Tim K. Lowenstein

Tropical Climate Changes at Millennial and Orbital Timescales on the Bolivian Altiplano

Tropical South America is one of the three main centres of the global, zonal overturning circulation of the equatorial atmosphere (generally termed the ‘Walker’ circulation). Although this area plays a key role in global climate cycles, little is known about South American climate history. Here we describe sediment cores and down-hole logging results of deep drilling in the Salar de Uyuni, on the Bolivian Altiplano, located in the tropical Andes. We demonstrate that during the past 50,000 years the Altiplano underwent important changes in effective moisture at both orbital (20,000-year) and millennial timescales. Long-duration wet periods, such as the Last Glacial Maximum—marked in the drill core by continuous deposition of lacustrine sediments—appear to have occurred in phase with summer insolation maxima produced by the Earths precessional cycle. Short-duration, millennial events correlate well with North Atlantic cold events, including Heinrich events 1 and 2, as well as the Younger Dryas episode. At both millennial and orbital timescales, cold sea surface temperatures in the high-latitude North Atlantic were coeval with wet conditions in tropical South America, suggesting a common forcing.

Melting behavior of fluid inclusions in laboratory-grown halite crystals in the systems NaClH2O, NaClKClH2O, NaClMgCl2H2O, and NaClCaCl2H2O☆

Abstract The chemical composition of aqueous fluid inclusions in crystals of halite can be accurately determined from observed melting behaviors of ice, hydrohalite, and sylvite. Some fluid inclusion melting behaviors observed in laboratory-grown halite crystals (systems NaClH 2 O, NaClKClH 2 O, NaClMgCl 2  H 2 O, and NaClCaCl 2 H 2 O) differ from predicted stable equilibrium relations. In the NaClH 2 O and NaClKClH 2 O systems, observed first melt temperatures are up to 15°C below the equilibrium eutectic temperatures of −21.2° and −22.9°C, respectively. The final melting temperature of ice, in the presence of hydrohalite, and the final melting temperature of hydrohalite are reproducible and match predicted melting temperatures. The limit of detection of sylvite daughter crystals in the NaClKClH 2 O system is approximately 5 wt% (≈ 1 molal) KCl. Final melting temperatures of sylvite match published equilibrium data to within 0.3°C. In the NaClKClH 2 O system at halite saturation, m KCl can be determined from the final sylvite dissolution temperature or from the final melting temperature of hydrohalite. Fluid inclusions in the NaClMgCl 2 H 2 O and NaClCaCl 2 H 2 O systems that form stable salt hydrates (MgCl 2 · 12H 2 O and CaCl 2 · 6H 2 O) during freezing first melt within 3°C of predicted eutectic temperatures (−37° and −52°C). However, fluid inclusions with MgCl 2 or CaCl 2 may also start melting at temperatures as low as −80°C. Such low first melt temperatures indicate the presence of metastable salt hydrates (presumably MgCl 2 · 8H 2 O, MgCl 2 · 6H 2 O or CaCl 2 · 4H 2 O). The formation of metastable phases during freezing of fluid inclusions can lead to misinterpretation of the chemical composition of fluid inclusions in natural samples. This is especially true for fluid inclusions with first melt temperatures below −37°C which may be erroneously interpreted as being rich in CaCl 2 . The final melting of ice in the presence of hydrohalite may vary by more than 15°C in fluid inclusions of different size but identical bulk composition, and occurs at lower temperatures than predicted in fluid inclusions from the NaClMgCl 2 H 2 O and NaClCaCl 2 H 2 O systems. However, the final melting temperature of ice in inclusions which fail to nucleate hydrohalite, and the final melting temperature of hydrohalite are reproducible to within ±0.1 ° C and can be used to determine MgCl 2 and CaCl 2 molalities.

Secular variation in seawater chemistry and the origin of calcium chloride basinal brines

CaCl 2 basinal brines, which are present in most Phanerozoic sedimentary basins, inherited their chemistries and salinities from evaporated paleoseawaters when the world oceans were Ca rich and SO 4 poor (CaCl 2 seas). CaCl 2 seas coincided with periods of rapid seafloor spreading, high influxes of mid-ocean-ridge brines rich in CaCl 2 , and elevated sea levels, conditions that favored accumulation of marine CaCl 2 brines in marginal and interior continental basins. Typical basinal brines in Silurian-Devonian formations of the interior Illinois basin, United States, show the same compositional trends as those of progressively evaporated CaCl 2 -rich Silurian seawater. Chemical deviations can be accounted for quantitatively by brine-rock reactions during burial (dolomitization, dolomite and K-feldspar cement). This explanation for the origin of CaCl 2 basinal brines contrasts with others that assume constancy of seawater chemistry and involve more complex brine-rock interactions.

Atmospheric pCO2 since 60 Ma from records of seawater pH, calcium, and primary carbonate mineralogy

A 60 m.y. record of atmospheric p CO2 has been refined from knowledge of (1) secular changes in the major ion composition of seawater (particularly Ca and Mg) and (2) oscillations in the mineralogy of primary oceanic carbonate sediments. Both factors have had a significant impact on the chemistry of the ocean carbonate buffer system. Calculated atmospheric p CO2 oscillated between values of 100–300 ppm and to maxima of 1200–2500 ppm from 60 to 40 Ma and varied between 100 and 300 ppm from 25 Ma to the present. The refined p CO2 values are significantly lower than previous estimates made from seawater pH data where total dissolved inorganic carbon was assumed constant and more in line with modeling and stomatal index estimations of atmospheric p CO2 for the Tertiary.

Microbial communities in fluid inclusions and long-term survival in halite

Fluid inclusions in modern and ancient buried halite from Death Valley and Saline Valley, California, USA, contain an ecosystem of “salt-loving” (halophilic) prokaryotes and eukaryotes, some of which are alive. Prokaryotes may survive inside fluid inclusions for tens of thousands of years using carbon and other metabolites supplied by the trapped microbial community, most notably the single-celled alga Dunaliella, an important primary producer in hypersaline systems. Deeper understanding of the long-term survival of prokaryotes in fluid inclusions will complement studies that further explore microbial life on Earth and elsewhere in the solar system, where materials that potentially harbor microorganisms are millions and even billions of years old.

200 k.y. paleoclimate record from Death Valley salt core

A 186-m-long core (DV93-1) from Death Valley, California, composed of interbedded salts and muds contains a 200 k.y. record of closed-basin environments and paleoclimates, interpreted on the basis of sedimentology, ostracodes, homogenization temperatures of fluid inclusions in halite, and correlation with shoreline tufa. The 200 k.y. paleoclimate record is dominated by two dry and/or warm and wet and cold cycles that occurred on a 100 k.y. time scale. These cycles begin with mud-flat deposits (192 ka to bottom of core, and 60 ka to 120 ka). Wetter and/or colder conditions produced greater effective moisture; saline pan and shallow saline lake evaporites overlie mud-flat sediments (186 ka to 192 ka and 35 ka to 60 ka). Eventually, enough water entered Death Valley to sustain perennial lakes that had fluctuating water levels and salinities (120 ka to 186 ka and 10 ka to 35 ka). When more arid conditions returned, mud-flat deposits accumulated on top of the perennial lake sediments, completing the cycle (120 ka and 10 ka). Of particular significance are the major lacustrine phases, 10 ka to 35 ka and 120 ka to 186 ka (oxygen isotope stages 2 and 5e‐6), which represent markedly colder and wetter conditions than those of modern Death Valley. Of the two major lake periods, the penultimate glacial lakes were deeper and far longer lasting than those of the last glacial.

A 100 ka record of water tables and paleoclimates from salt cores, Death Valley, California

Abstract Sedimentary and petrographic features of evaporites and associated sediments from a 185 m deep core taken in Death Valley, CA, together with uranium-series dating have been used to reconstruct the history of water table fluctuations and climate changes in Death Valley for the past 100 ka. Death Valley has been arid during the Holocene (0–10 ka), with predominantly mudflat and saline pan subenvironments. A perennial lake, up to 90 m deep, existed in Death Valley from 10 to 35 ka. Saline pan and mudflat subenvironments dominated Death Valley from 35 to 100 ka. The chronology of changing subenvironments and water table fluctuations in Death Valley generally correlates with other climate records in the western US (Owens Lake and Searles Lake, CA, Browns Room cave calcite, NV), the marine oxygen isotope record, and the Vostok ice core record. Core intervals through saline pan sediments are composed of interbedded halite, chaotic muddy halite, and mud. The halite contains abundant vertical dissolution pipes, cemented with clear halite. These sediments record repeated flooding by dilute waters, dissolution of subaerially exposed surface salt crusts, deposition of mud from suspension, precipitation of halite during the saline lake phase, and cementation by diagenetic halite. Mudflat sediments consist of clayey silt, with sand patches and mud cracks, which document long periods of desiccation and the formation of efflorescent salt crusts from the evaporation of groundwater brines. Saline pan and mudflat deposits formed during periods when Death Valley was relatively arid, similar to the modern climate. Lacustrine deposits consist of mudhalite cycles, accumulated during the early lake stage, bedded thenardite (Na 2 SO 4 ) and mud above, and a cap of massive halite formed during the latest lake stage, all of which record fluctuating salinities and lake levels in a perennial system. Such deposits document a relatively wet climate with a high ratio of water inflow to evaporation. Ostracodes in mud layers represent the least saline, deepest lake phases. Halite layers are made of fine grained cumulates and clear, vertically-oriented crystals precipitated during shallower, perennial lake stages. Of significance is the nearly complete absence of syndepositional dissolution of saline minerals in the lacustrine interval, indicating that accumulated salts were permanently protected from dissolution by saline lake waters. Such evidence strongly suggests that lakes existed continually, without desiccating, for 25 ka between 10 and 35 ka B.P.

Origin of Ancient Potash Evaporites: Clues from the Modem Nonmarine Qaidam Basin of Western China

Modern potash salt deposits and associated brines of the Qaidam Basin, western China, demonstrate that some anomalous marine evaporites may have formed from nonmarine brines instead of seawater. Qaidam Basin brines are derived from meteoric river inflow mixed with small amounts of CaCl spring inflow similar in composition to many saline formation waters and hydrothermal brines. Evaporation of springenriched inflow yields a predicted mineral sequence including carnallite, bischofite, and tachyhydrite that is identical to several anomalous marine evaporites. Other mixtures of river and spring inflow produce the salt assemblage expected from evaporation of seawater.

Model of seawater composition for the Phanerozoic

We present an inverse model of Phanerozoic seawater com- position calibrated against updated paleoseawater compositions from fluid inclusions in marine halites. The model considers step- wise alteration of seawater composition via: (1) variable input of river water, (2) variable rates of alteration of seawater through reactions at mid-ocean ridges, and (3) variable rates of alteration of seawater through reactions on ridge flanks and across the ocean floor in general. The model achieves agreement with paleoseawater fluid inclusion data for Na 1 ,C a 2 1 ,S O 4 2 2 , and K 1 , particularly when variable runoff is considered. Variable rates of basalt- seawater interactions at both ridges and ridge flanks are required to understand the evolution of seawater, particularly the observed, near-constant concentration of K 1 through time.

The major-ion composition of silurian seawater

One-hundred fluid inclusions in Silurian marine halite were analyzed in order to determine the major-ion composition of Silurian seawater. The samples analyzed were from three formations in the Late Silurian Michigan Basin, the A-1, A-2, and B Evaporites of the Salina Group, and one formation in the Early Silurian Canning Basin (Australia), the Mallowa Salt of the Carribuddy Group. The results indicate that the major-ion composition of Silurian seawater was not the same as present-day seawater. The Silurian ocean had lower concentrations of Mg2+, Na+, and SO42−, and much higher concentrations of Ca2+ relative to the ocean’s present-day composition. Furthermore, Silurian seawater had Ca2+ in excess of SO42−. Evaporation of Silurian seawater of the composition determined in this study produces KCl-type potash minerals that lack the MgSO4-type late stage salts formed during the evaporation of present-day seawater. The relatively low Na+ concentrations in Silurian seawater support the hypothesis that oscillations in the major-ion composition of the oceans are primarily controlled by changes in the flux of mid-ocean ridge brine and riverine inputs and not global or basin-scale, seawater-driven dolomitization. The Mg2+/Ca2+ ratio of Silurian seawater was ∼1.4, and the K+/Ca2+ ratio was ∼0.3, both of which differ from the present-day counterparts of 5 and 1, respectively. Seawaters with Mg2+/Ca2+ 2 (e.g., modern seawater) facilitate the precipitation of aragonite and high-magnesian calcite. Therefore, the early Paleozoic calcite seas were likely due to the low Mg2+/Ca2+ ratio of seawater, not the pCO2 of the Silurian atmosphere.