Earth Climate Part 2 Chapter 6 and 7 Notes. Notes below are the things I highlighted in my book.

Chapter 6: Greenhouse Climate

6-1 Model Simulation of the Cretaceous Greenhouse

• Temperatures near the equator 100 million years ago were estimated to be a few degrees higher than those today, but those at the North Pole were as much as 20C higher and at the South Pole, 40C or higher.
• The reason for the especially large temperature difference at the South Pole is the absence of the Antarctic ice sheet in the Cretaceous model simulation.
  • Today, this ice sheet reaches elevations of 3 to 4 kilometers, where temperatures are 20 to 25C colder than at sea level because of the lapse rate effect.
• Geologic evidence indicates that this ice sheet did not exist during much of Cretaceous time
• GCMs can simulate only a few years or decades of elapsed time, and ice sheets cannot grow or melt that quickly.
  • Because of this time limitation, the presence or absence of ice sheets is specified in advance as boundary-condition input to this kind of GCM simulation.
  • Without the Antarctic ice sheet, Earth’s surface at the South Pole was close to sea level and the simulated polar temperatures were cold, but much less frigid than if an ice sheet had been present
• Multiple simulations occurred to test the results and lead to slightly different results due to different quantities of carbon dioxide

6-2 What Explains the Data-Model Mismatch

• the explanation for the data-model mismatches could lie either in the geologic data, in the climate model simulation, or in some combination of the two.
• One possible explanation for the mismatch has focused on shortcomings in the climate models.
  • ocean circulation in O-GCMs was still very crude because it lacked upwelling along coastlines and near the equator as well as deep water circulation
• The possibility that the Cretaceous Ocean operated in a fundamentally different way compared to the present ocean.
  • If the ocean carried twice as much heat to the poles in the Cretaceous as it does today, the two major problems in the data-model mismatch might be resolved.
  • In this ocean heat transport hypothesis, the poles would be warmed by the greater influx of ocean heat from low latitudes, and the tropics would not be heated as much because of their greater loss of heat to the poles
• A related idea was that deep water might have formed in the northern subtropics in regions of very high ocean salinity, rather than in polar regions like today.
  • High salinities could have made surface waters dense enough to sink into the deep ocean as warm saline bottom water. A strong flow of this warm deep water from the tropics to the poles might then have contributed to the poleward heat flux needed to warm polar regions and resolve the data-model mismatch.
• More recent experiments with improved ocean models have not supported the ocean heat transport hypothesis.
• A second kind of data-model mismatch occurred in climate simulations for 100 million years ago, and this mismatch persists in model simulations of the warm climates that continued over the next several tens of millions of years.
• Data from warm-adapted vegetation and from fossil reptiles suggest that continents at high and middle latitudes had moderate climates and did not freeze in winter. In contrast, all the GCM experiments described simulated hard freezes in winter across the interiors of the northern hemisphere continents.
  • Warm Arctic Ocean failed to keep the interiors of the northern continents warm enough in winter to prevent freezing because the winter heat losses were too large
• A second path of investigation of the data-model mismatch focused on the possibility that the proxy data used to reconstruct past climate might have been giving an incorrect “target signal” for the models. Several studies have produced evidence in support of this explanation.
• Most of the data used to reconstruct Cretaceous temperatures at lower latitudes have come from geochemical analyses of ocean plankton using oxygen isotope ratios.
  • New evidence (shells) suggests that tropical temperatures were considerably warmer than the target signal values
• The best-preserved shells indicated that Cretaceous temperatures in the tropics were as warm as 35C, or at least 5C above the initial target signal.
• The much warmer tropical target signal suggests that CO2 levels were actually considerably higher than those used in the model simulation with altered geography and higher CO2 levels.
• These higher CO2 levels also warmed the higher latitudes enough to explain why the interiors of the mid-latitude continents didn’t freeze in winter.

6-3 Relevance of Past Greenhouse Climate to the Future

• Simulations shows that global average temperature rises with increasing CO2 levels, but the relationship is not linear (directly proportional). Instead, Earth’s temperature reacts strongly to CO2 changes at the lower end of the range, but much less so to changes at the high end of the range.
• A major reason for the shape of this curve is the positive feedback effect provided by snow and sea ice. 
• At low (200 ppm) CO2 values the albedo temperature feedback makes the climate system in an icehouse world react strongly to changes in CO2.
• In contrast, much higher CO2 values reduce the average amount of snow and ice present at high latitudes provide little positive feedback.
• A second factor that accounts for the slower rise of temperature at higher CO2 levels is CO2 saturation. As CO2 concentrations rise, the atmosphere gradually reaches the point at which further CO2 increases have little effect in trapping additional back radiation from Earth’s surface.
• A third factor is water vapor feedback. A warm atmosphere with CO2 values of 1,000 ppm can hold much more water vapor than a cold atmosphere with values of 100 ppm. In this case, the positive feedback effect of water vapor on temperature grows stronger in warmer, higher CO2 conditions, in contrast to the diminished effect of albedo-temperature feedback.
• These sensitivity experiments, and others with different models, show that large ice sheets cannot exist anywhere on Earth when CO2 concentrations exceed 1,000 ppm.
• In the last 100 years, the CO2 concentration in the atmosphere has risen from 280 ppm to nearly 400 ppm because of the industrial activities of humans.
• Projections for the next two centuries indicate increases to at least 550 ppm, and possibly 1,000 ppm or more, nearly as high as those in the Cretaceous.
• Over tectonic time scales, the average level of the world ocean has risen and fallen by 200 meters or more against the margins of the continents
• During times of high sea level, the ocean flooded low-lying interiors of continents and formed large seas, which deposited and eroded thick sequences of coastal sediments
• Local tectonic factors that cause uplift or subsidence of the land also affect the relative vertical position of the ocean margin against the land
• Evidence for higher global sea levels in the past comes from marine sediments deposited both on coastal margins and in shallow interiors of continents
• The higher sea levels 80 to 100 million years ago have been attributed to two kinds of factors: tectonically driven changes that altered the volume of the ocean basins and their capacity to hold water and changes in the volume of water in the ocean basins caused by variations in climate.

6-4 Changes in volume of the Ocean Basins

• Heating causes the rocks in the ocean ridges to expand, and expansion of the rock causes the surface of the ocean crust to rise.
• All ocean ridges today have crests that lie at an average depth of 2,500 meters below the sea surface. Away from the crest, the depth profiles of these ridges follow a simple equation: Ridge Depth = 2500m (at age 0) + 350 (Crust age)^1/2
• This equation describes a ridge crest that starts at an initial depth of 2,500 meters below the sea surface and gradually deepens with age away from the ridge crest.
• Paleomagnetic evidence from today’s ocean shows that different ridges spread at different rates
• Because all ridge depths are nearly constant with age, crust of a given age will have been carried much farther from the ridge crest in a given amount of time in fast-spreading areas like the South Pacific than in slow-spreading ones like the North Atlantic.
• Fast spreading gives those ridges a wider elevation profile than the slow-spreading ones and through time, ridge profiles also vary on a globally averaged basis.
• Currently the globally averaged rate of seafloor spreading is relatively slow, mean ridge profiles are relatively narrow, and little water is displaced onto the continents.
• At times in the past when the average spreading rate was faster, mean ridge profiles would have been wider, pushing more ocean water up onto the land.
• Ridge profiles in the past can be reconstructed by resetting the ages of the past ridge crests to zero for the time being examined, using the dated magnetic lineations to recalculate the past ages of the ridge flanks relative to this adjusted “zero” age, and applying the equation shown above.
• Most plate tectonic movements do not change the net area of either the oceans or the continents
  • creation of new ocean crust at ocean ridge crests is balanced by destruction of ocean crust subducting into trenches, leaving the area of the ocean basins constant.
• However, collision of continents does alter the area of the ocean basins and in that way affects sea level.
• Because continental crust is low in density, colliding continents tend to float near Earth’s surface rather than being pushed or pulled down into Earth’s mantle.
• This process builds a high plateau that rises well above sea level and at the same time thickens the subsurface low-density “root” of the plateau as far as 60 to 70 kilometers below Earth’s surface
• Because collision drives continents together and forms plateaus with a double-thick crust, this thickening must result in a net loss in the area of continental crust.
• Viewed in the reverse sense, the absence of continental collisions 80 to 100 million years ago would have kept sea level tens of meters higher than now.
• Along many ocean margins, sediments eroded from the continents are delivered to ocean trenches where they are subducted.
• But along those margins without subduction zones, ocean sediments from major rivers pile up on the seafloor and displace ocean water by reducing the volume of the ocean basins.
• The Bengal Fan in the northern Indian Ocean is an enormous sediment pile built by eroded sediments from the Himalaya Mountains to the north during the last 55 million years. By some estimates, sediment filling of ocean basins by these eroded deposits can cause sea level increases of as much as 50 meters.

6-5 Climatic Factors

• Continent-sized ice sheets several kilometers in thickness and thousands of kilometers in lateral extent can extract enormous volumes of water from the ocean and store it on land.
• Today, the Antarctic ice sheet holds the equivalent of almost 60 meters of global sea level, and the Greenland ice sheet another 6 meters.
  
  • Their absence 80 to 100 million years ago would have made sea level 65 meters higher than today.
• Ocean water has the capacity to expand and contract with temperature changes. The thermal expansion coefficient of water averages about 1 part in 7,000 for each 1C of temperature change.
• Thermal expansion of the warmer ocean that existed 80 to 100 million years ago would have resulted in a global sea level about 7 meters higher than today
• One problem is the fact that water moving into the ocean basins represents a large weight added to the underlying ocean crust, which sags or rebounds accordingly.
  • This response of the ocean crust decreases the net magnitude of the change in sea level.
• The other complication has to do with translating a change in the volume of ocean water or in the volume of the ocean basins into actual movement of sea level against the complex shapes of the world’s continental margins.

6-6 Assessment of Higher Cretaceous Sea Level

• Estimates of the highest Cretaceous Sea level have ranged widely.
• During the 1980s, research scientists at the Exxon Oil Company estimated sea levels of 250 meters or more above present values.
• Subsequent estimates that have considered important processes that occurred after the sediments were deposited have yielded lower estimates.
  • The soft, water-rich sediments that were initially deposited have been compacted both by the weight of later-deposited sediments and by ocean water moving in and out across the continental shelf.
  • The underlying continental crust below has also been weighed down by the weight of the water, the sediments deposited, and by local tectonic effects that are difficult to distinguish from global trends.
• Attempts to remove these post-depositional effects have yielded much lower sea level estimates.
• Estimates of higher Cretaceous Sea levels now reach no more than about 125 meters above present levels
• A considerable fraction of these reduced estimates can be accounted for by the two well-constrained climatic factors: the 65 meter contribution from ice now stored in Antarctica and Greenland, and the 7 meters from thermal expansion because of warmer deep water during the Cretaceous.
• The remaining amount must be accounted for by some combination of the other factors. At this point, a full understanding of both the amount and the causes of the higher Cretaceous Sea levels remains elusive.

6-7 Effect of Changes in Sea Level on Climate

• The most likely effect of sea level change on climate is linked to the very different thermal responses of land and water
• flooding of the land will tend to moderate nearby extremes of climate and produce milder winters and cooler summers. Withdrawal of the sea should have the opposite effect.
• For large changes in global sea level, the synchronous invasion and withdrawal of the sea across low-lying portions of many continents should result in simultaneous fluctuations between harsh continental and mild maritime climates in these widely separated regions.
• Decades ago, climate scientists thought that sea level might be a factor in the long-term succession of glacial (icehouse) versus nonglacial (greenhouse) climates, or maybe even the main control on these changes, but not viable anymore.
• As a result, the hypothesis that sea level is the major control of long-term glaciation finds little or no support today. Glaciation is now seen as a major cause of low sea level (because of storage of ocean water in ice sheets) rather than a result.

Chapter 7. From Greenhouse to Icehouse: The last 50 million years

7-1 Evidence from Ice and Vegetation

• two kinds of glacial ice form on land.
  • Small mountain glaciers appear on the tops of high mountains and large ice sheets cover much larger areas of the continents.
• Because temperatures vary widely from region to region and with altitude, the conditions that permit ice to persist year round do not appear at the same time in all areas
• The size of the Antarctic ice sheet has increased irregularly toward the present, with a major growth phase near 13 million years ago.
• Greater amounts of ice-rafted debris in nearby ocean sediments suggest additional increases in Antarctic ice during the last 10 million years.
• Today, even in a warm interglacial interval, more than 97% of Antarctica is buried under ice
• The first continental ice sheets of significant size appeared 2.75 million years ago. These ice sheets grew and melted in repeated cycles, and their maximum size in North America and Eurasia increased after 0.9 million years ago.
• Fossil remains of vegetation in both hemispheres also indicate a progressive cooling over the last 50 million years.
• Today, the only vegetation in Antarctica is lichen and algae found in summer meltwater ponds in ice-free regions of a few coastal valleys.
• A similar long-term cooling trend is evident in north polar regions. Gradually the warm conditions in the Arctic gave way to colder climates.
• The shapes of tree leaves have also been used to reconstruct past climate.
  • Leaves of trees living today in the warm tropics tend to have smoothly rounded margins and leaves of trees in cooler climates generally have irregular edges, usually jagged or serrated in outline.
• One low-resolution record derived from leaf-margin evidence in western North America shows an ongoing cooling over the last 55 million years. Although interrupted by small warm intervals, the longer-term trend toward lower temperatures persists.

7-2 Evidence from Oxygen Isotope Measurements

• The first occurrences of ice sheets on land are difficult to define and their subsequent fluctuations in size even more so.
  • lakes that accumulate remains of past continental vegetation in their muddy sediments rarely persist for millions of years.
  • parts of the deep ocean have accumulated a continuous climatic record with quantitative information about climate change across the entire 50-million-year interval.
• The most important climatic record in the ocean is the oxygen isotope signal.
  • Most of the oxygen in nature occurs either as the very abundant 16O isotope or the less abundant 18O isotope. Scientists refer to changes in the relative amounts of these two isotopes as variations in 𝛅18O, measured as changes in parts per thousand.
• Typical modern δ18O values are 0 to -2% for the surface ocean, 3 to 4% for the deep ocean, and -30 to -55% for ice sheets. 
• Foraminifera shells gives climate scientists information on past variations in the two isotopes of oxygen in the ocean, as quantified by the δ18O signal.
• Changes in size of the ice sheets also alter δ18O values in foraminiferal shells.
• Ice sheets are formed from water vapor evaporated from the oceans and later precipitated as snow.
  • Because the snow and ice are enriched in the lighter 16O isotope, more of the heavier 18O isotope is left behind in the oceans. This enrichment process is called fractionation. The δ18O value of ocean water becomes more positive as ice sheets grow and preferentially store the 16O isotope.
• The opposite is true if ice melts. If all the ice present on Antarctica and Greenland today melted and flowed back into the ocean, it would deliver a large volume of 16O-rich meltwater that would shift the ocean’s average δ18O value from its present value of 0% to a value of 21%.
•  A simple equation that summarizes these processes is
  • • Δδ18Oc = Δδ18Ow x 0.23ΔT
    • This equation indicates that measured changes in δ18O values in the shells of foraminifera (Δδ18Oc) result from changes in the mean δ18O value of the oceans (Δδ18Ow) caused by storage of 16O in ice sheets, and from variations in the temperature of the water in which the shell forms (Δ T). The value 0.23 results from inverting the 4.28C value noted earlier (1/4.2 5 0.23).
•  A record of δ18Oc over the last 70 million years has been compiled from benthic foraminifera living on the ocean floor.
  • This signal begins to trend erratically toward more positive values near 50 million years ago, and intervals of fastest change occur near 35 million years ago, 13 million years ago, and in the last 3 million years.
• These changes toward more positive δ18Oc values are caused by a combination of cooling of the deep ocean, and growth of ice sheets on land.
• No evidence exists of significant amounts of ice on Antarctica or anywhere else on Earth prior to 35 million years ago.
  • During the interval between 50 and 35 million years ago, the δ18Oc values increased from 20.75 to 10.75%, a net change of 11.5%. The temperature/ δ18Oc relationship tells us that deep waters must have cooled by more than 6C during this interval before major ice sheets appeared.
• Between 35 million years ago and today, the deep ocean δ18Oc values increased from about 10.75 to 13.5%, a further increase of 2.75%.
• Because the temperature of today’s deep ocean averages about 2C and has cooled by about 14C over the last 50 million years, the average deep-ocean temperature must have been near 16C before 50 million years ago.
• If deep water formed mainly in high latitudes as it does today, the polar climates that sent such warm water into the deep ocean must have been much warmer than they are today

7-3 Evidence from Mg/Ca Measurement

• Another valuable index of the climatic response of the ocean comes from analyzing the ratio of the elements magnesium (Mg) and calcium (Ca) in the shells of foraminifera.
  • The process by which Mg substitutes for Ca in the foraminiferal shells is dependent on the temperature of the waters in which the shells form.
  • Across the cold temperature range of the deep ocean, the relationship is nearly linear.
  • As was the case with the δ18O signal, adjustments must also be made for long-term changes in the Mg concentration of the global ocean.
• The trend in deep-water temperature reconstructed from Mg/Ca changes is very similar to that of δ18O. It confirms a long-term cooling of about 14C inferred from δ18O evidence and from changes on land and it shows particularly large steps at or near 35, 13, and 3 million years ago.
• In detail, however, these and other indices of past climate at times disagree.
  • One reason for the disagreements could be the lack of enough samples to resolve finer detail, especially in the leaf-margin reconstructions of climate.
  • Another reason is the fact that no one signal can possibly represent all aspects of global climate.
• A wide and convincing array of evidence documents the progressive cooling of both poles and of mid-latitude areas during the last 50 million years.
• Several hypotheses have been put forward to explain this gradual greenhouse-to-icehouse transition.

7-4 Evaluation of Gateway Hypothesis

• Some climate scientists have called on the opening or closing of ocean gateways to explain the onset of both southern and northern glaciation during the last 50 million years.
• These hypotheses have focused on narrow passages that allow or impede exchanges of ocean water between ocean basins. They have proposed that changes in key gateways caused glaciations by altering the poleward transport of heat or salt.

Case Study 1: Antarctica

• During the last 50 million years, the last of the Gondwana continents connected to Antarctica split off and moved north, leaving Antarctica isolated and surrounded by a circumpolar ocean.
• James Kennett proposed that this breakup caused the onset of glaciation in Antarctica.
  • Before the continents separated, oceanic flow around Antarctica had been impeded by the land connections with South America and Australia.
  • Kennett hypothesized that these barriers had diverted warm ocean currents poleward from lower latitudes and had delivered enough heat to Antarctica to prevent glaciation.
  • After the continents separated, a strong, unimpeded west-to-east flow developed around Antarctica.
  • Kennett proposed that the loss of the warm poleward flow of heat isolated Antarctica, causing it to cool and glaciation to begin.
  • Australia separated from Antarctica around 37 to 33 million years ago, the same interval as the first glaciation in Antarctica.
  • Climate scientists have used sensitivity tests with atmosphere/ocean general circulation models to evaluate this hypothesis.
    • Drake’s Passage was closed in one simulation and left open in a companion simulation, while all other features of Earth’s geography were kept the same.
    • The model results suggested that opening Drake’s Passage did not significantly alter ocean (or atmospheric) temperatures near Antarctica. Instead, the models simulated a frigid climate over Antarctica regardless of the pattern of ocean flow.
    • The combined heat transport by the ocean and the atmosphere remained about the same in both experiments.
    • Apparently, the opening of the circum-Antarctic gateway was not a critical factor in the onset and development of Antarctic glaciation.

Case Study 2: The Central American Seaway

• During the last 10 million years, uplift in Central America gradually closed a deep ocean passage that had previously separated North and South America in the region of Panama. Several climate scientists have suggested that closure of this passage and the start of northern glaciation are linked.
• They hypothesized that construction of the Isthmus of Panama blocked the strong flow of warm, salty tropical water that had previously been driven westward out of the tropical Atlantic Ocean and into the eastern Pacific by trade winds.
• The newly formed isthmus should have redirected this flow into the Gulf Stream and toward the high latitudes of the Atlantic.
• They further hypothesized that this strengthened northward flow of warm, salty water would have suppressed the formation of sea ice in north polar regions because saltier waters resist freezing better than fresh waters. According to this hypothesis, the reduced cover of sea ice would have made more moisture from the ocean available to nearby landmasses and triggered the growth of ice sheets
• Climate modelers have used atmosphere/ocean general circulation models to test this hypothesis by running pairs of experiments with the Panama region configured both as an open gateway passage and as a closed-off isthmus.
  • These two configurations roughly correspond to the end points of the tectonic changes that occurred between about 10 and 4 million years ago.
  • These simulations confirmed the prediction that warm, salty water would have been retained in the Atlantic Ocean and redirected toward northern latitudes.
  • Closing of the Panamanian isthmus also cut off a return flow of low-salinity Pacific water into the Atlantic. Blockage of this low-salinity return flow by the isthmus further increased the salinity of the northward-flowing Atlantic water
• In a critical respect, however, the model simulations contradicted the gateway hypothesis.
  • The stronger northward flow of salty water in the Atlantic and the resulting reduction of sea ice caused by closing the Isthmus of Panama did not greatly alter precipitation patterns around the high-latitude North Atlantic.
  • As a result, the increase in moisture hypothesized as necessary to grow ice sheets did not occur. Instead, the stronger northward flow of water from the tropics and subtropics transferred a large amount of heat to the atmosphere and warmed the regions where ice sheets were eventually to form. This warming increased summer melting of snow and opposed glacial inception, contrary to the gateway hypothesis.
  • Another problem with the gateway hypothesis is that northern hemisphere climate continued to cool, and ice sheets grew larger in size, millions of years after the Panamanian isthmus had emerged and closed off all flow between the Atlantic and Pacific oceans

Assessment of Gateway Changes

• These two case examples invoke very different (actually opposed) roles for the ocean in glacial inception.
  • For Antarctica, a reduced poleward flow of warm ocean water was proposed to have caused a cooling and subsequent glaciation. For the Isthmus of Panama, an increased poleward flow of warm ocean water was invoked as the cause of an increase in moisture flux that promoted glaciation.
• These differing assumptions reflect past disagreements among climate scientists about how the ocean affects ice sheets.
• Some climate scientists have suggested that a warmer ocean will release more latent heat (water vapor) to the atmosphere and thereby supply more moisture (snow) to aid ice growth.
• Most climate scientists, however, emphasize the fact that a warmer ocean will release more sensible heat to the overlying atmosphere and thereby potentially melt more ice.
• In any case, neither set of modeling experiments supports the hypothesis that changes in poleward flow of warm ocean water tied to gateway changes would have had a large enough effect on climate to initiate the growth of ice sheets.
  • This criticism needs to be tempered by the realization that the oceanic components of general circulation models are still at a relatively early stage of development and that they simulate precipitation patterns rather poorly.
  • A second criticism of the gateway hypothesis is that each opening or closing of a gateway occurs over perhaps 10 million years, while the cooling has lasted far longer.
• Major gateway changes redistribute heat and salt at the high-latitude sites where deep waters form, and these surface-water changes may affect formation of deep water.
• Higher-salinity surface waters promote stronger deep-water formation because the water is already dense when it encounters cold winter air masses at high latitudes.
• The results from this experiment agree with independent evidence that indicates increased formation of deep water in the North Atlantic between 10 and 4 million years ago, the interval over which the Central American seaway was gradually reduced and finally closed off by the emergence of the Central American isthmus

Hypothesis that Invokes Changes in CO2

• Most climate scientists regard falling CO2 levels as the cause of 50 mil years cooling, whether by slower CO2 input (BLAG spreading rate) or by faster removal (uplift weathering). Simulations with climate models support the importance of CO2.
• The climate modelers Rob DeConto and David Pollard and colleagues used a general circulation model to run a series of sensitivity tests with CO2 concentrations specified at values ranging from 280 to 1,400 ppm.
  • major ice sheets formed on the pole-centered Antarctic continent when CO2 values fell below 750 ppm, but did not form in Greenland or elsewhere in the Northern Hemisphere until the concentrations reached 280 ppm or lower. This offset sequence of glacial onsets is broadly consistent with geologic evidence.
• The proxy reconstructions suggest a major decrease in CO2 levels during the interval between 50 and 30–25 million years ago. This trend is broadly consistent with the first appearance of substantial ice in Antarctica by 34 million years ago and supports the idea that decreasing CO2 levels drove the cooling around that time.
• But the reconstructed CO2 concentrations for later intervals pose a major problem. By 25 to 20 million years ago, the proxy CO2 estimates suggest values around 250 ppm. Values that low should have caused northern ice sheets to appear by 20 million years ago, but they did not
• Even more perplexingly, no further decrease in CO2 occurred between 20 million years ago and today. Yet evidence from both poles and lower latitudes clearly shows that the planet continued to cool during this interval
• If the proxy CO2 estimates are accurate, CO2 could not have been the cause of the cooling after 20 million years ago, but this would leave climate scientists with no widely accepted explanation for that cooling.
• The other alternative is that the CO2 proxy estimates are not valid. Some support for this idea comes from the fact that multiple proxies spanning the same intervals can disagree by very large amounts. 
• Beginning after 5 to 4 million years ago, many areas of the ocean show a marked trend toward colder sea surface temperatures, often with a particularly prominent step near 3 to 2.5 million years ago.
• Some scientists have proposed that cooling in one or another of these regions triggered the onset of northern hemisphere glaciation by 2.75 million years ago but cause and effect are difficult to disentangle when changes occur nearly synchronously.
• Other scientists have interpreted the nearly global expression of these changes as evidence of a common driving factor, with decreasing CO2 the cause most often invoked.

7-5 Evaluation of the BLAG (Spreading Rate) Hypothesis

• To explain the global cooling of the last 50 million years, the spreading rate hypothesis must pass one critical test.
• A slowing of global mean spreading and subduction rates must have occurred through that interval, leading to slower rates of CO2 input to the atmosphere.
• Before 15 million years ago, global mean spreading (and subduction) rates show a decrease consistent with the spreading rate hypothesis. But since 15 million years ago, the mean rates of spreading and subduction have leveled out or even slightly increased to a present value comparable to the one that existed almost 30 million years ago.
• The slight increase in spreading rates since 15 million years ago should have put more CO2 into the atmosphere and warmed global climate. Instead, climate continued to cool, with a substantial increase in size of the Antarctic ice sheet
• Could additional volcanic input of CO2 at sites away from ocean ridges and subduction zones explain this discrepancy? No
• In Summary, the evidence indicates that the spreading rate hypothesis may have been a cause of global cooling before 15 million years ago, and particularly before 30 or 40 million years ago. But it predicts a warming during the last 15 million years, when in fact a substantial cooling has actually occurred.
• An alternative possibility has been informally considered by some scientists, but not yet posed as a full hypothesis. The amount of carbon carried down into ocean trenches may have varied in the past because of changes in the kind of sediments being subducted, even in the absence of changes in spreading rates.
• If a future change in the plate tectonic regime were to initiate subduction in the Atlantic Ocean, an enormous amount of carbonate would begin to be subducted in trenches for later melting and eventual release to the atmosphere through volcanoes. As a result, atmospheric CO2 values would increase, even in the absence of changes in spreading rates. At this point, however, this idea has not yet been tested.

7-6 Evaluation of the Uplift Weathering Hypothesis

• For uplift weathering hypothesis to explain the cooling, three main requirements must be met:
  • the amount of high-elevation terrain in existence today must be unusually large in comparison with earlier intervals
  • this high terrain must be causing unusual amounts of rock fragmentation
  • the exposure of fresh debris must be causing unusually high rates of chemical weathering.
• To determine whether or not these requirements are met, we have to compare the present with some interval in the past.
  • The last half of the Cretaceous interval, from 100 to 65 million years ago, is a useful basis for comparison, for two reasons:
    • abundant evidence of this interval is still left in the geologic record
    • it was an interval of full greenhouse climate

Prediction 1: Extensive High Terrain

• Marine sediments deposited at or below sea level 100 to 65 million years ago are now found at high elevations at many mountains.
• These sediments have been uplifted from sea level to their present heights in the last 70 million years or less
• Plate tectonic processes continually cause uplift in many regions throughout geologic time, while erosion continually attacks the highest topography and wears it down.
  • the highest topography during any interval of geologic time is always recent in origin
• In contrast, no major continental collisions occurred from 100 to 65 million years ago, and no massive plateaus existed then, or for the preceding 150 million years. The presence of the Tibetan Plateau and Himalayan complex provides a conclusive argument that an unusually massive amount of high topography exists at the present.
• Most other high-elevation regions on Earth have been formed by subduction of ocean crust beneath continental margins.
• Because subduction is an ongoing process, mountain terrain has existed continuously through time, in contrast to plateaus produced by sporadic continental collisions.
• Another kind of high terrain is the extensive low plateau in eastern and southern Africa at an elevation of 1 km.
  • results from deep seated heating that causes a broad upward doming and outpouring of volcanic lava
• Subduction has also occurred along western North America for some 200 million years, but scientists disagree about the history of uplift in this region over the last 50 million years
• In Summary, the existence of the massive Tibetan Plateau alone makes modern topography unusual, consistent with the uplift weathering hypothesis. Other regions of recently elevated terrain also exist in the South American Andes Mountains and in East Africa.

Prediction 2: Unusual Physical Weathering

• The second test of the uplift weathering hypothesis is whether or not today’s high topography has caused higher rates of physical weathering and rock fragmentation than intervals farther in the past.
• The measurements clearly indicate intense physical weathering at the present in the two regions with highest terrain. (Himalayas and Amazon River)
• The best record of rates of erosion lies in sediments deposited in ocean basins by rivers.
• Climate model experiments indicate that this influx of sediment to the Indian Ocean is a result of two factors: (1)
  • creation of steep terrain along the southern Himalayan margin of the Tibetan Plateau
  • the fact that a plateau the size of Tibet in effect creates its own weather, including the powerful South Asian monsoons.
• Monsoons result from different rates of heating of continents and oceans due to the different heat capacities of land and water.
  • In summer, solar heating warms the continents more rapidly than the nearby oceans. The heated air over the land rises and pulls in moist ocean air. The high Himalayas on the southern margins of the plateau form an obstacle to the incoming ocean air, forcing it to rise and its water vapor to condense in cooler temperatures at high altitudes. As a result, these steep slopes become a natural focal point for strong summer monsoon rains, which release latent heat and fuel even more powerful monsoons.
  • The powerful South Asian monsoons owe part of their existence to the rise of the Tibetan Plateau.
• Unfortunately, ocean sediments cannot give us a definitive estimate of global rates of physical weathering in the past.
• Many sediments were subducted into nearby trenches and cannot be quantified.
• some sediments deposited on the seafloor are eroded and redeposited, and this reworking skews compilations of sediment deposition rates through time toward younger ages.
• Rapid deposition of huge amounts of Himalayan sediment thus supports the hypothesis that physical weathering is stronger today on a global basis than in earlier times, but this conclusion remains tentative because of the sediment lost to subduction.

Prediction 3: Unusual Chemical Weathering

• The final test of the uplift weathering hypothesis is whether or not the global average rate of chemical weathering is higher today than it was in the past.
• Unfortunately, chemical weathering rates are difficult to determine
• Climate scientists quantify modern rates of chemical weathering on a regional basis by measuring the total amount of ions dissolved and transported in rivers.
  • This measure reflects the amount of chemical weathering in the watershed drained by each river, but disturbances of natural weathering processes by humans complicate such studies.
  • it is difficult to distinguish between the ions provided by slow weathering of silicate rocks and those resulting from rapid dissolution of carbonate rocks.
    • Only hydrolysis affects the CO2 balance in the atmosphere.
  • it is virtually impossible to study enough rivers to reach an accurate estimate of the global weathering rate because too many rivers contribute significantly to the global total
• the case for unusual chemical weathering during the last tens of millions of years rests on a plausibility argument based on several observations:
  • the unusual height and extent of the Tibetan-Himalayan complex, the unusual strength of the monsoon rains, and the unusual volume of physically eroded sediments deposited in the nearby ocean.
  • this combination of favorable factors should promote unusually rapid chemical weathering and cause CO2 removal from the atmosphere. But inference is not proof.
• In Summary, the cause of global cooling during the last 50 million years remains uncertain. The most likely culprit is a decrease in atmospheric CO2, driven by some combination of decreased input from seafloor spreading processes and increased removal tied to chemical weathering.

Future Climate change at tectonic time scales

• it is impossible to predict the changes in climate that will be caused by tectonic processes in the future.
  • the driving forces are not sufficiently understood.
• The tectonic-scale climatic trends of the distant future will also be influenced by positive and negative feedback processes.
  • negative feedback from chemical weathering is an integral part of the BLAG hypothesis, but a rough calculation shows that it could also have acted to offset much of the increase in chemical weathering driven by uplift (Figure 7-19).
  • A 50-fold increase in weathering over 1% of Earth’s land surface would increase global chemical weathering by 50%
• the moderating effects of the chemical weathering thermostat may have balanced the effects of uplift-driven weathering during the last 50 million years (or nearly so). Calculation in book
• possibility that the process of global cooling could produce positive feedbacks that keep driving climate toward even colder conditions.
• An increase in weathering is particularly likely if the lower limits of the mountain glaciers move up and down the sides of mountains, alternately grinding fresh rock and then exposing it to the atmosphere for chemical weathering.
• Similar positive feedback may occur in connection with large, continental-scale ice sheets
• If the ice sheets simply remain at or near their maximum extents for millions of years, they will likely slow the overall rate of chemical weathering.
• Further research is obviously needed to assess the role of tectonic forcing and internal feedbacks, both positive and negative.
  • projections say that Earth’s long-term “forecast” over tectonic time scales would entail colder temperatures and more ice
  • But in the meantime, other factors operating on shorter time scales will drive climate changes that are far more relevant to our immediate concerns