SAP 1.2 DRAFT 3 PUBLIC COMMENT 1 CCSP Synthesis and Assessment Product 1.2 2 Past Climate Variability and Change in the Arctic and at High Latitudes 3 4 Chapter 1 — Executive Summary 5 6 Chapter Lead Authors 7 Richard B. Alley, Pennsylvania State University, University Park, PA 8 Julie Brigham-Grette, University of Massachusetts, Amherst , MA 9 Gifford Miller, University of Colorado, Boulder, CO 10 Leonid Polyak, Ohio State University, Columbus, OH 11 James White, University of Colorado, Boulder, CO 12 13 Chapter 1 Executive Summary 1 SAP 1.2 DRAFT 3 PUBLIC COMMENT 13 1.1 Introduction 14 15 Paleoclimate records play a key role in our understanding of Earth’s past and present 16 climate system and in our confidence in predicting future climate changes. Paleoclimate data 17 help to elucidate past and present active mechanisms of climate change by placing the short 18 instrumental record into a longer term context and by permitting models to be tested beyond the 19 limited time that instrumental measurements have been available. 20 Recent observations in the Arctic have identified large ongoing changes and important 21 climate feedback mechanisms that multiply the effects of global-scale climate changes. Ice is 22 especially important in these “Arctic amplification” processes, which also involve the ocean, the 23 atmosphere, and the land surface (vegetation, soils, and water). As discussed in this report, 24 paleoclimate data show that land and sea ice have grown with cooling temperatures and have 25 shrunk with warming ones, amplifying temperature changes while causing and responding to 26 ecosystem shifts and sea-level changes. 27 28 1.2 Major Questions and Related Findings 29 30 How have temperature and precipitation changed in the Arctic in the past? What does this tell 31 us about Arctic climate that can inform projections of future changes? 32 The Arctic has undergone dramatic changes in temperature and precipitation during the 33 past 65 million years (m.y.) (the Cenozoic Era) of Earth history. Arctic temperature changes 34 during this time exceeded global average temperature changes during both warm times and cold 35 times, supporting the concept of Arctic amplification. Chapter 1 Executive Summary 2 SAP 1.2 DRAFT 3 PUBLIC COMMENT 36 At the beginning of the Cenozoic Era, 65 million years ago (Ma), there was no sea ice on 37 the Arctic Ocean, and neither Greenland nor Antarctica supported an ice sheet. General cooling 38 since that time is attributed mainly to a slow decrease in greenhouse gases, especially carbon 39 dioxide, in the atmosphere. Ice developed during this slow, “bumpy” cooling, first as mountain 40 glaciers and as seasonal sea ice in the Arctic Ocean. Following a global warm period about 3.5 41 Ma in the middle Pliocene, when extensive deciduous forests grew in Arctic regions now 42 occupied by tundra, further cooling crossed a threshold about 2.6 Ma, allowing extensive ice to 43 develop on Arctic land areas and thus initiating the Quaternary ice ages. This ice has responded 44 to persistent features of Earth’s orbit over tens of thousands of years, growing when sunshine 45 shifted away from the Northern Hemisphere and melting when northern sunshine returned.. 46 These changes were amplified by feedbacks such as greenhouse-gas concentrations that rose and 47 fell as the ice shrank and grew, and by the greater reflection of sunshine caused by more- 48 extensive ice. Human civilization developed during the most recent of the relatively warm 49 interglacials, the Holocene (about 11.5 thousand years ago (ka) to the present). The penultimate 50 warm interval, about130–120 ka, received somewhat more Northern-Hemisphere summer 51 sunshine than the Holocene owing to differences in Earth’s orbital configuration. Because this 52 more abundant summer sunshine warmed the Arctic about 5°C above recent temperatures, the 53 Greenland ice sheet was substantially smaller than its current size and almost all glaciers melted 54 completely at that time. 55 The last glacial maximum peaked at about 20 ka when the Arctic was about 20°C colder 56 than at present. Ice recession was well underway by 16 ka, and most of the Northern Hemisphere 57 ice sheets melted by 7 ka. Summer sunshine rose steadily from 20 ka to a maximum (10% higher 58 than at present due to the Earth’s orbit) about 11 ka ago, and has been decreasing since then. The Chapter 1 Executive Summary 3 SAP 1.2 DRAFT 3 PUBLIC COMMENT 59 extra energy received in summer in the early Holocene resulted in warmer summers throughout 60 the Arctic. Temperatures were 1°–3°C above 20th century averages, enough to completely melt 61 many small glaciers in the Arctic and to slightly shrink the ice sheet on Greenland. Summer sea- 62 ice limits were significantly less than their 20th century average. As summer sunshine decreased 63 in the second half of the Holocene, glaciers re-established or advanced, and sea ice became more 64 extensive. Late Holocene cooling reached its nadir during the Little Ice Age (about 1250–1850 65 AD), when most Arctic glaciers reached their maximum Holocene extent. Warming during the 66 19th century has resulted in Arctic-wide glacier recession, the northward advance of terrestrial 67 ecosystems, and the reduction of perennial (year-round) sea ice in the Arctic Ocean. These trends 68 will continue if greenhouse gas concentrations continue to increase into the future. 69 Paleoclimate reconstructions of Arctic temperatures compared with global temperature 70 changes during four key intervals during the past 4 m.y. allow a quantitative estimate of Arctic 71 amplification. These data suggest that Arctic temperature change is 3 to 4 times the global 72 average temperature change during both cold and warm departures. 73 74 How rapidly have temperature and precipitation changed in the Arctic in the past? What do 75 these past rates of change tell us about Arctic climate that can inform projections of future 76 changes? 77 As discussed with the previous question, climate changes on numerous time scales for various 78 reasons, and it has always done so. In general, longer-lived changes are somewhat larger but 79 much slower than shorter- lived changes. 80 Chapter 1 Executive Summary 4 SAP 1.2 DRAFT 3 PUBLIC COMMENT 81 Processes linked to continental drift (plate tectonics) have affected atmospheric and oceanic 82 currents and the composition of the atmosphere over tens of millions of years; in the Arctic, a 83 global cooling trend has switched conditions from being ice-free year-round near sea level to icy 84 conditions more recently. Within the icy times, variations in Arctic sunshine in response to 85 features of Earth’s orbit have caused regular cycles of warming and cooling over tens of 86 thousands of years that were roughly half the size of the continental-drift-linked changes. This 87 “glacial-interglacial” cycling was amplified by colder times bringing reduced greenhouse gases 88 and greater reflection of sunlight, especially from expanded ice-covered regions. This glacial- 89 interglacial cycling has been punctuated by sharp-onset, sharp-end (in as little as 1–10 years) 90 millennial oscillations, which near the North Atlantic were roughly half as large as the glacial- 91 interglacial cycling but which were much smaller Arctic-wide and beyond. The current warm 92 period of the glacial-interglacial cycling has been influenced by cooling events from single 93 volcanic eruptions, slower but longer lasting changes from random fluctuations in frequency of 94 volcanic eruptions and from weak solar variability, and perhaps by other classes of events. Very 95 recently, human effects have become evident, not yet showing both size and duration that exceed 96 peak values of natural fluctuations further in the past, but with projections indicating that human 97 influences could become anomalous in size and duration and, hence, in speed. 98 99 What does the paleoclimate record tell us about the past size of the Greenland ice sheet and its 100 implications for sea level changes? 101 The paleo-record shows that the Greenland ice sheet has consistently lost mass when the 102 climate warmed and grown when the climate cooled, even at times of negligible sea-level 103 change. In contrast, no changes in the ice sheet have been documented independent of Chapter 1 Executive Summary 5 SAP 1.2 DRAFT 3 PUBLIC COMMENT 104 temperature changes. Moreover, snowfall has increased when the climate warmed, but the ice 105 sheet lost mass nonetheless; increased accumulation in the ice sheet center was not sufficient to 106 counteract increased melt and flow near the edges. Most of the documented changes (of both ice 107 sheet and forcings) spanned multi-millennial periods, but limited data show rapid responses to 108 rapid forcings have also occurred. In particular, regions near the ice margin have been observed 109 to respond within a few decades. However, major changes of the ice sheet are thought to take 110 centuries to millennia, and this is supported by the limited data. 111 The paleo-record does not yet give any strong constraints on how rapidly a near-complete loss of 112 the ice sheet could occur, although the paleo-data indicate that onset of shrinkage will be 113 essentially immediate after forcings begin. The available evidence suggests such a loss requires 114 a sustained warming of at least 2-7oC above mean 20th century values, but this threshold is 115 poorly defined. The paleo-archives are sufficiently sketchy that temporary ice sheet growth in 116 response to warming, or changes induced by factors other than temperature, could have occurred 117 without being recorded. 118 119 What does the paleoclimate record tell us about past changes in Arctic sea ice cover, and what 120 implications does this have for consideration of recent and potential future changes? 121 Although incomplete, existing data outline the development of Arctic sea-ice cover from 122 the ice-free conditions of the early Cenozoic. Some data indicate that sea ice has consistently 123 covered at least part of the Arctic Ocean for the last 13–14 million years, and it has been most 124 extensive during the most recent approximately 2 m.y. Other data argue against the development 125 of perennial (year-round) sea ice until the most recent few million years. Nevertheless, episodes 126 of considerably reduced ice cover, or even a seasonally ice-free Arctic Ocean, probably Chapter 1 Executive Summary 6 SAP 1.2 DRAFT 3 PUBLIC COMMENT 127 punctuated even this latter period. Warmer climates associated with the orbitally paced 128 interglacials promoted these episodes of diminished ice. The current sea ice reduction in the 129 Arctic began during the late 19th century and has accelerated during the last several decades. It is 130 the largest ice reduction during at least the last few thousand years, and it is progressing at a very 131 fast rate that appears to have no analogs in the past. 132 133 1.3 Recommendations 134 135 Paleoclimatic data on the Arctic are generated by numerous international investigators 136 who study a great range of archives throughout the vast reaches of the Arctic. The value of this 137 diversity is evident in this report. Many of the key results of this report rest especially on the 138 outcomes of community-based syntheses, including the CAPE Project, and multiply replicated, 139 heavily sampled archives such as the central Greenland deep ice cores. Results from the ACEX 140 deep coring in Arctic Ocean sediments were appearing as this report was being written. These 141 results are quite valuable and will become more so with synthesis and replication, including 142 comparison with land-based and marine records. The number of questions answered, and raised, 143 by this one new data set shows how sparse the data are on many aspects of Arctic paleoclimatic 144 change. We recommend that future research maintain and expand the diversity of 145 investigators, techniques, archives, and geographic locations, while promoting development of 146 community-based syntheses and multiply replicated, heavily sampled archives. Only through 147 breadth and depth can the remaining uncertainties be reduced while confidence in the results 148 is improved. 149 Chapter 1 Executive Summary 7 SAP 1.2 DRAFT 3 PUBLIC COMMENT 150 The questions asked of this study by the CCSP are relevant to public policy and require 151 answers. The answers provided here are, we hope, useful and informative. However, we 152 recognize that despite the contributions of many community members to this report, in many 153 cases a basis was not available in the refereed scientific literature to provide answers with the 154 accuracy and precision desired by policymakers. We recommend that members of the Arctic 155 paleoclimatic community formulate future research activities to address in greater detail the 156 policy-relevant questions motivating this report. 157 158 Paleoclimatic data provide very clear evidence of past changes in important aspects of the 159 Arctic climate system. The ice of the Greenland ice sheet, smaller glaciers and ice caps, the 160 Arctic Ocean, and in soils is shown to be vulnerable to warming, and Arctic ecosystems are 161 strongly affected by changing ice and climate. National and international studies generally 162 project rapid warming in the future. If this warming occurs, the paleoclimatic data indicate that 163 ice will melt and associated impacts will follow, with implications for ecosystems and 164 economies. We recommend that policymakers and science managers use the results presented 165 here in design of monitoring, process, and model-projection studies of Arctic change and 166 linked global responses. Chapter 1 Executive Summary 8 CCSP Synthesis and Assessment Product 1.2 Past Climate Variability and Change in the Arctic and at High Latitudes Chapter 2 — INTENTIONALLY BLANK This is a place holder in the event that an additional introductory chapter is deemed desirable based on the public comments. SAP 1.2 DRAFT 3 PUBLIC COMMENT 1 CCSP Synthesis and Assessment Product 1.2 2 Past Climate Variability and Change in the Arctic and at High Latitudes 3 4 Chapter 3 — Preface: Why and How to Use This Synthesis and Assessment 5 Report 6 7 Chapter Lead Author 8 Joan Fitzpatrick, U.S. Geological Survey 9 Contributing Authors 10 11 Richard B. Alley, Pennsylvania State University, University Park, PA 12 Julie Brigham-Grette, University of Massachusetts, Amherst , MA 13 Gifford Miller, University of Colorado, Boulder, CO 14 Leonid Polyak, Ohio State University, Columbus, OH 15 Mark Serreze, University of Colorado, Boulder, CO 16 Chapter 2 Preface 1