aA is the amount of warming relatively to 1960-1992 SAT mean, which equals to 10.07°C.

aA is the amount of warming relatively to 1960-1992 SAT mean, which equals to 10.07°C.

differences between model and measurements and the variance of the model. If the phenomenon described by the model plays significant role in the occurrence of the signal, the dispersion of the deviations from the model will be small in comparison with the dispersion of the model. For the Tachlovice boreholes the use of the model taking into account advective disturbances explains 83-95% of the transient signal, while the use of the pure conductive model could explain not more than 27-58% of the temperature signal. In the real field situations, when no or only limited hydrological information is available, the poor coincidence of the reduced and the calculated temperature log under the pure conductive approach may imply on the presence of advective disturbances.

The estimated velocities of fluid migration are almost identical for all boreholes and indicate a slow upward movement in the discharge area of topographically induced subsurface convection. All obtained POM-values are similar and somewhat exceed 10°C.

Fig. 65. Map of the rms misfit (in Kelvin) as a function of the POM and fluid velocities for borehole Tach-I. The position of minimum corresponds to the best fit for POM = 10.34°C and v = -2.22 X 10-9m/s.

Fig. 65. Map of the rms misfit (in Kelvin) as a function of the POM and fluid velocities for borehole Tach-I. The position of minimum corresponds to the best fit for POM = 10.34°C and v = -2.22 X 10-9m/s.

Calculated POM-values for pure conductive approach are lower than those obtained for conductive/advective approach; the differences amount to 0.3-0.5 K (Table 4). Mean POM-values calculated for the Tachlovice borehole T-z data under pure conductive (POM1 = 10.01°C) and for conductive/advective (POM2 = 10.39°C) approaches are shown in Figure 64. The difference is less than 0.4K; however, as this value is comparable with the amount of twentieth century warming, it is significant for the interpretation of the obtained POM-values.

The above analysis showed high sensitivity of borehole temperatures to the POM estimate and fluid flow velocity. Even a relatively small difference in POM and/or fluid velocity caused significantly poorer fit with observed data. A well-developed minimum in the misfit function indicates the robustness of the described method. Regardless of being only a first approximation to the full problem, the method offered reasonable POM-values that could explain near 80-95% of the transient borehole temperature signal and enabled to estimate Darcy fluid velocities directly from measured temperature logs. The study suggests that even advectively distorted borehole temperature log may still contain a valuable signal, which can be used in paleoclimate reconstruction to assess the POM-values. Except for extreme cases, such as rough topography, the climate signal cannot be completely "washed out" by hydraulic advection. Bodri and Cermak (2005a) method: (1) can help to distinguish boreholes affected by advective disturbances on the base of temperature log only, when independent hydrologic information is not available, and (2) saves the possibility to include some of the previously rejected temperature logs for the paleoclimate analysis. It is especially important for the regions affected by hydraulic disturbances and/or in the regions with a

Fig. 66. Reduced temperatures (dots) compared with the best-fit transient temperature logs calculated for pure conductive and/or for conductive/advective approaches for Tachlovice boreholes (Tach-I to Tach-IV).

limited number of borehole temperature logs, where an existing datum may be valuable for the informative reconstruction of the climatic history.

The above-cited and similar works represent mainly theoretical and/or local studies of the groundwater flow problem. More comprehensive research was performed in the recent work by Ferguson et al. (2006). To provide general criteria for the screening of the temperature logs for the presence of the groundwater disturbances, the authors have examined what are the hydrological conditions that could produce significant advective disturbances and thus distort climatic signal to an unrecoverable state. The authors have modeled recharge15 area conditions, where advective perturbations are expected to be the largest. The 2-D aquifer models were constructed for the wide range of the

15Recharge area represents a land area where water reaches the zone of saturation from surface infiltration, e.g. where precipitation soaks through the ground to reach an aquifer. Generally, it is connected with the underground aquifer by a highly porous soil or rock layers. Water entering recharge area may flow for kilometers underground. On the contrary, a net annual transfer of water from the ground to the surface (e.g. to streams, lakes, wetlands) occurs in discharge area. Recharge areas are usually in topographic highs, while discharge areas are located in topographic lows.

downward fluid velocities from 2 X 10~10 to 2 X 10~8m/s. To estimate the maximum influence of the possible advective disturbance in every modeled situation, the temperature-depth profile corresponding to maximum downward water flow were chosen for further GST inversions. Applied SVD inversion procedure has assumed pure conductive heat transfer regime; thus, obtained apparent GST histories gave account on the advective disturbances only. Numerical trial calculations have shown that for a 200 m deep and 1 km long basin the groundwater flow can produce a significant perturbation (0.3 K or more) to the GST signal only when the Darcy velocity exceeds 3.3 X 10~9m/s. Higher velocities were necessary to produce similar perturbation at shallower and/or longer basins. Thus, for a 100 m deep basin of the same length the fluid velocity of more than 2X10~8m/s was required. In all cases advectively induced perturbations were significantly lower than a typical climatic signal. The 1.5-3 times larger Darcy velocities would be required to produce an apparent GST history comparable in the order of magnitude with generally observable climatic trends. Because (1) recharge area is generally connected with the underground aquifer by highly porous soil or rock stratum and (2) a net infiltration of rainwater depends on the precipitation amount, Ferguson et al. (2006) have also studied an influence of the hydrogeological parameters of the medium and meteorological conditions on the groundwater disturbances and have concluded that advection will have detectable effect on the GST history only in (1) moderately to highly permeable (k> 10"3 darcy)16 aquifers with (2) high depth to length ratio that (3) can be also characterized by significant amount of precipitation (>2000mm/year).17 Borehole temperature log can be rejected from the GST reconstruction when all three conditions are fulfilled. Their inspection of available database has shown that available temperature profiles were measured in the environments well below these parameters. Moreover, according to the survey by Ferguson et al. (2006) the coincidence of all three conditions occurs only rarely. The above study has indicated that advective disturbances unlikely represent serious problem for the majority of the GST reconstructions that have used the International Heat Flow Commission (IHFC) global database (

2.8 Climate Change and Permafrost

In Section 2.6.2 we described the GST-SAT coupling in the case of occasionally frozen soils and have concluded that on the long timescales the effect of soil freeze/thaw cycles on the GST-SAT decoupling is not so important. In the most of the land mass of the Northern Hemisphere actual freezing and their occurrence appears to be quite sporadic, and its effect vanishes during averaging over large temporal/spatial scales and likely cannot create a false systematic secular trend in the GST. What about perennially frozen ground, so-called permafrost? It is frequent at high latitudes and was even more widespread during the last glaciation. The thermal effect of the stored and released latent heat may influence subsurface thermal field. In what way GST-SAT coupling occurs in such environments? Can GST changes be inferred from temperature logs measured in

16The permeability of a consolidated rock equals to (2.5-3.0) X 10~3 darcy (Fetter, 1988).

17Average annual precipitation amounts to, e.g. 1190mm (New York, NY), 818mm (Toronto, Canada), 752mm

boreholes drilled in the deep permafrost? Twenty years ago, Lachenbruch and Marshall (1986) have aroused an interest of the scientific community by highlighting the potential of permafrost borehole temperature data to be used for the climate change reconstruction.

Term permafrost means the soil that remains in frozen state for more than two years. Nowadays permafrost conditions characterize about 25% of the land in the Northern Hemisphere (together with discontinuous permafrost and/or glacial ice), including vast regions of Canada, Russia, and Alaska. Most permafrost is located at high latitudes, but there is also alpine permafrost and other small permafrost areas at high altitudes in the mountain chains in both the Northern and Southern Hemisphere. Permafrost was more widespread during the past continental glaciation episodes. Evidence of former existence of permafrost has been found in the regions of North America and Eurasia that are far away from the present permafrost boundary.

The physics of the permafrost phenomenon is as follows. At high-latitude regions where the mean annual temperature is below 0°C, some of the ground frozen in the winter does not completely thaw in the summer. The thickness of the permafrost layer is controlled by the energy balance at the soil-air surface and the geothermal gradient (Lachenbruch et al., 1988; Smith and Riseborough, 1996). Annual fluctuation of the air temperature from winter to summer is reflected in an attenuated manner in the subsurface (for details see Section 1.3 and Figures 14 and 15). As the surface temperature signal propagates downward, it is delayed in time and its amplitude decreases exponentially with depth due to the diffusive process of heat conduction. Each variation vanishes over a vertical distance related to the period of change and to the thermal diffusivity of the ground. Shorter period fluctuations attenuate more rapidly. The daily temperature wave is practically not observable below 1m depth (Figure 15, Chapter 1). Similarly, annual GST oscillations are detectable at the 10-15m deep uppermost layer. This interval is known as the active layer. It thaws during summer and generally overlays permafrost (Figure 67). Thickness of the active layer varies by time and location. The thermal regime of the seasonally thawed active layer is highly complex, owing to non-conductive heat transfer processes that operate much of the year (Hinkel et al., 1997). Permafrost represents zone just below the active layer that tends to be ice-rich. The annual temperature variations are not visible in this layer and the temperature gradient corresponding to the outflow of heat from the Earth's interior becomes discernible. If the permafrost is in the thermal equilibrium, the temperature at the level where annual amplitude vanishes is generally regarded as the minimum temperature of the permafrost. It can vary from close to 0°C at the southern rim of the permafrost area to — 10°C in northern Alaska (Hinkel et al., 2001) and — 13°C in northeastern Siberia (Pavlov, 1994,1996).

The state of permafrost depends strongly on climate. Global climate models predict that the greatest increase in the average annual SAT will occur at high latitudes over 50°N for the few next decades as anthropogenic carbon builds up in the atmosphere (e.g. Woo et al., 1992; Flato et al., 2000). Most simulations suggest that the warming should be enhanced due to ice-albedo feedback mechanisms.18 Because of expected serious impact

18Ice (and snow) can have very high albedo. It reflects more than 80% of the incident sunlight, while only some 20% is absorbed. If there is climate change that increases the temperature of the surface it will cause melting that lowers the albedo. It results in more absorbed sunlight increasing melting and lowering albedo even more. It is a straightforward positive feedback mechanism. The net result of such feedback could be a rapid increase of the temperature of the Earth's surface.

Diseases Permafrost
Fig. 67. Permafrost thermal conditions in winter.

of the global warming on the circumpolar regions, this research is of high priority for the global change research community in the recent time. Long-term monitoring of the permafrost began in the last some decades. In 1967, the Greenlandic Geological Survey (GEUS) have started a monitoring of the permafrost temperatures in borehole from Ilulissat (West Greenland), where 21 thermistors have recorded ground temperatures from 0.25 to 15m depth (Van Tatenhove and Olesen, 1994; fgdc/ggd631_boreholes_greenland; see also Figure 68). The monitoring experiment continued up to 1982. From the mid-1960s to early 1990s the Geological Survey of Canada (GSC) performed regular ground temperature and permafrost measurements at near 40 deep wells previously used for hydrocarbon exploration at the Canadian Arctic Archipelago to north of 75°N latitude. Since 1990 the GSC has established a set of instrumented sites and boreholes for permafrost thermal monitoring. This system crosses several permafrost boundaries with measurement locations in a variety of climatic and/or environmental conditions ( In Canada several groups of researchers perform measurements of permafrost temperatures that began in different times during the last 35 years. The GSC also provides the international data management for the Global Terrestrial Network for Permafrost (GTN-P) borehole temperature monitoring program (see below). The U.S. Geological Survey (USGS) plays an important role in providing field-based data on Arctic-Alaska climate monitoring

Was this article helpful?

0 0

Post a comment