5. Table of Contents
| Authors and affiliations | S280 |
| a. Overview | S283 |
| b. Atmosphere | S285 |
| 1. The arctic troposphere in 2023 | S286 |
| 2. The arctic stratosphere in 2023 | S286 |
| Sidebar 5.1: The February 2023 major sudden stratospheric warming | S288 |
| c. Surface air temperature | S291 |
| 1. Brief summary of impacts and overview | S291 |
| 2. Annual perspectives | S291 |
| 3. Seasonal patterns | S291 |
| Sidebar 5.2: Summer 2023 weather and climate impacts | S293 |
| d. Precipitation | S295 |
| 1. Introduction | S295 |
| 2. 2023 Summary | S295 |
| 3. Regional anomalies | S295 |
| 4. Historical perspective | S296 |
| 5. Heavy precipitation events | S296 |
| e. Sea-surface temperature | S298 |
| f. Sea ice | S301 |
| 1. Sea-ice extent | S301 |
| 2. Sea-ice age, thickness, and volume | S302 |
| g. Greenland Ice Sheet | S304 |
| h. Glaciers and ice caps outside Greenland | S307 |
| i. Terrestrial snow cover | S311 |
| j. Permafrost | S314 |
| 1. Permafrost temperatures | S314 |
| 2. Active layer thickness | S316 |
| k. Tundra greenness | S318 |
| Appendix 1: Acronyms | S321 |
| Appendix 2: Datasets and sources | S322 |
| References | S326 |
a. Overview
—M. L. Druckenmiller, R. L. Thoman, and T. A. Moon
Arctic observations in 2023 provided clear evidence of rapid and pronounced climate and environmental change, shaped by past and ongoing human activities that release greenhouse gases into the atmosphere and push the broader Earth system into uncharted territory. This chapter provides a snapshot of 2023 and summarizes decades-long trends observed across the Arctic, including warming surface air and sea-surface temperatures, decreasing snow cover, diminishing sea ice, thawing permafrost, and continued mass loss from the Greenland Ice Sheet and Arctic glaciers. These changes are driving a transition to a wetter, greener, and less frozen Arctic, with serious implications for Arctic peoples and ecosystems, as well as for low- and midlatitudes.
Average surface air temperatures for 2023 (January–December) for the Arctic as a whole were the fourth highest since 1900, with the Arctic summer (July–September) being the warmest on record. These unprecedented surface temperatures aligned with record-positive geopotential height anomalies in the polar troposphere, which have been increasing alongside warming air temperatures since 1958, indicating the strong connection between long-term atmospheric circulation and regional temperature patterns.
Large-scale atmospheric circulation also strongly influences year-to-year variability and regional differences. For example, in 2023, a colder-than-normal spring across Alaska slowed snowpack and sea-ice melt, while parts of north-central Canada experienced their highest spring average temperatures on record. Short-term atmospheric events can also influence Arctic and midlatitude connections. A major Arctic sudden stratospheric warming (SSW) event in February 2023 is described in Sidebar 5.1—an event that can increase the likelihood of midlatitude cold-air outbreaks for several weeks to months, influencing subseasonal-to-seasonal predictability for midlatitude surface weather.
Warming seasonal air temperatures together with the timing and extent of summer sea-ice loss significantly influence multi-decadal trends and the substantial regional and year-to-year variability seen across both marine and terrestrial systems. Driven by accelerated sea-ice retreat and melt that started in July, the September 2023 sea-ice monthly extent, which is the lowest monthly extent of the year, was 4.37 million square kilometers—about 10% lower than the past two years and overall the fifth lowest in the 45-year satellite record. Additionally, the 17 lowest September sea-ice monthly extents have all occurred in the last 17 years. Spring and early-summer sea-ice loss exposes the dark ocean surface and allows time for solar heating of the ocean. Linked to early sea-ice loss, average sea-surface temperatures for August 2023 were much higher than the 30-year average in the Barents, Kara, Laptev, and Beaufort Seas. Anomalously low August 2023 sea-surface temperatures were observed in Baffin Bay and parts of the Greenland, Bering, and Chukchi Seas. Despite considerable year-to-year variability, almost all Arctic Ocean and marginal seas studied show a statistically significant 1982–2023 warming trend.
On land, the Arctic tundra is greening due to its sensitivity to rapidly increasing summer temperatures, as well as to rapidly evolving sea-ice, snow, and permafrost conditions. In 2023, circumpolar average peak tundra greenness was the third highest in the 24-year Moderate Resolution Imaging Spectroradiometer (MODIS) satellite record, a slight decline from the previous year. Closely aligned with air temperatures and nearshore sea-ice anomalies, peak vegetation greenness in 2023 was much higher than usual in the North American tundra, particularly in the Beaufort Sea region. In contrast, tundra greenness was relatively low in the Eurasian Arctic, particularly in northeastern Siberia.
Long-term changes in permafrost conditions are also largely controlled by changes in air temperature. Across all Arctic regions, permafrost temperatures and active layer thickness (i.e., thickness of the soil layer above the permafrost that seasonally thaws and freezes) continue to increase on decadal time scales. In 2023, permafrost temperatures were the highest on record at over half of the reporting sites across the Arctic. Permafrost thaw disrupts Arctic communities and infrastructure and can also affect the rate of greenhouse gas release to the atmosphere, potentially accelerating global warming.
Analyses of Arctic precipitation reveal additional connections between a changing atmosphere and land. Precipitation in 2023 was above normal in all seasons for the Arctic as a whole, with short-duration heavy precipitation events breaking existing records at various locations. Arctic precipitation in the past year was also marked by important seasonal and regional variations. Unusually low precipitation and high temperatures produced severe drought and contributed to the record-breaking wildfire season in Canada’s Northwest Territory. Snowpack in early spring 2023 was above normal for North America and Eurasia, but then rapid snow loss in much of the Arctic resulted in record-low average snow-water equivalent for the North American Arctic in May and near-record-low snow cover for the Eurasian Arctic in June.
Precipitation patterns also influence the Greenland Ice Sheet. Above-average snowfall over parts of the Greenland Ice Sheet between autumn 2022 and spring 2023 contributed to a relatively low (for the twenty-first century) total mass loss from the Greenland Ice Sheet despite extensive late-June-to-September ice melt. So, while the Greenland Ice Sheet lost mass in the past year, as it has every year since 1998, the loss for September 2022 to August 2023 was much lower than the 22-year average and similar to that of 2020/21. However, the cumulative melt-day area during summer 2023 was the second-highest in the 45-year satellite observational record.
Beyond the Greenland Ice Sheet, the Arctic’s other glaciers and ice caps show a continuing trend of significant ice loss, especially in Alaska and Arctic Canada. All of the 25 monitored Arctic glaciers reported in this chapter for the 2022/23 mass balance year show an annual loss of ice, and for many glaciers these data indicate continued rapid wastage with substantial total contributions to global sea level.
The exceptionally warm Arctic summer alongside persistent long-term climate changes contributed to a range of societal and environmental impacts in 2023 (see Sidebars 5.2 and 7.1), providing stark reminders of the varied climate disruptions that Arctic peoples and broader societies increasingly face. For example, Canada experienced its worst national wildfire season on record. Multiple communities in the Northwest Territories were evacuated during August, including more than 20,000 people from the capital city of Yellowknife. In August 2023 near Juneau, Alaska, a glacial lake on a tributary of the Mendenhall Glacier burst through its ice dam and caused unprecedented flooding and severe property damage on Mendenhall River, a direct result of dramatic glacial thinning over the past 20 years. With increasing seasonal shifts and widespread disturbances influencing the flora, fauna, ecosystems, and peoples of the Arctic, the need for ongoing observation and collaborative research and adaptation has never been higher.
Special Note: This chapter includes a focus on glaciers and ice caps outside of Greenland (section 5h), which alternates yearly with a section on Arctic river discharge, as the scales of regular observation for both of these climate components are better suited for reporting every two years.
b. Atmosphere
—A. H. Butler, S. H. Lee, G. H. Bernhard, V. E. Fioletov, J.-U. Grooß, I. Ialongo, B. Johnsen, K. Lakkala, R. Müller, T. Svendby, and T. J. Ballinger
The Arctic is warming rapidly, not only at the surface but vertically throughout the troposphere (Cohen et al. 2020). Against the background of long-term warming, the atmospheric circulation contributes to the large year-to-year variability in regional temperature and precipitation patterns across the Arctic. The chemical composition in the Arctic stratosphere, which overlies the troposphere, may also have important climate effects (Polvani et al. 2020; Friedel et al. 2022). The Arctic atmosphere in 2023 was marked by a major sudden stratospheric warming (SSW) in February (Sidebar 5.1) and a persistent anticyclonic high-pressure system during summer that corresponded to record surface temperatures over the Arctic (section 5c), higher-than-normal melt days in Greenland (section 5g), and enhanced wildfire activity in the Northwest Territories (see Sidebar 7.1 for details).
One measure of large-scale atmospheric circulation is geopotential height, which is the altitude of a given atmospheric pressure (Fig. 5.1). The geopotential height tends to be higher where the atmosphere is warmer and lower where it is colder. In general, when the polar cap (60°N–90°N) averaged geopotential heights (PCHs) are anomalously positive, the stratospheric polar winds are weaker than normal, and the tropospheric jet stream is shifted equatorward (and vice versa when the PCHs are anomalously negative). Therefore, the PCHs encapsulate both the thermodynamic and dynamic variability of the high-latitude atmosphere and indicate when the polar atmosphere is vertically coupled (i.e., have the same-signed anomalies from the surface to the upper stratosphere). In 2023 (Fig. 5.1), a major SSW is evident as positive anomalies that first appeared in the stratosphere in February and descended to the troposphere (Sidebar 5.1). The other notable feature is a period of persistent, record-high PCH from July to September that extended vertically from the surface to the mid-stratosphere.
Vertical profile of daily Arctic polar cap (60°N–90°N) standardized geopotential height anomalies (hPa) during 2023. Anomalies are shown with respect to a 30-day centered running-mean 1991–2020 climatology and standardized at each pressure level by the standard deviation of each calendar day during 1991–2020 (smoothed with a 30-day running mean). Data are from once-daily 0000 UTC ERA5 reanalysis (Hersbach et al. 2020).
Citation: Bulletin of the American Meteorological Society 105, 8; 10.1175/BAMS-D-24-0101.1
Figure 5.2 illustrates the large year-to-year variability of the Arctic atmospheric circulation, particularly in winter, and places 2023 in the context of the historical record. The 2023 PCH anomalies in the troposphere (500 hPa) and stratosphere (50 hPa) were generally close to 1991–2020 climatological values in winter (January–March), spring (April–June), and autumn (October–December ); however, record positive PCH anomalies in both the troposphere and stratosphere were observed in summer (July–September ). In the troposphere, the record-high summer value is consistent with a significant linear trend in summer towards increasing tropospheric heights and thus warming air temperatures since 1958 (also evident in spring and autumn). In the stratosphere, linear trends since 1958 are negative in all seasons but generally not significant, except in spring (indicative of cooling stratospheric temperatures).
Time series over the 1958–2023 period of polar-cap averaged height anomalies (m) at (a)–(d) 50 hPa and (e)–(h) 500 hPa for the four seasons: (a),(e) winter (JFM), (b),(f) spring (AMJ), (c),(g) summer (JAS), and (d),(h) autumn (OND). The dashed line is the linear least-squares fit, where the trend value ± the standard error of the trend (m decade−1) is shown in the upper left. Geopotential height anomaly data are from monthly-mean ERA5 reanalysis; anomalies are calculated relative to the 1991–2020 climatology. The 2023 values are marked by a star.
Citation: Bulletin of the American Meteorological Society 105, 8; 10.1175/BAMS-D-24-0101.1
1. The arctic troposphere in 2023
Figure 5.3 shows the seasonally averaged 500-hPa geopotential height and wind anomalies across the Arctic in 2023. Winter (Fig. 5.3a) was marked by anomalously positive heights near the North Pacific and central Arctic and anomalously negative heights across northeastern Eurasia and North America. This pattern was associated both with La Niña teleconnections and the downward coupling of the stratospheric anomalies following the SSW (Fig. 5.1).
500-hPa geopotential height (m; shading) and 200-hPa wind (m s−1; vectors) anomalies for (a) winter, (b) spring, (c) summer, and (d) autumn. Anomalies are calculated relative to the 1991–2020 climatology. Stippling indicates that the anomaly exceeds ±2 std. dev. of the 1991–2020 mean. The dashed circle indicates the 60°N latitude, and the area within denotes the polar-cap region. (Source: ERA5 reanalysis.)
Citation: Bulletin of the American Meteorological Society 105, 8; 10.1175/BAMS-D-24-0101.1
Spring (Fig. 5.3b) was characterized by negative height anomalies over the central Arctic and Alaska, associated with anomalous cold, and positive height anomalies over Canada and Scandinavia, associated with anomalous warmth. However, the seasonal average does not reflect strong monthly variations that occurred. In particular, PCH anomalies at 500 hPa were at their second most positive value since 1958 for April but were moderately negative in May (Fig. 5.1).
Summer (Fig. 5.3c) exhibited strongly anomalous positive heights (anticyclonic wind flow) across a broad region of the Arctic. This is consistent with the observed record-high surface temperatures (section 5c). The persistence and vertical extent (Fig. 5.1) of positive height anomalies likely contributed to higher-than-normal melt days in Greenland (section 5g) and enhanced wildfire activity in the Northwest Territories (see Sidebar 7.1).
A notable feature of autumn (Fig. 5.3d) was the presence of strongly negative height anomalies over the Scandinavian region, linked to cold anomalies there. Height anomalies were otherwise broadly positive, particularly over Canada, where the associated strong anticyclonic wind anomalies likely contributed (via advection) to above-normal temperatures over the Canadian Arctic Archipelago (section 5c).
2. The arctic stratosphere in 2023
In January 2023, the Arctic stratospheric polar vortex was anomalously strong and cold, leading to strong chlorine activation and initiating chemical ozone loss. This was interrupted, however, by a major SSW on 16 February (Sidebar 5.1), which resulted in higher-than-average polar total ozone column (TOC) in March. The stratospheric winds at 10 hPa and 60°N weakly returned to westerlies after the SSW and had a slightly later-than-average spring transition to easterly summer conditions. After the westerly winds returned in autumn, their strength stayed near climatological values until November when they strengthened for a couple of weeks (Fig. 5.1), setting near-records for daily zonal-mean wind speeds at 10 hPa and 60°N.
March has historically been the month with the largest potential for chemical ozone depletion in the Arctic (WMO 2022). In March 2023, the minimum Arctic daily TOC was 3.5% (13 Dobson units; DU) above the average since the start of satellite observations in 1979 (Fig. 5.4a). While the recovery of Arctic TOC to pre-1979 levels is expected due to the phase-out of ozone-depleting substances by the Montreal Protocol, it is difficult to detect due to large year-to-year variability (WMO 2022). Spatially, Arctic TOC anomalies varied between −8% and +24% but stayed within 2 std. dev. of past observations from the Ozone Monitoring Instrument (OMI; 2005–22), with the exception of a small area in northern Scandinavia and the adjacent Barents Sea (Fig. 5.4b). This enhancement of TOC was related to the February 2023 SSW, which transported ozone into the polar stratosphere and raised stratospheric temperatures enough to halt chemical processing and ozone loss.
(a) Time series of the minimum daily-mean total ozone column (TOC; Dobson units, DU) for March poleward of 63°N equivalent latitude, which represents the area enclosed by the stratospheric polar vortex (Butchart and Remsberg 1986) and is determined using ERA5 reanalysis data (adapted from Müller et al. [2008] and WMO [2022]). The blue line indicates the average TOC for 1979–2023. Open circles represent years in which the polar vortex was not well-defined in March. Ozone data for 1979–2019 are based on the combined NIWA-BS total column ozone database version 3.5.1 (Bodeker and Kremser 2021). Ozone data for 2020–23 are from the Ozone Monitoring Instrument (OMI). Monthly mean anomaly maps of (b) total ozone column (%) and (c) noontime ultraviolet index (UVI; %) for Mar 2023 relative to 2005–22 means, based on the OMTO3 Version 3 total ozone product (Bhartia and Wellemeyer 2002), which is derived from OMI measurements. (c) compares UVI anomalies from OMI (first value in parenthesis) with ground-based measurements at nine locations (second value presented). Site acronyms of ground stations are ALT: Alert; EUR: Eureka; NYA: Ny-Ålesund; RES: Resolute; AND: Andøya; SOD: Sodankylä; TRO: Trondheim; FIN: Finse; and OST: Østerås. White areas centered at the North Pole indicate latitudes where OMI data are not available because of polar darkness. Stippling in (b) and (c) indicates pixels where anomalies exceed ±2 std. dev. of the 2005–22 OMI measurement climatology.
Citation: Bulletin of the American Meteorological Society 105, 8; 10.1175/BAMS-D-24-0101.1
Anomalies in monthly averages of the noontime ultraviolet (UV) Index (a measure of the intensity of solar ultraviolet radiation in terms of causing erythema [sunburn] in human skin) for March 2023 varied spatially between −55% and +67% and generally did not exceed 2 std. dev. of past OMI (2005–22) observations (Fig. 5.4c). Areas with high UV index values roughly match areas with low TOCs and vice versa, but UV index anomalies have larger spatial variability because of their added dependence on clouds (Bernhard et al. 2023). Anomalies calculated from satellite data (OMI instrument) and ground-based measurements generally agree well (Fig. 5.4c). Differences in excess of 5% can be explained by coastal (Andøya: OMI anomaly −6%; ground-based anomaly 0%) or urban (Trondheim: OMI anomaly −6%; ground-based anomaly +2%) effects.
The February 2023 major sudden stratospheric warming
S. H. Lee, G. Manney, and A. H. Butler
A major sudden stratospheric warming (SSW) occurred in the Arctic on 16 February 2023. Major SSWs, which occur in the Arctic on average six times per decade, are characterized by a rapid warming of the Arctic stratosphere by as much as 50°C in less than a week and a breakdown and reversal of the mean westerly circulation of the stratospheric polar vortex. Sudden stratospheric warming events induce long-lasting impacts on stratospheric chemical composition (notably ozone; section 5b) and can increase the likelihood of midlatitude cold-air outbreaks for several weeks to two months afterward, acting as a source of subseasonal-to-seasonal predictability for midlatitude surface weather (Domeisen et al. 2020).
CAUSE AND EVOLUTION OF THE EVENT
The SSW in February 2023 was the fourth major SSW in six consecutive winters, part of a recent clustering of events following no major SSWs during the preceding four winters from 2013/14 to 2016/17. The major 2023 SSW was preceded by a minor warming during the last few days of January that was driven by a pulse of enhanced upward-propagating planetary wave activity (Fig. SB5.1a, shading) that weakened the zonal-mean zonal winds in the mid-stratosphere to ∼10 m s−1 (Fig. SB5.1a, contours). Around 14 February, another pulse of anomalous wave activity confined mostly within the stratosphere fully disrupted the vortex, and the winds at 10 hPa and 60°N reversed from westerly to easterly on 16 February, marking the date of the major SSW. During an SSW, the polar vortex either splits into two or more smaller vortices or is displaced away from the Arctic. The February 2023 SSW fell into the latter category, with the vortex in the stratosphere displaced toward Eurasia.
(a) Vertical profile of daily 40°N–80°N eddy heat flux anomalies (std. dev.; shading) and 60°N zonal-mean zonal winds (m s−1; gray contours, with the zero-wind line in black) for 30 days before to 30 days after the 16 February 2023 sudden stratospheric warming (SSW). (b) Average 2-m temperature anomalies (°C, shading) and mean sea-level pressure anomalies (hPa, contours) for the 15 days prior to the SSW (1–15 February) and (c) during a period of stratosphere–troposphere coupling following the SSW (1–15 March). Data are from the ERA5 reanalysis (Hersbach et al. 2023a,b) and all anomalies are shown with respect to a 30-day centered smoothed 1991–2020 climatology.
Citation: Bulletin of the American Meteorological Society 105, 8; 10.1175/BAMS-D-24-0101.1
Because the SSW was not preceded by sustained anomalous tropospheric wave activity, the circulation anomalies prior to the event (Fig. SB5.1b) do not strongly resemble precursors of many SSWs. Nonetheless, pressure near the Aleutian Islands was slightly lower than normal during this time, while an anomalous anticyclone extended across parts of northwest Europe. Both of these features have been shown to contribute to SSWs by constructively interfering with the mean stationary wave pattern in the troposphere (Martius et al. 2009; Garfinkel et al. 2010).
At 10 hPa, the winds then returned to westerly during 22–23 February, reversed back to easterly on 24 February, became westerly again on 26 February, and then easterly once again through 10 March. Although several zonal wind reversals occurred, these all formed part of a single SSW event. Such fluctuations occasionally occur during SSWs, but are not typical. The multiple zonal wind reversals resulted from continued wave activity (Fig. SB5.1a) that eventually destroyed the vortex in the lower to mid-stratosphere sufficiently (Karpechko et al. 2017) for likely downward impacts on the troposphere in early March.
INFLUENCE ON WEATHER PATTERNS AND THEIR PREDICTABILITY
Following the February 2023 SSW, there was no immediate coupling between the stratosphere and the troposphere; in fact, for the first two weeks after the SSW, geopotential heights over the Arctic in the troposphere (below ∼6 km) were anomalously low, in direct contrast to those in the stratosphere. However, during the first half of March, a brief period of stratosphere–troposphere coupling occurred, characterized by a negative North Atlantic Oscillation pattern at the surface (Fig. SB5.1c) as is typical following SSWs. The coupling occurred around 28 February together with the downward propagation of the weakened vortex into the lower stratosphere. This is consistent with the role of lower-stratospheric circulation anomalies in modulating the surface response to SSWs (e.g., Afargan-Gerstman et al. 2022). During this period of stratosphere–troposphere coupling, anomalously high surface temperatures were present around the Labrador Sea and Baffin Bay, with marginally below-normal temperatures across northwest Europe and northern Eurasia. This pattern of temperature anomalies is consistent with the average surface response to SSWs, albeit weaker and more transient. Unusually low temperatures also occurred after the SSW in western North America; however, this is more likely related to North Pacific ridging arising from the then-ongoing La Niña conditions, rather than the SSW itself. The lack of prolonged downward coupling, combined with onset of spring, meant that surface impacts from the February 2023 SSW were relatively minimal.
TRANSPORT OF WATER VAPOR FROM 2022 HUNGA-TONGA HA’APAI ERUPTION
The January 2022 eruption of the underwater Hunga Tonga–Hunga Haʻapai (HTHH) volcano increased the mass of water vapor in the stratosphere by about 10% (e.g., Millán et al. 2022). Water vapor injected in the southern tropics spread across the globe, with high anomalies extending above 60-km altitude in the tropics and midlatitudes and concentrated in the middle stratosphere (around 25 km–35 km) in the polar regions (see section 2g7 for details). The influence on radiative forcing of surface climate from the HTHH stratospheric water vapor increase is uncertain (including whether it produced net heating or cooling), but the impact is minor compared to that of climate change (e.g., Schoeberl et al. 2023).
High water vapor concentrations from the HTHH eruption reached the Arctic stratospheric polar vortex edge in early January 2023. By that time, the vortex was well-developed, and the excess water vapor was largely blocked from crossing its edge (Fig. SB5.2e). Water vapor concentrations are typically high inside the vortex and low outside the vortex (Figs. SB5.2c,d show 2020, a year with a strong vortex). Prior to the SSW (Fig. SB5.2a), exceptionally high water vapor concentrations outside the vortex were well separated across the vortex edge from even higher water vapor concentrations inside (but the high water vapor concentrations inside the vortex were not as anomalous; Fig. SB5.2e).
(a)–(d) Maps of water vapor concentration (mixing ratio in parts per million by volume [ppmv]) in the Northern Hemisphere mid-stratosphere near 27-km altitude (approx. 18 hPa) on the same two days of year in (a),(b) 2023 and (c),(d) 2020, from a gridded product based on Aura Microwave Limb Sounder (MLS) data (Global Modeling and Assimilation Office 2022; Wargan et al. 2023). (e) Time series of anomalies (departure from the daily mean for 2005–21) of MLS water vapor at the same altitude as the maps (Lambert et al. 2021). The purple vertical line is the initial date of the sudden stratospheric warming. In all panels, the black overlaid lines demarcate the stratospheric polar vortex edge, based on MERRA2 reanalysis (Global Modeling and Assimilation Office 2015).
Citation: Bulletin of the American Meteorological Society 105, 8; 10.1175/BAMS-D-24-0101.1