Isotopic composition of 1. glacier
Glaciers are divided into two types: polar glaciers and temperate glaciers. Polar glaciers are distributed at the two poles of the earth, and the temperature of the whole ice remains below the melting point, while mild glaciers are distributed outside the poles.
Polar glaciers and temperate glaciers are generally composed of surface snow, granular snow (dense perennial snow) and ice. When the snow on the surface becomes thicker and thicker, under the action of gravity, the structure of deep snow will become more compact and the pressure of snow will increase. After a long time, these snow particles will become transparent ice. Because of the close genetic relationship between them, the isotopic composition of granular snow and ice in the same glacier is mainly controlled by the preservation state of surface snowfall and surface snow and the degree of later disturbance.
According to whether the surface snow melts or not and the degree of water infiltration into the snow layer, it can be divided into three main zones: in the process of surface snow preservation, the snow does not melt and there is no water infiltration, which is called dry snow zone; The area where the snow surface partially melts and the dissolved water penetrates into the snow layer for a certain distance and then freezes again is called the infiltration zone; Dissolved water seeps into the snow layer all year round, that is, the snow in the previous winter, and the temperature rises to 0℃ in the next summer or even the whole autumn. These areas are called soaking areas. The dry snow zone mainly appears in polar glaciers, while the saturated zone and soaking zone mainly exist in temperate glaciers.
Isotopic changes of (1) polar glaciers
Polar glaciers are located at or near the poles, and ice is stored in an environment where the temperature is lower than the melting point. The stored ice is rarely affected by melting or refreezing, and the isotope content of continuous snow layers can often be preserved in the granular ice and snow of glaciers, which often reflects the environmental characteristics of snow formation.
Figure 15- 17 ice core profile (according to Dansgaard et al., 197 1)
See Figure 15- 17 for the representative ice core data. It is taken from CampCentury near Greenland in the Arctic, and its isotope profile shows some regular changes. The δ value varies periodically with the depth of ice core in a sine wave shape. From the ice surface down to 1 15m, the wavelength (λ) of the periodic vibration wave is close to 0.35. With the increase of depth, the wavelength gradually becomes shorter and disappears at several kilometers (that is, the periodic change disappears), while the δ value changes. At this time, the large vibration wave tends to be smooth. The study of isotopic composition proves that the highest δ value represents the isotopic composition of summer snow, and the lowest δ value (amplitude bottom) represents the isotopic composition of winter snow. The maximum difference of δ value between summer and winter is about 20, which indicates that there is isotopic seasonal effect in glaciers. The difference of δ values in similar years generally reflects the difference of climatic conditions when it snows every year. The fluctuation of δ value in the same year (one vibration wave period) is more complicated, because different environmental conditions will cause the fluctuation of δ value during snowfall. If the δ value of snow changes little in summer and winter (that is, there is no difference in environmental conditions), or the isotope effect increases after the snow-covered period, the δ value will lose its regularity or even disappear. This situation is more obvious in temperate glaciers. The overall trend of Byrd ice core in Antarctica is similar, but the range and period of δ value sometimes appear confusion. Without regularity, it is difficult to explain seasonal changes.
Epstein and Sharp( 1959) first pointed out that in the dry snow zone of polar glaciers, the infiltration of snow-soluble water is occasionally seen in a very small infiltration zone, and most of the seasonal variation of δ can be retained in granular snow and glacier ice. In the infiltration zone, especially in the soaking zone, the seasonal variation amplitude of δ will be obviously weakened.
Dansgaard et al. (1973) published the data of three observatories in Greenland, Dye Island, Crete and CampCentury, and one observatory in Antarctica. The data of three stations in Greenland show that the seasonal variation of δ remains unchanged during the whole snowfall and lasts for thousands of years. However, the seasonal variation of Antarctic Byrd value disappeared after 3 years.
Due to the special environment of polar glaciers, most of the snow is preserved in dry snow areas. So there are two special cases: one is the drift of snow, which makes the seasonal variation of δ in snow unrecognizable; Second, the snow has not drifted, which will accumulate to form snow, and the seasonal information of its δ value will be preserved for a long time.
In polar glaciers, the seasonal variation range of δ value in the process of granular snow changes greatly. With the passage of time, the accumulation of granular snow deepens and isotopes become uniform. Because there is almost no dissolved snow water between the granular snow in polar glaciers, and there is almost no refreezing of the sub-layer dissolved water, there is no isotope exchange between the dissolved water and the granular snow. Therefore, the change of δ value of polar glaciers can only be realized by molecular diffusion in gas phase or solid ice. These later isotopic effects are so slow that the seasonal δ values stored in the materials piled up in the form of granular snow are not significantly affected.
Johnson( 1977) and others made a detailed study on the granular snow in the dry snow area of polar glaciers. He believes that polar snow may have two isotope homogenization effects: first, the air pressure change at the air interface above the glacier causes the particle snow vapor to move vertically upward; The other is the diffusion of gas phase in snow layer. These effects will cause the vertical exchange of granular snow. The former is extremely active in the upper part of snow, while the latter mainly exists in the critical depth of granular snow. There, snow particles will become very dense, and the isotope exchange rate through the gas phase is very slow below the critical depth, but this exchange continues until the pores are completely closed, and then only the slow diffusion of molecules in solid ice exists. In this depth of ice, the δ value can be stably preserved for thousands of years. The cumulative amount of Greenland ice sheet is 24 g cm-2 a-1,and the seasonal variation of δ value tends to remain unchanged during the critical depth of granular snow. In Antarctica, snow drifts due to local storms, and the accumulation speed of snow there should reach at least 34 g cm-2 a-1,so information that changes with seasons can be retained, but it is often not so ideal in practice. For example, in the warm season, a little melting of the earth's surface may form an ice layer in the granular snow, thus closing the upward passage of water vapor in the lower granular snow, thus hindering the material exchange caused by the vertical movement of air. Johnson et al. (1972) also mentioned the above situation when analyzing the seasonal variation of the δ value of the ice core at the depth of 776m in CampCentury.
In the solid ice of polar glaciers, molecular diffusion is the only factor that affects the change of ice δ, and vertical diffusion can be described by the following equation:
isotope geochemistry
Where: d is the diffusion coefficient; X is the depth (thickness) of the ice. If a sine wave with wavelength λ is used to represent the period of seasonal variation of average δ value, the variation of maximum amplitude Q will decrease exponentially with time due to diffusion, just like damping vibration. For a diffusion equation with time constant t, it is
isotope geochemistry
With d = 2.6×10-12cm2 s-1(at-24℃), according to the seasonal variation period λ=38cm of the δ value of the uppermost part of the Kemp ice core, assuming that the ice layer does not thin all year round, t is about 450,000 years. It can be seen that the molecular diffusion in solid ice is very slow.
When the ice layer becomes thinner with the increase of depth and pressure, diffusion will have a great influence on the seasonal variation of δ value with gentle fluctuation, and even make the δ value signal completely disappear in deep places. For example, the annual layer at the depth of Century Ying Ice Core 100m is reduced to about 5cm, the temperature is -20℃, and t is 5000 years. The experiment of Johnson et al. (1972) also shows that the seasonal variation of the δ value in the ice center of Camp Century is basically consistent with the experimental results at the depth of 1 100 m ... The annual ice layer at this depth is reduced from 38 cm to 4 cm.
For the Camp Century Ice Center, Johnson found that it takes 5000 years, 8000 years and 10%, 50% and 90% to reduce the seasonal thickness of granular snow, which is quite consistent with the experimental data.
Johnson( 1977) also calculated the molecular diffusion, the thinning of ice layer with depth and the dependence on temperature, and obtained the results. In most cases, this result can be used to calculate the decrease of isotope δ value in polar ice.
To sum up, with the increase of time, the δ value of polar glacier snow changes greatly. These changes can be long-term, seasonal or even short-term. Due to the drift or melting of snow, the top layer of snow can change δ, and its annual isotope distribution characteristics will be strongly disturbed. In granular snow, the fluctuation amplitude of δ value will decrease strongly. This is caused by the mass exchange of steam between layers. The change of δ value in a year often disappears quickly. Under certain conditions, although the seasonal variation of δ is strongly weakened, the δ information retained in granular snow can still be traced back to thousands of years in the ice core. The seasonal variation intensity of snow δ value is negatively correlated with snow speed. Molecular diffusion is the only way of glacier isotope exchange, which is extremely slow, making the δ information in the ice layer last for a long time.
(2) Isotopic changes of temperate glaciers
Mild glaciers are distributed outside the poles, and their temperature has been kept within the pressure melting point. Storage snow is often affected by melting and water infiltration, especially at the end of winter and just after the melting season, its δ will be very strongly disturbed.
In some areas, such as the Alps in North America and Europe, glacier granular snow is rich in heavy isotopes, which may be related to the infiltration of summer precipitation and the capture of summer precipitation in the cracks of granular snow (Sharp et al., 1960). Iceland's glaciers are different. Due to the mild maritime climate, there is almost no difference in isotopic composition between winter snow and summer precipitation, and its isotopic enrichment can only be attributed to isotopic exchange between solid and liquid phases. Table 15-6 lists three isotopic equilibrium fractionation coefficient values between ice and water. As can be seen from the table, in the solid-liquid system, during isotope exchange, solid is richer in heavy isotopes than liquid.
Table 15-6 D and 18O isotopic equilibrium fractionation coefficients of ice and water
Figure 15- 18 Iceland Vatnaj? Relationship between δD and depth of shallow borehole in kall Glacier (according to Dnsgaaard et al., 1974)
The figure 15- 18 shows Vatnaj in Iceland. Correlation between δD and depth of three shallow holes in Karl Glacier. Samples were collected before the snow began to melt in spring. This section is the most typical temperate glacier. The seasonal variation of δ is very small, and the average isotopic composition of precipitation in summer is similar to that of snowfall in winter. V- 1, V- 10 and V- 18 are located at 1300, 1400 and 2000m, respectively, and their isotopic uniformity largely depends on the runoff ratio, that is, on the amount of infiltration water in the snow.
According to the data of Icelandic Meteorological Bureau, the annual average precipitation at V- 1 is estimated to be 300 g cm-2 a-1. The density measurement of granular snow shows that about 50% of precipitation is lost in the form of runoff in summer. At V- 18, the amount of dissolved water is very small, which does not affect the original δ value of stored snow.
The δ values of V- 1, V- 10, and V- 18 in the surface layer of snow in recent winter are the same as those of snowfall. However, in the snow before summer, the V- 1 profile is affected, and the preserved granular snow is obviously rich in heavy isotopes, even more abundant in deuterium than summer precipitation. The reason for this phenomenon is that the snow stored in the snow water system has undergone recrystallization and isotope exchange. The ice ratio D, which is in isotopic equilibrium with water, is about 20‰ richer than that of 18O, which is about 3.0‰.
During the snow cover period, the evaporation and condensation of the surface will also change the isotopic composition of the surface snow. Moser and Sti-Chler( 1970, 1974) observed that the evaporation of surface snow would be rich in D and 18O even at 0℃. Because enrichment only occurs in the surface layer, it will not change the overall isotope content of stored snow. In different summer snow layers, the homogenization degree of isotopes is different, and the homogenization degree in the previous summer is deeper than that in the next summer. This situation is particularly obvious at V- 1
In V- 18, there is little dissolved water, which has no obvious effect on the deuterium change in stored snow, and the δ value change is obviously stored in granular snow with a depth of about 20 m. This depth is equivalent to six years of snow.
Fig. 15- 19 experimental curve of snow melting amount and δD (according to Moser et al., 1980)
In order to quantitatively study the isotope exchange of snow particles in temperate glaciers, Arnason et al. (1969) conducted a simulation experiment. Using a thick and long snow column, the dissolved water slowly melts from the top of the wave, then gradually passes through the snow column, and finally is discharged from the bottom of the snow column, and the deuterium content of the discharged water is measured. It is observed that there is isotope exchange between snow crystal and water, and a linear equation describing this exchange quantitatively is derived, that is, δ d = ax+b, where δ d is the instantaneous value of snow; X is the proportion of snow dissolved in the total snow; B is the original δD value when the snow does not melt; A is the change rate of Δ Δ d Δ x, and the results of experiment and simulation are very consistent (Figure 15- 19). The figure shows that in the early stage of snow column melting, water did not completely penetrate the snow column, but flowed down from the inner wall of the glass column, resulting in errors in the measurement of snow melting. Under natural conditions, the snow and precipitation in Fluhioch, Switzerland were measured by a snow melting instrument, and the obtained isotope data were compared with the experimental results, which confirmed that the experiment was successful.
In temperate glaciers, the molecular diffusion of solid ice plays no important role in isotope homogenization. The annual accumulation of temperate glaciers is often much stronger than that of polar glaciers, and quite thick ice layers are accumulated every year. But the ice at the bottom of temperate glaciers is much younger than that at the bottom of polar glaciers. There may be other effects that affect the change of δ value.
In Vatna? In a 450-meter-deep borehole of kall Glacier, the ice from 20 meters below the surface to the bottom is within the pressure melting point, and the deep part may also contain water. When the ice near the surface sinks to a depth of 500 meters, some water will be produced due to the increase of pressure and the decrease of melting point, and there will be various stresses, especially on the sliding surface, and a certain water level will also be produced. This water can pass through the whole ice body, causing some substances to migrate and changing the δ value.
Yes, Vatna, Iceland? The crystal size and chloride content of the 4 15m deep ice core of Karl Glacier were studied. The results show that the whole ice body has undergone continuous infiltration and isotope exchange (similar to the snow at the top of the infiltration zone and soaking zone). With the increase of depth, the ice crystals grow gradually, so recrystallization runs through the whole ice core.
The measurement of chloride in ice core shows that the chloride content in ice crystals gradually decreases from 1.0μg/g on the surface to 0. 1μg/g at the depth of 4 15m when the chloride changes gradually. It is estimated that chloride may dissolve in the ice layer between ice crystals, accompanied by water escaping from the ice. It has been found that this effect may affect the preservation of δ value information by residual granular snow.
According to the changing characteristics of isotopic composition of temperate glaciers, it is not suitable to study various environmental parameters related to atmospheric precipitation, even in the infiltration zone with little interference, we should be especially careful. In the period of granular snow, when light snow turns into coarse snow and finally becomes transparent ice, this process will be recrystallized, which is very important for understanding the formation, development and characteristics of mild glaciers.
2. Study on glacier isotope and paleoenvironment.
(1) Ice and snow accumulation speed
In granular ice and snow, the area where the seasonal variation information of δ value is kept. The δ profile of ice and snow can be used to determine the accumulation speed. The determination method is simple. As long as the number of summer δ values is counted from the surface layer and the density of each layer of granular snow is measured, the annual accumulation can be calculated.
However, this method has great limitations. It is limited in areas disturbed by snowdrifts and in temperate glaciers where the seasonal variation of δ value is submerged by homogenization or the leakage of dissolved water is dominant. Even the glacier zone that meets the conditions of accumulation speed is limited to the upper part. In the deep, various natural forces, especially vertical pressure, have different degrees of influence on the annual layer thickness. The degree of influence depends on the changing history of glacier temperature, the original thickness of granular snow and the slope of ice layer, so the situation is quite complicated.
Nevertheless, in many parts of Greenland and some coastal areas of Antarctica, especially in the upper part of glaciers, satisfactory and credible results can still be obtained by using isotopes to determine the accumulation speed.
(2) Bing Xin Dating
Dating has always been an important topic in the study of paleoenvironment. Without time-scale environmental data, it completely loses its significance and value.
The basis of ice core δ 18O dating is that the summer precipitation δ 18O value is high, the winter precipitation δ 18O value is low, and the winter and summer cycles repeatedly, forming the cycle characteristics of ice core δ18o. Ice core dating is based on calculating the maximum δ value in summer, but it is only effective when there is a significant difference between the δ value in summer and the δ value in winter. For the δ 18O value of ice core, the difference is not less than 2, and for d, the difference is not less than 20, which meets the age resolution. Therefore, ice core dating is especially suitable for high polar glaciers, because the seasonal variation of δ value of high polar glaciers can be preserved for thousands of years. From the point of view of meteorology and climatology, this method depends on the transportation and condensation of water in the ocean atmosphere, so it is especially suitable for areas with large seasonal temperature changes. Antarctic ice sheet is an ideal area for δ 18O dating because of its high latitude and large seasonal temperature difference. Although the δ 18O seasonal information dating is very effective, there are some preconditions for the application of this method: First, δ 18O dating is suitable for areas with high ice and snow accumulation; Secondly, δ 18O dating is only applicable to the ice-covered area where the wind is light and the snowdrift is undisturbed, and the δ 18O seasonal signal can be preserved.
According to the research of Dansgaare, Johnson( 1969, 1972, 1973) and Hammer et al. (1978), the age (t) of Bing Xin can be calculated by the following formula.
isotope geochemistry
Where: τ represents one year; H is the total thickness of the ice layer; Y is the vertical distance between the ice layer and the bottom every year; λ is the average annual thickness, which can be obtained from the annual variation period of ice core δ 18O.
The determination of ice core age can also be obtained by counting the maximum peak value of δ in each summer from the table. Using this method, two ice cores of Greenland glacier were measured: Dyez, the ice core with a depth of 340 1m was formed in A.D.1233 ~1971ad; Grete, an ice core with a depth of 404m, was formed in 1 177 ~ 1973. The error of these ages is within a few thousandths.
As mentioned earlier, this method is only applicable to polar glaciers, not temperate glaciers. Deutsch( 1966) and others tried to date alpine glaciers by this method, but failed, although they found that there was a great difference in rainfall δ between winter and summer. But there are exceptions For example, Koerner et al. (1973) dated the Meighen ice cap in the Canadian Arctic and determined that it was formed in the late Wisconsin because of the high content of 18O there.
(3) Ice core isotopes and paleoclimate
Temperature is the main factor that restricts the isotopic composition of precipitation, which is reflected by the seasonal variation of isotopes. In other words, the change information of isotope δ value stored in the ice core profile can reflect the changes of climate and environment in different periods in the past to some extent.
The above problems were first put forward by Dansgaard( 1954), and were extensively studied, and valuable enlightenment was obtained.
The most representative data is the ice core profile with the depth of 1390m in the Camp Century of Greenland (Figure 15-20). On the profile, the time scale is thousand years, and the variation pattern of δ value on the profile reflects the climatic conditions in cold and mild periods in detail. The low δ value is in the cold period and the high δ value is in the mild period. The δ value of ice core profile is "["distribution, corresponding to the ice age in Wisconsin. Because Wisconsin was in a cold climate throughout the ice age, its δ value was generally low. However, the low δ value in the profile is not entirely caused by climatic factors, and there may be superimposed effects such as the increase of ice thickness caused by low temperature, the height effect of precipitation and the lower temperature change trend. The dotted line in the figure is a rough correction considering the above situation.
Due to the large age span of the deeper ice layer in the lower part, the relative amplitude of annual δ seasonal variation of some profiles becomes smaller, and some residual signals disappear, so it is impossible to date by δ value in ice. The deepest ice core in Greenland may have recorded a climate exceeding 1Ma. If the precise dating of deep ice cores can be solved, the profile data of 18O obtained by drilling ice cores on the seabed can be supplemented to Emiliani Ice Age.
Another deep ice core was obtained in Antarctic ice drilling. The total length of the ice core is 2000 meters, and the content of 18O has been measured in the whole ice core. However, due to the slow accumulation speed and the lack of seasonal variation of δ value, it is difficult to date and a suitable model has not been established. But the general trend of climate change shows that the northern and southern hemispheres are almost the same.
The significance of these studies is that we can trace the climate change in the history of the earth and use this information to predict the future. Johnson et al. (1970) predicted the climate in the next 39-50 years by using the δ fluctuation of ice core profiles in 78 years and 180 years. If human activities (such as pollution) are not considered, the temperature will rise immediately after 20 10 ~ 2020, and the next 10 ~ 2020.
Vatna, Iceland? Kall glacier belongs to temperate glacier type, and the ice core with a length of 4 15m is obtained from the highest point. By studying it, we can provide short-term climate change information. Detailed deuterium measurements in the upper part of the ice core show that the δ value of snow is only slightly affected by the infiltration of dissolved water. At the depth of 20m, consistent with 1963, the annual variation of δ value is almost gentle, and the average δ value has no obvious change. Comparing the δ value information recorded before 1960 with the meteorological data in Iceland, we can see that in some years before 1920 and 1948 ~ 1952, the δ value is also low when the average temperature is relatively low. Although its time span is short, it may provide a record of the climate in the North Atlantic.
Due to the lack of obvious seasonal variation of temperate glaciers, δ value can not be used for isotopic dating. However, the situation of Icelandic glaciers is quite special, because there are more than 50 meters of pyroclastic rocks in the 4 15m ice core, and the time of volcanic eruption is known, so it may be helpful for dating and providing short-term climate information.
(4) Glacier isotope and ice flow model
Ice is a plastic substance, which will flow down the mountain under the action of gravity and other forces. Reid( 1896) first proposed the normal glacier flow model. According to the envisaged model, the snow deposited in the accumulation area sinks downwards and flows down the slope. As a result of the flow, the ice flow terminal in the ablation area will become the water on the lowest surface of the ice layer (Figure 15-2 1).
Figure 15-20 Paleoclimate Change Reflected by δ Value of Kemp Century Ice Center (according to Dnsgaaard et al., 1974).
Figure 15-2 1 schematic diagram of ice flow pattern (according to Reid, 1896)
On the longitudinal section of the glacier on the normal granular snow line, the change of δ value should also follow the above assumed law, that is, the δ value of snow deposited at the highest point of the glacier is the lowest due to the isotope height effect, so when it sinks and flows to the end of the ice flow in the ablation area, the δ value of ice is the lowest. On the other hand, the residual snow is richer in heavy isotopes than the original accumulated snow, and its homogenization and enrichment degree depend on the infiltration of melted water in the snow storage layer. In the lower accumulation area, the snow melts and takes away more water, and the granular snow there will be richer in heavy isotopes than the granular snow in the high place. Therefore, the homogenization process of accumulation area tends to increase the δ gradient. Below the snow line, the change of δ value profile is just the opposite, because the lower δ value should be close to the end point because of the flow of ice. On the cross section, the δ value increases gradually from the peak to the edge, but it should decrease with the increase of depth when drilling through the ice layer. In this way, it is possible to establish the corresponding ice flow model by isotope δ value profile method.
Dansgaand( 196 1) traced the formation site of the iceberg 1 1 in West Greenland by isotope method. According to the change of the content of 18O, he found that the iceberg came from the ice sheet at the height of 60 ~ 460 km in the inland area.
In addition, climate change may change the isotopic composition of the accumulation area in a certain area, and sometimes it is difficult to show the characteristics of isotopic ice flow model, which should be paid attention to in application.