Over the past 2.5m year ice age numerous glaciated periods (glacials) and warm periods (interglacials) have come and gone. The most recent continental glaciers began to melt around 20,000 years ago and really got underway away about 11,000 years ago leaving ice sheets only in Antarctica and Greenland. The end of glacials appears to be synchronous with one of the Milankovitch cycles; namely the variation in the tilt (obliquity) of the earth's axis. At the beginning of the present ice age, which we are in the middle of, the cycle was 41,000 years. However over the past million years of the present 2.5 million year Glacial Age, only every third or so obliquity nudge has resulted in an interglacial. Glacials over the past million years or so have been lasting on the order of 100,000 years. Coincidentally with the melting of continental glaciers there is a sharp rise in Carbon dioxide
As suggested in a previous blog it seems unlikely that some sudden source of Carbon dioxide would occur exactly in sinc with the Milankovitch cycle. As odd as it seems it is more likely that the rise in CO2 is somehow the result of the melting. Of course, once sufficient CO2 is released, a run-away melting will occur. One likely positive feed back (warming causing more warming) is the ability of the oceans to hold less Carbon dioxide when they are warm than when they are cold. As warming starts, presumably as a result of obliquity, the oceans could give out Carbon dioxide or at the very least, not absorb as much.
In a previous blog, I suggested that methane clathrates and carbon dioxide clathrates could accumulate under an ice sheet once it had thickened to a few hundred meters. Sources of methane and Carbon dioxide include coal measures, liquid and gaseous hydrocarbon deposits and shale beds as well as the decomposition of organic material buried by the accumulating ice. If the ice covered over an area of permafrost**, the methane stored in the permafrost could also find its way to the bottom of the ice sheet. Ice sheets cause the depression of the land by about a third of a km for every km of ice added# and this might well act as a natural 'fracting', increasing the escape of such gases from all the mentioned sources. All this carbon would be sitting there at the bottom of the ice sheet ready to be released if the continental ice sheet started to melt. If sufficient was released in a burst, the green house effect could lead to a feedback, melting more ice causing the release of more gas and causing more melting etc. This run-away greenhouse effects would only end when the ice sheet was all melted. Following the melt, the slow ever present sequestering of carbon in the various carbon sinks would continue until it was possible for the accumulation of snow to start again.
# The basaltic rock on which the continents float has a specific Gravity (SG) of just over 3. Hence a km of ice with a specific gravity of a tad under 1 would push down the continent about a third of a km.
**It is interesting to note that if snow is accumulating on an area of permafrost, the snow will insulate the underlying ground. The 0 degree contour at the bottom of the permafrost will move upward as geothermal heat melts it. Eventually, as the ice sheet deepens, all the permafrost, often rich in organic soil and methane clathrate, will be melted. The permafrost undergoing anaerobic break down and any already stored up methane* would be available to combine with the bottom layer of ice and form methane clathrate.
In this blog, I would like to suggest mechanisms which would explain why every nudge from the Milankovitch cycle does not end an ice age. Lets do a mind exercise.
Consider for simplicity a large continent like Australia. It is shaped like a hockey puck, flat on top with very little slope in any direction. Enough carbon has left the atmosphere and become sequestered in sinks for some snow to last through the summer. The process of going into a glacial is gradual. The accumulation of snow can only proceed at the rate of precipitation in the area of accumulation minus sublimation and melting. The process is somewhat accelerated by the albedo effect. When a significant area is covered with white snow, incident light is mostly reflected back into space increasing the cooling.
Snow occupies about 10 times as much volume as an equivalent weight of water but as the snow deepens, the weight of overlying snow on the bottom layers increases and air is squeezed out. By the time there is a hundred or so meters of snow, the bottom layers have been squeezed into ice with some inclusions of air. The ice at the bottom occupies about 10% more volume than an equivalent weight of water.
When the ice is only a few hundred meters thick it just sits there getting deeper and deeper. The land is flat so it doesn't move down hill and there isn't enough pressure yet to squeeze ice outward. At about 300m depth, there is enough pressure at the bottom of the ice layer for clathrates to form. Any methane or Carbon dioxide which is coming from the underlying land combines with the ice and is trapped.
When the ice has reached a km or so in depth, the pressure is great enough that ice begins to be squeezed toward the edges. Right in the middle of our hockey puck continent, there is no motion with respect to the underlying land. As you go toward the edges, the motion is faster and faster. Fast is all relative. Even in glaciated continents such as Greenland with 3 or so km of ice at the center, the motion at the edges is only a few to a few tens of meters per year. There are some individual glaciers which carry ice from the interior which are moving as much as 12km per year but this is down specific valleys and not along the entire perimeter of the glacier. Averaged over the years, ice can only fall off the edges at the rate that it accumulates on top. Since continental glaciers reach depths of at least 3km, clearly, less ice was expelled than has was accumulated over the formation of the 3km deep ice sheet. At some point, as the ice thickens, the rate of loss of ice will equal the rate of accumulation. It is likely that toward the middle of the continent, the ice doesn't move at the bottom relative to the land but is rather squeezed out of the middle layers of the ice. Toward the edges, ice would be moving over the ground.
Each Milankovitch nudge will probably result in some melting. If it is correct that clathrates have been collecting at the bottom of the ice sheet, this will cause an increase in the output of green house gases and the thicker the ice the faster this might occur due to faster rate of spread caused by the thicker ice. Also, the faster the ice is moving, the further into a melting climate the ice will be pushed. This may be the explanation for the start of an interglacial only every few Milankovitch nudges. Presumably a certain amount of green house gas is necessary to initiate a run away melting. The older an ice sheet, the more clathrate could accumulate at the bottom and the thicker the ice sheet, the faster its borders are moving outward. Therefore, the older and thicker an ice sheet and the more clathrate it has accumulated, the greater the chance of a run away melting when an obliquity nudge occurs.
As a further factor, with a thick ice sheet, the glacier at the edge will be moving across the ground and expelling clathrate. It would be expected that the concentration of clathrate would increase as you go toward the centre of an ice sheet. As the ice sheet edge melts back, more and more clathrate breaks down into the atmosphere. The thicker the ice, the more clathrate is likely to be stored under the ice and the faster the ice is moving at the edges. Thick ice should be much less stable than thin ice.
Another factor which might be relevant is the heat coming out of the earth. Although it varies widely from location to location, the temperature increases as you go down into the earth at about 25 degrees C per km. Put a layer of ice on the ground and this heat has to work it's way up to the surface of the ice. I haven't been able to find the factor for heat transmission in rock and in ice in order to compare them but for the sake of the argument let's say it is the same. Let's also assume that the average temperature at the top of the ice sheet is -50
The result is that with over 2km of ice depth and given some time to reach equilibrium, there should be water at the bottom of the ice sheet. Here we run into another wee codicil. If, as was suggested in a previous blog, methane clathrate has accumulated at the bottom of the ice sheet, it can stay frozen up to 18 degrees centigrade with sufficient pressure. Whatever the actual case, the general principle is that with a sufficiently deep ice sheet, the bottom layer should be melting. This may be another part of the explanation as to why only every three or so nudges by the Milankovitch cycle sets off an interglacial period. A sufficient depth of ice has to first collect to cause heat from the earth to liquefy its bottom and increase its horizontal movement on this lubricating layer . So what sort of evidence would support this hypothesis.
Test
A) at the bottom of present ice sheets, the temperature should be around 0 degrees.
B) There should be lakes below deep ice sheets where the topography allows.
C) It should be possible in some locations at least, to detect methane and possibly Carbon dioxide being evolved from the edges of ice sheets where they are melting.
D) Where ice sheets exit into the ocean and are at least 30m above sea level (and hence their base is 300m below sea level) there may be clathrates on the bottom layers in some locations.
E) Since the carbon released from the bottom of a 100,000 year ice sheet would be "old carbon" (in other words, carbon depleted in C14) There should be a dating anomaly from the end of the last ice period, 11,000 years ago*. This sudden influx of old carbon into the atmosphere should make wood, growing after the melting, look older. For wood from about 11,000 years ago, one might see successive growth rings from a tree looking older and older despite the fact that they each successive growth ring is younger than the previous one. A place to find suitable wood might be in tropical swamps where a log had sunk into the anaerobic mud or high mountains where trees such as the Bristle Cone Pine exist.
F) if the bottom layer of an ice sheet is composed of clathrates, you might find that even though a core found solid ice right to the bottom of the core, the temperature could be above 0 degrees. Clathrates can exist up to 18 degrees centigrade with sufficient pressure. In fact, you might find a liquid layer at the bottom of the ice with a clathrate layer below the liquid layer.
Of course, if the bottom of the glacier is moving horizontally at locations where there is permafrost, it will be scraping off the permafrost layer and carrying it toward the edge of the ice sheet with it's entrained load of clathrates. If the ice sheet is frozen to the base and is only moving laterally due to the middle being squeezed out, the clathrates will only be released when that part of the glacier melts.
A last contributor to sudden melt down and release of Carbon dioxide is Moulins. Moulins are vertical shafts which are caused by melt water pouring down fissures in continental glaciers. At present this phenomenon is best observed on the Greenland ice sheet where there is increased melting each summer. Pools of water form on the surface of the ice and if they find a fissure, they pour down to the bottom of the ice sheet. This water has to come out somewhere and presumably it will find its way out at the edges of the ice sheet. It would be expected that it would carry with it the material from the bottom of the ice sheet. Part of this would be the clathrates that have accumulated there. At each Milankovitch nudge there would be expected to be surface melting and a wash out of some of the bottom material. If great enough, this would result in a run away green house effect. It should be possible to detect methane where streams come out under continental ice sheets.
All the above scenarios depend on the supposition that clathrates will accumulate under continental glaciers ready to be released when the glacier melts. The longer the glacier exists, the greater the accumulation should be. If this is indeed happening, it should be observable under our two remaining continental glaciers on Antartica and Greenland and even possibly Iceland.
*At a pinch, carbon dating can go back 50,000 years. Hence it would be perfectly useful for dating objects from the end of the last glacial but wouldn't extend back to the previous interglacial which was 125,000 years ago.
In summary
The older and thicker an ice sheet, the more unstable it should be. This may be due to:
1) Higher temperatures at the bottom of thicker ice sheets than more shallow ice sheets due to geothermal heat being insulated from escape due to the insulating properties of the ice. At a sufficient thickness a layer of water at the bottom of the ice sheet would accelerate its flow outward.
2) A greater accumulation of carbon in the form of clathrates the longer a glacial lasts and hence the larger available green house effect if the ice sheet starts to melt.
3) A greater speed of spread at the edges of an ice sheet, the deeper the ice is, pushing ice into geographical areas where it will melt. If this ice has got a bottom layer of clathrate, this will be entering the environment. Above some critical amount of carbon added to the atmosphere, a run away green house effect would occur. A nudge by the Milankovitch cycle would release more methane from a thick ice sheet than a shallower one.
4) Outwash of bottom material by surface melt and Moulins at each Milankovitch nudge. The longer the ice sheet exists, the more carbon there should be available to be washed out.
Note that methane is often quoted to be 20 to 30 times as effective a greenhouse gas as carbon dioxide. This only holds on a 100 year basis. Methane oxidizes in the atmosphere to Carbon dioxide with a half life of about 8 years. If we look at the effect of, say, a cubic meter of methane over 100 years and calculate how much warming it will cause, it will cause 20 to30 times as much warming as a similar amount of CO2. However, the actual strength of methane as a greenhouse gas is more like a hundred times as much as Carbon dioxide. This is only important if methane is being introduced into the atmosphere in very large quantities (as seems to be the case now).
PS (Dec24, 2012)
A recent paper by German authors has shown that volcanic activity increases following strong ice melt. This would also go some way to explaining the increase in carbon dioxide in the atmosphere following, rather than before ice melt.
As suggested in a previous blog it seems unlikely that some sudden source of Carbon dioxide would occur exactly in sinc with the Milankovitch cycle. As odd as it seems it is more likely that the rise in CO2 is somehow the result of the melting. Of course, once sufficient CO2 is released, a run-away melting will occur. One likely positive feed back (warming causing more warming) is the ability of the oceans to hold less Carbon dioxide when they are warm than when they are cold. As warming starts, presumably as a result of obliquity, the oceans could give out Carbon dioxide or at the very least, not absorb as much.
In a previous blog, I suggested that methane clathrates and carbon dioxide clathrates could accumulate under an ice sheet once it had thickened to a few hundred meters. Sources of methane and Carbon dioxide include coal measures, liquid and gaseous hydrocarbon deposits and shale beds as well as the decomposition of organic material buried by the accumulating ice. If the ice covered over an area of permafrost**, the methane stored in the permafrost could also find its way to the bottom of the ice sheet. Ice sheets cause the depression of the land by about a third of a km for every km of ice added# and this might well act as a natural 'fracting', increasing the escape of such gases from all the mentioned sources. All this carbon would be sitting there at the bottom of the ice sheet ready to be released if the continental ice sheet started to melt. If sufficient was released in a burst, the green house effect could lead to a feedback, melting more ice causing the release of more gas and causing more melting etc. This run-away greenhouse effects would only end when the ice sheet was all melted. Following the melt, the slow ever present sequestering of carbon in the various carbon sinks would continue until it was possible for the accumulation of snow to start again.
# The basaltic rock on which the continents float has a specific Gravity (SG) of just over 3. Hence a km of ice with a specific gravity of a tad under 1 would push down the continent about a third of a km.
**It is interesting to note that if snow is accumulating on an area of permafrost, the snow will insulate the underlying ground. The 0 degree contour at the bottom of the permafrost will move upward as geothermal heat melts it. Eventually, as the ice sheet deepens, all the permafrost, often rich in organic soil and methane clathrate, will be melted. The permafrost undergoing anaerobic break down and any already stored up methane* would be available to combine with the bottom layer of ice and form methane clathrate.
In this blog, I would like to suggest mechanisms which would explain why every nudge from the Milankovitch cycle does not end an ice age. Lets do a mind exercise.
Consider for simplicity a large continent like Australia. It is shaped like a hockey puck, flat on top with very little slope in any direction. Enough carbon has left the atmosphere and become sequestered in sinks for some snow to last through the summer. The process of going into a glacial is gradual. The accumulation of snow can only proceed at the rate of precipitation in the area of accumulation minus sublimation and melting. The process is somewhat accelerated by the albedo effect. When a significant area is covered with white snow, incident light is mostly reflected back into space increasing the cooling.
Snow occupies about 10 times as much volume as an equivalent weight of water but as the snow deepens, the weight of overlying snow on the bottom layers increases and air is squeezed out. By the time there is a hundred or so meters of snow, the bottom layers have been squeezed into ice with some inclusions of air. The ice at the bottom occupies about 10% more volume than an equivalent weight of water.
When the ice is only a few hundred meters thick it just sits there getting deeper and deeper. The land is flat so it doesn't move down hill and there isn't enough pressure yet to squeeze ice outward. At about 300m depth, there is enough pressure at the bottom of the ice layer for clathrates to form. Any methane or Carbon dioxide which is coming from the underlying land combines with the ice and is trapped.
When the ice has reached a km or so in depth, the pressure is great enough that ice begins to be squeezed toward the edges. Right in the middle of our hockey puck continent, there is no motion with respect to the underlying land. As you go toward the edges, the motion is faster and faster. Fast is all relative. Even in glaciated continents such as Greenland with 3 or so km of ice at the center, the motion at the edges is only a few to a few tens of meters per year. There are some individual glaciers which carry ice from the interior which are moving as much as 12km per year but this is down specific valleys and not along the entire perimeter of the glacier. Averaged over the years, ice can only fall off the edges at the rate that it accumulates on top. Since continental glaciers reach depths of at least 3km, clearly, less ice was expelled than has was accumulated over the formation of the 3km deep ice sheet. At some point, as the ice thickens, the rate of loss of ice will equal the rate of accumulation. It is likely that toward the middle of the continent, the ice doesn't move at the bottom relative to the land but is rather squeezed out of the middle layers of the ice. Toward the edges, ice would be moving over the ground.
Each Milankovitch nudge will probably result in some melting. If it is correct that clathrates have been collecting at the bottom of the ice sheet, this will cause an increase in the output of green house gases and the thicker the ice the faster this might occur due to faster rate of spread caused by the thicker ice. Also, the faster the ice is moving, the further into a melting climate the ice will be pushed. This may be the explanation for the start of an interglacial only every few Milankovitch nudges. Presumably a certain amount of green house gas is necessary to initiate a run away melting. The older an ice sheet, the more clathrate could accumulate at the bottom and the thicker the ice sheet, the faster its borders are moving outward. Therefore, the older and thicker an ice sheet and the more clathrate it has accumulated, the greater the chance of a run away melting when an obliquity nudge occurs.
As a further factor, with a thick ice sheet, the glacier at the edge will be moving across the ground and expelling clathrate. It would be expected that the concentration of clathrate would increase as you go toward the centre of an ice sheet. As the ice sheet edge melts back, more and more clathrate breaks down into the atmosphere. The thicker the ice, the more clathrate is likely to be stored under the ice and the faster the ice is moving at the edges. Thick ice should be much less stable than thin ice.
Another factor which might be relevant is the heat coming out of the earth. Although it varies widely from location to location, the temperature increases as you go down into the earth at about 25 degrees C per km. Put a layer of ice on the ground and this heat has to work it's way up to the surface of the ice. I haven't been able to find the factor for heat transmission in rock and in ice in order to compare them but for the sake of the argument let's say it is the same. Let's also assume that the average temperature at the top of the ice sheet is -50
The result is that with over 2km of ice depth and given some time to reach equilibrium, there should be water at the bottom of the ice sheet. Here we run into another wee codicil. If, as was suggested in a previous blog, methane clathrate has accumulated at the bottom of the ice sheet, it can stay frozen up to 18 degrees centigrade with sufficient pressure. Whatever the actual case, the general principle is that with a sufficiently deep ice sheet, the bottom layer should be melting. This may be another part of the explanation as to why only every three or so nudges by the Milankovitch cycle sets off an interglacial period. A sufficient depth of ice has to first collect to cause heat from the earth to liquefy its bottom and increase its horizontal movement on this lubricating layer . So what sort of evidence would support this hypothesis.
Test
A) at the bottom of present ice sheets, the temperature should be around 0 degrees.
B) There should be lakes below deep ice sheets where the topography allows.
C) It should be possible in some locations at least, to detect methane and possibly Carbon dioxide being evolved from the edges of ice sheets where they are melting.
D) Where ice sheets exit into the ocean and are at least 30m above sea level (and hence their base is 300m below sea level) there may be clathrates on the bottom layers in some locations.
E) Since the carbon released from the bottom of a 100,000 year ice sheet would be "old carbon" (in other words, carbon depleted in C14) There should be a dating anomaly from the end of the last ice period, 11,000 years ago*. This sudden influx of old carbon into the atmosphere should make wood, growing after the melting, look older. For wood from about 11,000 years ago, one might see successive growth rings from a tree looking older and older despite the fact that they each successive growth ring is younger than the previous one. A place to find suitable wood might be in tropical swamps where a log had sunk into the anaerobic mud or high mountains where trees such as the Bristle Cone Pine exist.
F) if the bottom layer of an ice sheet is composed of clathrates, you might find that even though a core found solid ice right to the bottom of the core, the temperature could be above 0 degrees. Clathrates can exist up to 18 degrees centigrade with sufficient pressure. In fact, you might find a liquid layer at the bottom of the ice with a clathrate layer below the liquid layer.
Of course, if the bottom of the glacier is moving horizontally at locations where there is permafrost, it will be scraping off the permafrost layer and carrying it toward the edge of the ice sheet with it's entrained load of clathrates. If the ice sheet is frozen to the base and is only moving laterally due to the middle being squeezed out, the clathrates will only be released when that part of the glacier melts.
A last contributor to sudden melt down and release of Carbon dioxide is Moulins. Moulins are vertical shafts which are caused by melt water pouring down fissures in continental glaciers. At present this phenomenon is best observed on the Greenland ice sheet where there is increased melting each summer. Pools of water form on the surface of the ice and if they find a fissure, they pour down to the bottom of the ice sheet. This water has to come out somewhere and presumably it will find its way out at the edges of the ice sheet. It would be expected that it would carry with it the material from the bottom of the ice sheet. Part of this would be the clathrates that have accumulated there. At each Milankovitch nudge there would be expected to be surface melting and a wash out of some of the bottom material. If great enough, this would result in a run away green house effect. It should be possible to detect methane where streams come out under continental ice sheets.
All the above scenarios depend on the supposition that clathrates will accumulate under continental glaciers ready to be released when the glacier melts. The longer the glacier exists, the greater the accumulation should be. If this is indeed happening, it should be observable under our two remaining continental glaciers on Antartica and Greenland and even possibly Iceland.
*At a pinch, carbon dating can go back 50,000 years. Hence it would be perfectly useful for dating objects from the end of the last glacial but wouldn't extend back to the previous interglacial which was 125,000 years ago.
In summary
The older and thicker an ice sheet, the more unstable it should be. This may be due to:
1) Higher temperatures at the bottom of thicker ice sheets than more shallow ice sheets due to geothermal heat being insulated from escape due to the insulating properties of the ice. At a sufficient thickness a layer of water at the bottom of the ice sheet would accelerate its flow outward.
2) A greater accumulation of carbon in the form of clathrates the longer a glacial lasts and hence the larger available green house effect if the ice sheet starts to melt.
3) A greater speed of spread at the edges of an ice sheet, the deeper the ice is, pushing ice into geographical areas where it will melt. If this ice has got a bottom layer of clathrate, this will be entering the environment. Above some critical amount of carbon added to the atmosphere, a run away green house effect would occur. A nudge by the Milankovitch cycle would release more methane from a thick ice sheet than a shallower one.
4) Outwash of bottom material by surface melt and Moulins at each Milankovitch nudge. The longer the ice sheet exists, the more carbon there should be available to be washed out.
Note that methane is often quoted to be 20 to 30 times as effective a greenhouse gas as carbon dioxide. This only holds on a 100 year basis. Methane oxidizes in the atmosphere to Carbon dioxide with a half life of about 8 years. If we look at the effect of, say, a cubic meter of methane over 100 years and calculate how much warming it will cause, it will cause 20 to30 times as much warming as a similar amount of CO2. However, the actual strength of methane as a greenhouse gas is more like a hundred times as much as Carbon dioxide. This is only important if methane is being introduced into the atmosphere in very large quantities (as seems to be the case now).
PS (Dec24, 2012)
A recent paper by German authors has shown that volcanic activity increases following strong ice melt. This would also go some way to explaining the increase in carbon dioxide in the atmosphere following, rather than before ice melt.
No comments:
Post a Comment