Cryospheric Changes and Their Impacts on Water Resources in the Belt and Road Regions

“Belt and Road” regions include Asia, Europe and eastern and northern Africa, with a wide spatial distribution. The cryosphere is undergoing rapid changes in the Belt and Road regions with global warming, and has an important impact on water resources, ecosystems and Arctic waterways in these regions. This article reviewed recent cryospheric changes and associated impacts on water resources in the Belt and Road regions during the last decades. The main cognitions are as follows: \mathrm{①} Most glaciers are shrinking and glacier mass balances are most negative, but there are regional differences in the changes of glaciers. \mathrm{②} Global temperature rise has resulted in permafrost degradation, including a rise in permafrost temperature and decreasing permafrost thickness as well as an increase in active layer thickness. \mathrm{③} There is a significant decrease in snow cover extent and an increase in snow depth. Snow cover duration has shortened, the onset of snow cover has delayed, and the end of snow cover has advanced. However, there are still obvious regional differences in the changes of snow cover. \mathrm{④} Arctic sea ice has declined precipitously in both extent and thickness in summer, and multi-year sea ice has decreased,indicating the precipitous retreat of sea ice. The freeze-up date of some lakes has been delayed, the break-up date has advanced, and the ice cover duration of river/lake ice has significantly shortened. \mathrm{⑤} Glacial runoff has increased significantly in China. Snowmelt and permafrost degradation have also increased the basin runoff, which indicates the important impact of cryospheric changes on runoff. This study will provide a baseline and important scientific support for addressing climate change and regional sustainable development.     

The significance of global snow and ice cover for global change studies

The spatial extent and volume of global snow and ice cover at present and during glacial conditions are reviewed for each of the principal components (snow cover, land ice, sea ice and permafrost). The state of global monitoring of snow and ice conditions is shown to be adequate for some variables, but unsatisfactory for others. Recent trends are reported and expected changes projected for increasing concentrations of greenhouse gases are examined.     

极地海冰的研究及其在气候变化中的作用

极地海冰作为全球气候系统的一个重要组成部分,通过影响大洋表面的辐射平衡、物质平衡、能量平衡以及大洋温、盐流的形成和循环而影响全球气候变化.从最初研究极地海冰的强度和承载力到目前海/冰/气相互作用全球气候耦合模型的建立,使海冰变化和全球气候变化紧密结合起来.这些研究领域主要有:海冰及其表层雪的物理特性和过程、海冰区域生态特征、海冰区与气候相关的反照率和物质平衡研究以及海冰气候耦合模型等大的领域.模拟显示,21世纪因为全球变暖,南北极海冰都将减少.海冰和全球气候系统其它要素之间的相互作用问题、极地海冰的厚度季节性区域性分布问题、极地海冰边界及范围变化趋势问题、生消关键过程及其影响因素问题、冰间湖的作用以及海气相互作用等将是未来重要的研究方向.     

最近十多年来冰冻圈加速萎缩——IPCC第六次评估报告之冰冻圈变化解读

政府间气候变化专门委员会（IPCC）于2021年8月发布了第六次评估报告第一工作组报告《气候变化2021：自然科学基础》。该报告基于最新的观测和模拟研究,评估了冰冻圈变化的现状,并采用CMIP6模式对未来变化进行了预估。报告明确指出,近十多年来冰冻圈呈现加速萎缩状态：北极海冰面积显著减小、厚度减薄、冰量迅速减少;格陵兰冰盖、南极冰盖和全球山地冰川物质亏损加剧;多年冻土温度升高、活动层增厚,海底多年冻土范围减少;北半球积雪范围也在明显变小,但积雪量有较大空间差异。冰冻圈的快速萎缩加速海平面的上升。未来人类活动对冰冻圈萎缩的影响将愈加显著,从而导致北极海冰面积继续减少乃至消失,冰盖和冰川物质将持续亏损,多年冻土和积雪的范围继续缩减。报告也提出,目前冰冻圈研究仍存在观测资料稀缺、模型对各影响因素的敏感性参数和过程描述亟需提升、对吸光性杂质的变化机制认知不足等问题,从而影响了对冰冻圈变化预估的准确性,未来需要重点关注。     

The rapidly shrinking cryopshere in the past decade：an interpretation of cryospheric changes from IPCC WGI Sixth Assessment Report北大核心CSCD

The Working Group I report of the Sixth Assessment Report（AR6）of the Intergovernmental Panel on Climate Change（IPCC）was released in August 2021. Base on updated and expanding data, AR6 presented the improved assessment of past changes and processes of cryosphere. AR6 also predicted the future changes us⁃ ing the models in CMIP6. The components of cryosphere were rapid shrinking under climate warming in the last decade. There were decreasing trends in Arctic sea-ice area and thickness. Sea-ice loss was significant. The Greenland Ice Sheet, the Antarctic Ice Sheet and all glaciers lost more mass than in any other decade. Global warming over the last decades had led to widespread permafrost warming, active layer thickness increasing and subsea permafrost extent reducing. Snow cover extent in the Northern Hemisphere also decreased significantly. However, the variations of snow depth and snow water equivalent showed great spatial heterogeneity. The rapid shrinking of the cryosphere accelerated the global mean sea level rise. The impact of human activities on cryo⁃ sphere will become more significant in the future. The Arctic sea-ice area will decrease, and the Arctic Ocean will likely become practically sea ice-free. The Greenland Ice Sheet, the Antarctic Ice Sheet and glaciers will continue to lose mass throughout this century. Permafrost and Northern Hemisphere snow cover extent will con⁃ tinue to decrease as global climate continues to warm. In addition, there are still uncertainties in the prediction of cryosphere due to the absence of observations, the poor sensitivity of models to the components and processes of cryosphere, and the inexplicit represent of the mechanism of light-absorbing impurities. More attentions should be paid on these issues in the future. © 2022 Science Press (China). All rights reserved.     

The Last Permafrost Maximum (LPM) map of the Northern Hemisphere: permafrost extent and mean annual air temperatures, 25–17 ka BP

This paper accompanies a map that shows the extent of permafrost in the Northern Hemisphere between 25 and 17 thousand years ago. The map is based upon existing archival data, common throughout the Northern Hemisphere, that include ice‐wedge pseudomorphs, sand wedges and large cryoturbations. Where possible, a distinction is made between areas with continuous permafrost and areas where permafrost is either spatially discontinuous or sporadic. The associated mean annual palaeo‐temperatures that are inferred on the basis of present‐day analogues increase understanding of the possible changes in permafrost extent that might accompany current global warming trends. Areas with relict permafrost and areas that were formerly exposed due to lower sea level (submarine permafrost) are also mapped. Mapping is mostly limited to lowland regions (areas approximately <1000 m a.s.l.). Striking features that appear from the map are (i) the narrow permafrost zone in North America, which contrasts with the broader LPM permafrost zone in Eurasia (that may be related to different snow thickness or vegetation cover), (ii) the zonal extent of former LPM permafrost (that may reflect sea‐ice distribution), which contrasts with the present‐day pattern of permafrost extent (especially in Eurasia) and (iii) the relatively narrow zones of LPM discontinuous permafrost (that may indicate strong temperature gradients).     

Latest Pleistocene and Holocene glacier fluctuations in western Canada

We summarize evidence of the latest Pleistocene and Holocene glacier fluctuations in the Canadian Cordillera. Our review focuses primarily on studies completed after 1988, when the first comprehensive review of such evidence was published. The Cordilleran ice sheet reached its maximum extent about 16 ka and then rapidly decayed. Some lobes of the ice sheet, valley glaciers, and cirque glaciers advanced one or more times between 15 and 11 ka. By 11 ka, or soon thereafter, glacier cover in the Cordillera was no more extensive than at the end of the 20th century. Glaciers were least extensive between 11 and 7 ka. A general expansion of glaciers began as early as 8.4 ka when glaciers overrode forests in the southern Coast Mountains; it culminated with the climactic advances of the Little Ice Age. Holocene glacier expansion was not continuous, but rather was punctuated by advances and retreats on a variety of timescales. Radiocarbon ages of wood collected from glacier forefields reveal six major periods of glacier advance: 8.59–8.18, 7.36–6.45, 4.40–3.97, 3.54–2.77, 1.71–1.30 ka, and the past millennium. Tree-ring and lichenometric dating shows that glaciers began their Little Ice Age advances as early as the 11th century and reached their maximum Holocene positions during the early 18th or mid-19th century. Our data confirm a previously suggested pattern of episodic but successively greater Holocene glacier expansion from the early Holocene to the climactic advances of the Little Ice Age, presumably driven by decreasing summer insolation throughout the Holocene. Proxy climate records indicate that glaciers advanced during the Little Ice Age in response to cold conditions that coincided with times of sunspot minima. Priority research required to further advance our understanding of late Pleistocene and Holocene glaciation in western Canada includes constraining the age of late Pleistocene moraines in northern British Columbia and Yukon Territory, expanding the use of cosmogenic surface exposure dating techniques, using multi-proxy paleoclimate approaches, and directing more of the research effort to the northern Canadian Cordillera.     

Mapping exposure to snow in a developing African context: implications for human and livestock vulnerability in Lesotho

Alpine cryospheric hazards are becoming increasingly prominent under current global/regional climate change scenarios and receiving wide scientific coverage from, in particular, northern hemisphere mountain regions associated with glaciers, permafrost, and extensive seasonal snow cover. However, there is a general paucity of knowledge and attention on cryospheric hazards associated with mountain environments only occasionally/rarely impacted by heavy seasonal snowfalls or severe frost events, particularly those in developing and southern hemisphere regions. Prolonged snow cover in the Lesotho Highlands sometimes carries the consequence of human and livestock deaths owing to isolation and exposure in this developing region. We use daily Moderate Resolution Imaging Spectroradiometer snow cover images for the period 2003–2010, to establish the frequency, extent, and timing of snowfalls across Lesotho. In addition, a digital shape file containing the location, name, and district attributes of 2,016 villages across Lesotho was used to assist in the construction of a village exposure to snow index. A ranking system was applied to each village according to the seasonal duration of snow cover, and the accessibility and proximity to the nearest road. Snowfalls occur on average between 1 and 8 times per annum, with village exposure to snow (potential vulnerability) being generally low, particularly for the lowlands and Senqu River Valley. However, the study identifies that some high-altitude (>2,500 m) villages such as Thoteng, Letseng-la-Terae, and Mabalane are, on occasion, highly exposed to prolonged snow cover, and particularly so during the mid-snow season of July/August. We demonstrate the importance of applying spatiotemporal assessments on infrequent snow occurrences (which carry associated hazards) in developing mountain regions such as Lesotho, with implications to reduce livelihood risks through improved disaster preparedness and a well-informed, focused emergency response.     

The post to late glacial valley reconstruction on the Haramosh north side (Mani, Baska and Phuparash valleys)

The post to late glacial valley reconstruction is focused on the Mani- Baska and Phuparash valleys on the Rakaposhi- Haramosh Muztagh in the south Karakoram. The recently glaciated valleys join the Indus valley near Sassi at 1500 m. The knowledge of the tributary valley reconstruction can be seen in the context of the scientific discussion about the extent of glaciation along the main Indus valley. Today, the recent avalanche fed glaciers come down from high lying catchment areas with an average altitude of 6700–6800 m and terminate at 2700 m. Snow line runs at 4700–4800 m in the steep flanks which is common in the Karakoram Mountains. The postglacial extent is marked by the great lateral moraine (GLM) and reached down not more than 2.5–5 km away from the recent glaciers with a calculated snow line depression of 300 m in maximum. It can be shown that the valleys were already glaciated during the lastest Late Glacial down to the valley outlet at 1500 m. The snow line was depressed 600–700 m during that period. A high glacial ice filling of the Haramosh valley and glacial erosion of the flat top of the Darchan ridge as an intermediate valley head is strongly probable.     

面向对象-改进冰雪指数法消除冰湖干扰提取冰川边界的优越性分析——以各拉丹冬冰川为例

冰川作为气候变化的重要指标器,其范围监测对区域生态环境以及人类社会生活具有重要意义。目前基于遥感技术的冰川范围监测应用广泛,然而传统遥感监测方法中冰雪指数阈值法的提取结果存在无法区分冰川与冰湖的现象,面向对象分类法受地物光谱纹理信息限制无法避免同谱异物现象的出现。为弥补上述不足,提出一种可区分冰湖与冰川的改进冰雪指数,并将其融入面向对象分类法中,构建了一种面向对象-改进冰雪指数法。将各拉丹冬冰川作为试验区（该地区冰川表面洁净）,运用面向对象-改进冰雪指数法识别冰川边界,使用青藏高原冰川数据产品及常规遥感监测方法的提取结果作为参考数据,以验证此方法的有效性和稳健性。结果表明：面向对象-改进冰雪指数法综合了改进冰雪指数阈值法以及面向对象分类法的优点,冰川范围提取精度高达97.26%,与冰雪指数阈值提取结果相比精度提高了0.12%,与面向对象分类法提取结果相比精度提高了0.38%。此方法不仅有效地解决了地物错分的问题,还实现了冰川边界的精确识别。     

Nomenclature and resolution in Holocene glacial chronologies

Most Quaternary research in Canada during the first half of the twentieth century focused on Pleistocene glaciation. Given the dramatic shifts in climate during the Pleistocene, it is not surprising that the Holocene was viewed as a time of benign climate. Holocene climate variability was first recognized around the middle of the century when paleoecologists found evidence that the early part of the epoch was warmer and drier than the later part. In 1970s and 1980s, another generation of geologists, geographers, and botanists began to recognize more complexity in Holocene climate and vegetation in western Canada. Several millennial-scale glacier “advances” postdating the early Holocene warm interval were defined, including the Garibaldi Phase (6.9–5.6 ka), the Tiedemann–Peyto Advance (3.5–1.9 ka), and the Little Ice Age (AD 1200–1900). Subsequently, application of dendrochronological techniques and stratigraphic studies in glacier forefields showed that the Little Ice Age was itself more complex than previously thought. During that 700-year period, glaciers repeatedly advanced and retreated in response to climatic variability on time scales ranging from centuries to decades. Recent work shows that the glacier record of the Garibaldi Phase and the Tiedemann and Peyto advances are similar in complexity to the Little Ice Age, with multiple advances of glaciers separated by intervals of more restricted ice cover. Researchers have also identified other times in the Holocene when glaciers expanded from restricted positions – 8.20, 4.90–3.80, and 1.70–1.40 ka. Continued research undoubtedly will reveal additional complexities, but with what is currently known the appropriateness of terms such as “Tiedemann Advance,” “Peyto Advance,” and “Little Ice Age” can be questioned. Only short periods of time separate these episodes as currently defined, and it seems likely that intervals of restricted glacier cover within each of these millennial-length intervals are just as long as the intervals separating them.     

冰川/积雪-大气相互作用研究进展

冰川和积雪是冰冻圈的重要组成部分,在全球或区域气候系统中起着极其重要的作用.开展冰川/积雪-大气相互作用研究,是研究冰冻圈物理过程及其对气候系统反馈作用的必然需求,也是研究冰川/积雪对气候变化响应的有效手段,同时还可为全球(区域)气候和水文模式提供冰川/积雪面的地表特征参数.近一个世纪以来,在冰川/积雪面辐射特征、能量通量计算方法和平衡特征等方面开展了许多观测试验和理论研究,并取得了卓有成效的研究结果.但是在准确获取辐射通量、研发普适性较强的反照率参数化方案、复杂地形条件下能量通量的计算,以及发展分布式能量平衡模式等方面尚存在许多不确定性,仍面临许多技术难点,也是未来的研究重点.     

Dynamics of an active rock glacier (Ötztal Alps, Austria)

The rock glacier Innere Ölgrube, located in a small side valley of the Kauner Valley (Ötztal Alps, Austria), consists of two separate, tongue-shaped rock glaciers lying next to each other. Investigations indicate that both rock glaciers contain a core of massive ice. During winter, the temperature at the base of the snow cover (BTS) is significantly lower at the active rock glacier than on permafrost-free ground adjacent to the rock glacier. Discharge is characterized by strong seasonal and diurnal variations, and is strongly controlled by the local weather conditions. Water temperature of the rock glacier springs remains constantly low, mostly below 1°C during the whole melt season. The morphology of the rock glaciers and the presence of meltwater lakes in their rooting zones as well as the high surface flow velocities of >1 m/yr point to a glacial origin. The northern rock glacier, which is bounded by lateral moraines, evolved from the debris-covered tongue of a small glacier of the Little Ice Age with its last highstand around A.D. 1850. Due to the global warming in the following decades, the upper parts of the steep and debris-free ice glacier melted, whereas the debris-covered glacier tongue transformed into an active rock glacier. Due to this evolution and due to the downslope movement, the northern rock glacier, although still active, at present is cut off from its ice and debris supply. The southern rock glacier has developed approximately during the same period from a debris-covered cirque glacier at the foot of the Wannetspitze massif.     

20世纪全球变暖的冰冻圈证据

20世纪80年代以来，地面气象观测结果表明全球气温在明显升高，而探空资料和卫星遥感资却表明低层大气在降温，这使得人们对于全球变暖问题提出了质疑。通过对近100多年来冰冻圈各组成要素（冰川、积雪、冻土、海冰等）变化的综合分析，并结合全球不同地区的冰芯记录，证明了20世纪全球气候是在变暖，指出开展冰冻圈（尤其是冰盖、积雪、冻土和海冰）与大气／海洋之间相互作用（影响和响应）过程研究的必要性和迫切性。     

The pleistocene glaciation of Tibet and the onset of ice ages — An autocycle hypothesis

During seven expeditions new data were obtained on the maximum extent of glaciation in Tibet and the surrounding mountains. Evidence was found of moraines at altitudes as low as 980 m on the S flank of the Himalayas and 2300 m on the N slope of the Tibetan Plateau, in the Qilian Shan. On the N slopes of the Karakoram, Aghil and Kuen Lun moraines occur as far down as 1900 m. In S Tibet radiographic analyses of erratics document former ice thicknesses of at least 1200 m. Glacial polishing and knobs in the Himalayas, Karakoram etc. are proof of glaciers as thick as 1200–2000 m. On the basis of this evidence, a 1100–1600 m lower equilibrium line altitude (ELA) was reconstructed for the Ice Age, which would mean 2.4 million km² of ice covering almost all of Tibet, since the ELA was far below the average altitude of Tibet. On Mt. Everest and K2 radiation was measured up to 6650 m, yielding values of 1200–1300 W/m². Because of the subtropical latitude and the high altitude solar radiation in Tibet is 4 times greater than the energy intercepted between 60 and 70° N or S. With an area of 2.4 million km² and an albedo of 90% the Tibetan ice sheet caused the same heat loss to the earth as a 9.6 million km² sized ice sheet at 60–70° N. Because of its proximity to the present-day ELA, Tibet must have undergone large-scale glaciation earlier than other areas. Being subject to intensive radiation, the Tibetan ice must have performed an amplifying function during the onset of the Ice Age. At the maximum stage of the last ice age the cooling effect of the newly formed, about 26 million km² sized ice sheets of the higher latitudes was about 3 times that of the Tibetan ice. Nevertheless, without the initial impulse of the Tibetan ice such an extensive glaciation would never have occurred. The end of the Ice Age was triggered by the return to preglacial radiation conditions of the Nordic lowland ice. Whilst the rise of the ELA by several hundred metres can only have reduced the steep marginal outlet glaciers, it diminished the area of the lowland ice considerably.     

南极中山站附近固定冰生消过程观测

2005～2006年对南极中山站附近固定冰区海冰生消过程进行了系统观测。基于观测数据比较了2005年和2006年冰季固定冰生消过程的差异,分析了冰厚增长速率的日内变化和季节变化,讨论了积雪对固定冰生消的影响。2005年尽管冰生长期的负积温较大,但由于雪厚较大,其冰厚增长速率和最大冰厚都小于2006年的观测值;极夜期间冰厚增长速率的日内变化不明显,但极夜前和极夜后都存在明显的日内变化;中山站以北16km范围内的冰/雪厚度随离岸距离没有明显的变化趋势,但冰脊形成和冰山存在都会对局部的冰/雪厚度分布产生较大影响。     

The upper limit of glaciation in the Himalayas

On the slopes of Himalayan Mountains there is a reduction and culmination of glaciation at 7000–7200 m asl. The presumed cause for this is that the surface temperatures on these slopes are too low for glaciation. This working hypothesis was verified with temperature measurements using collected infra-red radiation. The regression analysis of the measurements taken in the Mt. Everest region during sunny weather conditions of the post-monsoon season resulted in a 0°C line at 7000–7200 m asl. The coincidence of the 0°C line with the upper limit of glaciation is causally definable with the copula between the function of temperature and snow metamorphism: since it is too cold above 7000–7200 m asl, metamorphism into perennial or galcial ice through settling or sintering is absent or simply too slow. High relief and drifting hinder here the processes of ice-formation through pressure compaction of the dry-snow accumulation caused by molecular diffusion and recrystallization. Above 7200 m only continuous leeward accumulations of shifting snow on wall sections with moderate inclination lead to the formation of seracs. However, glaciation generally ceases at this level. This additionally confirms another study. It has been proven that Himalayan glaciers with catchment areas over 7000 m do not extend further downward than those glaciers whose catchment areas just reach this altitude. A break in balance at 7100 m asl is thereby confirmed, and the upper glacial limit is proven. Above the glacial region a rocky zo ne adjoins with pergelic conditions even in the surface layer. This zone is covered by snow during monsoon season only. Here, the weathering processes take place in an arid environment without thawing and purely by means of temperature variations below 0°C. They could correspond to those occurring on a larger scale on the planets of our solar system.A lowering of the upper glacial limit by at least 660 or 1200 m respectively, analogous to the Pleistocene snow-line depression reconstructed in S Tibet and the Central Himalayas, is assumed during the Ice Age.The author gratefully acknowledges the translation of this paper rendered by Dr. J. A. Hellen, Newcastle-upon-Tyne.     

Mais comment s'écoule donc un glacier ? Aperçu historique

Ice and snow have often helped physicists understand the world. On the contrary it has taken them a very long time to understand the flow of the glaciers. Naturalists only began to take an interest in glaciers at the beginning of the 19th century during the last phase of glacier advances. When the glacier flow from the upslope direction became obvious, it was then necessary to understand how it flowed. It was only in 1840, the year of the Antarctica ice sheet discovery by Dumont d'Urville, that two books laid the basis for the future field of glaciology: one by Agassiz on the ice age and glaciers, the other one by canon Rendu on glacier theory. During the 19th century, ice flow theories, adopted by most of the leading scientists, were based on melting/refreezing processes. Even though the word ‘fluid’ was first used in 1773 to describe ice, more the 130 years would have to go by before the laws of fluid mechanics were applied to ice. Even now, the parameter of Glen's law, which is used by glaciologists to model ice deformation, can take a very wide range of values, so that no unique ice flow law has yet been defined. To cite this article: F. Rémy, L. Testut, C. R. Geoscience 338 (2006).     

Relationships between glacier and rock glacier in the Maritime Alps, Schiantala Valley, Italy

In the Schiantala Valley of the Maritime Alps, the relationship between a till-like body and a contiguous rock glacier has been analyzed using geomorphologic, geoelectric and ice-petrographic methodologies. DC resistivity tomographies undertaken in the till and in the rock glacier show the presence of buried massive ice and ice-rich sediments, respectively. Ice samples from a massive ice outcrop show spherical gas inclusions and equidimensional ice crystals that are randomly orientated, confirming the typical petrographic characteristics of sedimentary ice. The rock glacier formation began after a phase of glacier expansion about 2550 ± 50 ¹⁴C yr BP. Further ice advance during the Little Ice Age (LIA) overrode the rock glacier root and caused partial shrinkage of the pre-existing permafrost. Finally, during the 19th and 20th centuries, the glacial surface became totally debris covered. Geomorphological and geophysical methods combined with analyses of ice structure and fabric can effectively interpret the genesis of landforms in an environment where glaciers and permafrost interact. Ice petrography proved especially useful for differentiating ice of past glaciers versus ice formed under permafrost conditions. These two mechanisms of ice formation are common in the Maritime Alps where many sites of modern rock glaciers were formerly occupied by LIA glaciers.     

Climate-Volcanism Feedback and the Toba Eruption of 74,000 Years Ago

A general feedback between volcanism and climate at times of transition in the Quaternary climate record is suggested, exemplified by events accompanying the Toba eruption (74,000 yr ago), the largest known late Quaternary explosive volcanic eruption. The Toba paroxysm occurred during the δ¹⁸O stage 5a-4 transition, a period of rapid ice growth and falling global sea level, which may have been a factor in creating stresses that triggered the volcanic event. Toba is estimated to have produced between 10¹⁵ and 10¹⁶ g of fine ash and sulfur gases lofted in co-ignimbrite ash clouds to heights of at least 32 ± 5 km, which may have led to dense stratospheric dust and sulfuric acid aerosol clouds. These conditions could have created a brief, dramatic cooling or "volcanic winter," followed by estimated annual Northern Hemisphere surface-temperature decreases of 3° to 5°C caused by the longer-lived aerosols. Summer temperature decreases of 10°C at high northern latitudes, adjacent to regions already covered by snow and ice, might have increased snow cover and sea-ice extent, accelerating the global cooling already in progress. Evidence for such climate-volcanic feedback, following Milankovitch periodicities, is found at several climatic transitions.