevier.com/S2212-0947(24)00029-X/optWzHQbdyDxL http://refhub.elsevier.com/S2212-0947(24)00029-X/optWzHQbdyDxL http://refhub.elsevier.com/S2212-0947(24)00029-X/optWzHQbdyDxL http://refhub.elsevier.com/S2212-0947(24)00029-X/optWzHQbdyDxL http://refhub.elsevier.com/S2212-0947(24)00029-X/sref68 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref68 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref68 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref69 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref70 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref71 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref71 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref72 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref72 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref72 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref73 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref73 https://doi.org/10.17159/sajs.2016/20150221 https://doi.org/10.17159/sajs.2016/20150221 https://census.statssa.gov.za/#/province/9/2 https://www.nytimes.com/2023/07/12/world/africa/south-africa-snow-johannesburg.html https://www.nytimes.com/2023/07/12/world/africa/south-africa-snow-johannesburg.html http://refhub.elsevier.com/S2212-0947(24)00029-X/sref77 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref77 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref77 https://www.timeslive.co.za/news/south-africa/2021-07-26-some-relief-as-sa-to-ease-from-cold-to-cool-this-week/#google_vignette https://www.timeslive.co.za/news/south-africa/2021-07-26-some-relief-as-sa-to-ease-from-cold-to-cool-this-week/#google_vignette https://www.timeslive.co.za/news/south-africa/2021-07-26-some-relief-as-sa-to-ease-from-cold-to-cool-this-week/#google_vignette https://doi.org/10.5539/jgg.v3n1p13 https://doi.org/10.5539/jgg.v3n1p13 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref79 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref79 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref81 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref81 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref82 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref82 https://doi.org/10.17159/sajs.2020/7614 https://doi.org/10.17159/sajs.2020/7614 https://doi.org/10.1002/joc.6947 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref86 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref86 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref87 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref87 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref88 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref88 http://refhub.elsevier.com/S2212-0947(24)00029-X/optTzamWUHvXF http://refhub.elsevier.com/S2212-0947(24)00029-X/optTzamWUHvXF http://refhub.elsevier.com/S2212-0947(24)00029-X/sref89 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref89 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref90 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref90 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref90 http://refhub.elsevier.com/S2212-0947(24)00029-X/optpUDPQX08Jj http://refhub.elsevier.com/S2212-0947(24)00029-X/optpUDPQX08Jj http://refhub.elsevier.com/S2212-0947(24)00029-X/optpUDPQX08Jj http://refhub.elsevier.com/S2212-0947(24)00029-X/sref91 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref91 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref91 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref92 http://refhub.elsevier.com/S2212-0947(24)00029-X/sref92 Dynamics of an extreme low temperature event over South Africa amid a warming climate 1 Introduction 2 Data and methods 2.1 The South African Highveld 2.2 Analysis of observed and reanalysis data 3 Results and discussion 3.1 Climatologies 3.1.1 Seasonal mean temperatures 3.1.2 Seasonal cycles and interannual variability of Tmin, Tmax and ground frosts 3.1.3 Trends in temperature extremes over Johannesburg airport 3.1.4 Snow occurrence over South Africa 3.2 Case study: ELT weather during 22–25 July 2021 3.2.1 Temperature observations and anomalies 3.2.2 Circulation dynamics during 22–25 July 2021 4 Conclusion Funding CRediT authorship contribution statement Declaration of competing interest Data availability Acknowledgements APPENDICES Acknowledgements Appendix a Mann-Kendall and Sen’s slope trend test for Tmin, ground frost days and Tmax Appendix b RClimDex slope trend test (trends computed by linear least square and locally weighted linear regression) for CO ... References