Ice Core Evidence for Climate Change

PALEOCEANOGRAPHY Lecture 7 Ice core evidence for climate change I. Polar ice caps: characteristics and flow behavior y I Surface C Flow lines x C I Bed rock Vertical cross section of an ice sheet resting on a horizontal subsurface. Ice particles deposited on the snow surface will follow lines that travel closer to the base the farther inland the site deposition. An ice mass formed around the divide (I-I) will be plastically deformed (thinned) with depth as suggested by the lined areas. The dashed curve along the vertical ice core (C-C) shows the calculated horizontal velocity profile Vx (Weertman 1968b). The horizontal arrows along C-C show the adopted approximation to Vx (Dansgaard et al. 1969). Note: this figure does not illustrate bedrock depression A. Total ice mass on earth today: 29 x 106 km3, equivalent to ~80m of sea level rise (but some of this would fill in space filled by sub sea-surface ice).. B. Physical Properties and Transitions 1. Depth transitions a. Snow-->firn-->ice Catm, global Catm, local Density (kg/m3) Age (yr) 200 0,5 400 1 550 20 800 100 840 120 Large-scale transport atmospheric circulation Wet deposition turbulence air-snow partitioning Csnow Redistribution Cfirn Smoothing and decay Cice Decay Processes and steps involved in transfer function, which relates concentrations in ice to those in the global atmosphere. Depth and age scales are for Greenland. Snow-to-firn transition is defined by metamorphism and grain growth; firn-to-ice transition is defined by pore closure. b. Bubble closure and compression: appearance of air hydrate inclusions 1000-1300m; bubbles disappear completely 1500-1600m; c. Brittle ice zone (800-1200m); ice often fractures upon return to surface d. deep stable ice 2. Bubbles reform upon return of deep ice to surface 3. Exclusion of soluble salts from ice crystals and grain-boundary H2SO4 II. Chronology A. Annual counting 1.In the upper part of the ice, annual variations in O and H isotopes can be used to count annual layers. As the ice gets older, molecular diffusion blurs the cycles and they become ambiguous, hence limiting O18 cycle counting to the upper portion of the core (~1000 years or so, depending on accumulation rate). At low accumulation rates (e.g. South Pole), annual cycles are not at all useful; at higher accumulation rates (e.g. Dye 3), annual δ18O cycles can be discerned back as far as 3,000 years. 2. Other indicators can show seasonal cycles: a. dust b. chemical constituents (major ions) c. physical properties, such as electrical conductivity d. summer "hoar frost" formation (visually apparent on a light table) e. Since these properties do not diffuse (significantly), they can record older layers than can δ18O. 3. Any annual counting method will have some ambiguities that may lead to slight over-and under- counts. B. Flow models. Based on approximations of the physical equations driving ice flow. These may be decent, but they depend on a good knowledge of boundary conditions and their temporal evolution. These work best when used with chronological spikes deep in the record – the model helps “interpolate” between the chronological spikes. C. Correlation with other climate records 1. Climate record correlations 2. Gas correlations a. CO2 b. CH4 c. δ18O2 D. Direct dating methods 1. In principle, it should be possible to date the CO2 in the ice bubbles by AMS 14C. In reality, no one has reported a successful 14C date. One problem is that cosmic rays striking the ice convert some of the oxygen to carbon 14 (D. Lal). E. Other methods 1. Volcanic ash layers 2. Acidity spikes from volcanic eruptions 3. U-series dating of recoil products (Fireman) III. δ18O and δD evidence for T changes A. Stable isotope hydrology 10 δ18O in Precipitation (per mil) 0 -10 -20 -30 -40 -50 -60 -60 -40 -20 0 20 Surface Temperature (C) Observed δ18O versus observed T (annual mean). The annual means for the IAEA [1981a] sites are computed from monthly means through precipitation weighting. B. "Silty" ice near bottom - problematical δ18O. C. Camp Century Ice Core 1. Time scale is based on annual cycles of δ18O for the first millenium. 2. Below that level, time scale is based on flow model and on correlation with other climate records. Note big surprise awaiting on deep Camp Century time scale! 3. Glacial/interglacial climate signal; Younger Dryas; interstadials GRIP/GISP2 problems: boudinage, basal ice folding (I) (II) Ice Flow B B Depth A Light A Heavy Isotope Composition I) A typical shear fold in the basal part of a glacier. If the ice had not been previously folded, a stratigraphic sequence of a climate-related property (such as oxygen isotope composition) sampled by a borehole at point A might produce a simple monotonic trend as shown by line A in part II. Sampling at B, after folding, would yield the sequence shown as B in II. Multiple folding can complicate the sequence further. I. Taylor Dome J. EPICA Dome C Percentage of Surface Warming Depth (m) 0 0 20 40 60 80 100 1000 2000 3000 Calculated temperature perturbation in central West Antarctica, 10,000 years after a surface temperature change. B. Does spatially-calibrated ice core δ18O=temperature? a. Annual T cycle in central Greenland: ii. Holocene temperature and isotope variability: The "Little Ice Age" and "Medieval Warm Period" confirmed b. Borehole temperature analysis of the glacial section of GISP2. Adapted from source: Cuffey et al. (1995) b. Monte-Carlo borehole temperature analysis of the glacial section of GRIP.