C12A-01 INVITED 10:25h
Spatially Distributed Permafrost Models: Current Status, Problems and Needs
Permafrost is a central element of the cryospheric system. Its importance has become increasingly recognized in both scientific literature and popular media, especially in the last few years. The past two decades have seen a dramatic rise in the number of permafrost models used to evaluate permafrost parameters over geographic space, as well as spatial changes in permafrost-related phenomena that may follow from global climate change. Despite the importance of its roles in the geological, ecological, engineering, and climate-change sciences, modeling of permafrost has, for the most part, remained the domain of individuals and small groups of scientists, each utilizing its own methodological approach, resulting in a wide range of results. This situation has made it difficult to incorporate generated data sets and modelling products into the larger global-change research enterprise. Consequently, there has been little effort to develop an explicit hierarchy of permafrost models, to evaluate their performance using standardized validation tools and data sets, to rank the performance of various models in different applications and to explicitly link modeling results with observations. As a result, considerable uncertainties remain regarding appropriate methods for describing the ground thermal regime spatially, their accuracy, and their applicability to different scales and climatic conditions. To address these problems the Workshop on Spatially Distributed Modeling and Remote Sensing of Permafrost was held in Fairbanks, Alaska in October 2004. The main objectives of the workshop were: 1) access the current status of available permafrost models and observations and 2) develop a framework for intercomparison of spatially oriented approaches to permafrost modeling at a variety of geographical scales. This presentation provides an overview of the current status of permafrost modeling and report on recent activity aimed at intercomparison of permafrost models.
C12A-02 10:45h
Spatially Distributed Model of Permafrost Dynamics in Alaska
Given the possibility of climate warming in the near future, an evaluation of the magnitude of changes in the ground thermal regime becomes desirable for assessments of possible ecosystem responses and impacts on infrastructure in the Arctic and sub-Arctic regions. In the past, a soil model GIPL 1.0 developed at the Geophysical Institute Permafrost Lab was used to simulate the dynamics of the active layer thickness and mean annual ground temperature, both retrospectively and prognostically, using climate forcing from Global Climate Models. The GIPL 1.0 model is a quasi-transitional, spatially distributed, analytical model for the active layer thickness and mean annual ground temperature. This model is incorporated into GIS, which contains the information on geology, soils properties, vegetation, and snow distribution. GIS allows visualization of input and output parameters and their representation in the form of digital maps. As a further significant step in the GIPL model development, we replaced the analytical solution with a numerical model based on a finite difference method for the non-linear Heat Conduction Equation. In this model the process of soil freezing/thawing is occurring in accordance with the unfrozen water content curve, which is specific for each soil layer and for each geographical location. For each grid point on the map we used a one-dimensional multi-layer model of soil down to the depth of a constant geothermal heat flux (typically 500 to 1000 m). At the upper boundary, there are insulating layers of snow and vegetation that can change their properties with time. Special Enthalpy formulation of the energy conservation law makes it possible to use a coarse vertical resolution without loss of latent heat effects in phase transition zone even in case of fast temporally and spatially varying temperature fields. The new version of GIPL (GIPL 2.0) calculates soil temperature and liquid water content fields for the entire spatial domain with daily, monthly and yearly resolutions. The merge of the new GIPL and the GIS technique provides a unique opportunity to analyze spatial features of permafrost dynamics with high temporal resolution. The results of the new model application to study permafrost dynamics in Alaska will be presented.
C12A-03 11:00h
Spatial and Temporal Variations of the Annual Freezing/Thawing Index in the Northern Hemisphere
The annual freezing/thawing index (AF/TI) can be used to predict and map permafrost distribution and active layer extent, providing important information on climate variability. Reliable long-term measures of the freezing and thawing index are thus important variables for understanding and predicting high-latitude processes. The AF/TI is defined as the cumulative number of degree-days below/above $0\deg$C for a given time period. However, long-term daily air temperature measurements are not easily available and are prone to missing values which will influence the magnitude of the AF/TI. Monthly data are more easily obtained and less influenced by missing data issues. An important question regarding the use of monthly temperatures in the computation of AF/TI is how reliable the monthly approximation is. Using stations with long-term daily temperature observations at various locations in the Northern Hemisphere (NH) mid- to high-latitudes we assess the temporal reliability of AF/TI calculations based on monthly data, relative to the conventional daily computation. Gridded NH daily 2-m temperatures from the ECMWF 40+ year reanalysis (ERA-40), interpolated onto a 25x25 km grid, we assess the spatial accuracy. For both temperature products, the daily values are averaged to monthly resolution and AF/TI is computed based on daily, and monthly data. The accuracy of the monthly approximation is then assessed using a simple measure of percentage (relative) error. Based on station data, preliminary results indicate no significant changes in accuracy over time, but a latitudinal dependence of the accuracy of the monthly-based AF/TI, {\it e.g.}, the approximation works well in higher latitudes, but is less accurate south of $60\deg$N over North America. This is verified by the reanalysis temperatures, which provide a spatially continuous assessment over the NH. Freezing index is approximated well using monthly data (errors$<$10%) between the pole and $\sim$$50\deg$N over North America, and as far south as $\sim$$30\deg$N over central Asia. South of these regions and the oceans, freezing index based on monthly data is not reliable. Thawing index is approximated well over most Northern Hemisphere land areas, but not over the high-latitude oceans. Having verified regions where AF/TI calculations work well allows for the use of globally gridded monthly temperature products extending back to the beginning of the last century, and allows for the use of long-term monthly station records.
C12A-04 11:15h
Triggering of Active-layer Detachment Landslides: Field Observation and Laboratory Modelling
Permafrost degradation, resulting from Global Warming, is likely to increase the scale and frequency of thaw-related mass movements on slopes. Field studies of these processes present difficulties since prediction of where and when landslide events will occur is usually impossible, and long-term monitoring is necessary to collect substantial data sets. Shallow active-layer detachment slides in the polar desert environment of Ellesmere Island Canada, are described where triggering is apparently a response to unusually warm summer conditions, but landslide slope response is also apparently modulated by progressive reduction in shear strength over periods of one or two decades. An alternative approach to understanding such thaw-related slope failure mechanisms is through laboratory experiments in which model boundary conditions may be controlled. Since geotechnical centrifuge scaling laws show that time for both seepage force similarity and conductive heat transfer scale consistently, the centrifuge technique offers an ideal approach to the study of thaw consolidation processes in soils. Modelling experiments are described where the frozen model slopes were formed in silty clay. Models were thawed from the surface downwards at an acceleration of 20 gravities. Soil temperatures, pore water pressures during thaw and profiles of shear strain were recorded. Styles and mechanisms of failure are discussed in the context of the Ellesmere field study.
C12A-05 INVITED 11:30h
Permafrost and Railroad Construction on the Tibetan Plateau
The Qinghai-Xizang railroad is under construction on "The Roof of the World" --- the Tibetan Plateau, to be completed in 2007. The railroad will cross 550 km of permafrost region over the Tibetan Plateau, 50% of which is high-temperature permafrost and 37% of which is ice-rich permafrost. Predicted climate warming over the Tibetan Plateau in the coming decades would accelerate permafrost degradation. Surface disturbance due to the railroad construction would further destabilize permafrost conditions and seriously damage the ecosystem in the permafrost region. Thawing of warm permafrost over the Tibetan Plateau becomes one of the key issues in the cross-Plateau railroad construction. In this presentation, we will discuss techniques used to prevent permafrost from thawing due both to the climate warming and the surface disturbance of engineering construction. Although several techniques have been used over the Tibetan Plateau, application of crushed rock layer to cool permafrost and maintain permafrost stability is very successful at current stage although further observations are needed. We will also further demonstrate the principles of using the crushed rock layer to maintain permafrost based on data from field investigation, laboratory experiments, and numerical simulations.
C12A-06 11:50h
Role of Fire in the Permanent Loss of Permafrost under a Changing Climate
Climate conditions can be described as favorable (1), neutral (2), or unfavorable (3) for permafrost stability. When climate is favorable to permafrost, it takes only a few years to turn the soil below the active layer to a perennially frozen state (way 1). Permafrost thickness will grow with time until it reaches the maximum set by the geothermal gradient. Under the cold temperatures in the continuous permafrost zone, permafrost formation occurs independent of ecological processes. With climate neutral to permafrost, permafrost formation can occur in special topographic situations, such as north facing slopes. More commonly, however, permafrost formation is a result of ecosystem development and its effect on reducing soil temperatures (way 2). Permafrost in the discontinuous permafrost zone is highly dependent on type of soil, vegetation, and soil moisture. Therefore, permafrost is a product of landscape evolution, not just a product of climate, and it takes to an ecosystem hundreds of years to develop conditions that are favorable to permafrost under appropriate topographic conditions. Permafrost initially formed under a favorable climate can also persist under neutral or unfavorable climates because of surface conditions created by ecosystem development. Permafrost forms a drainage barrier that increases soil moisture and peat accumulation under anaerobic conditions, and influences vegetation succession in a direction favorable to moss growth. Thus, mosses, peat, and soil saturation produce an important positive feedback to permafrost. As a result, permafrost that formed in a favorable climate, such as during the Little Ice Age, can remain relatively stable in a neutral climate as long as it protected by other ecological components. Removal of vegetation and soil by natural or human disturbance typically leads to permafrost degradation. After fire, the permafrost table decreases for hundreds of years and can be stabilized only if ecosystem development after fire creates favorable conditions (way 2). The 6 million acres of the boreal forest burned in central Alaska during summer 2004 is a vivid illustration of the importance of the widespread impacts of fire. If permafrost was formed when the climate was favorable and remains protected in an unfavorable climate (way 3), it cannot recover after disturbance. Observations in West Siberia and Alaska show that restoration of permafrost after fire under the current climate of the discontinuous permafrost zone is unlikely under most conditions. Vast areas of boreal forest with evidence of permafrost existence in the recent past now are free of permafrost or have a permafrost table below 5-V10 m as a result of fire. It is typical for areas which are well-drained due to coarse soil or relief, and the changed thermal properties and successional pathways make recovery unlikely. Restoration of permafrost probably can occur only in areas of the flat relief and fine soil, where poor drainage makes possible permafrost development in way 2 and it takes more than 100 years. Such conditions exist in the Copper River Basin where soil is glacial-lacustrine clay. Even under the most conservative scenarios of climate change, which predict an increase in mean annual air temperatures of more than 3,aC in next 100 years, the climate will become unfavorable to permafrost. It means that in areas affected by fire during the last 20-30 years, permafrost in the boreal forest of the discontinuous permafrost zone will not recover. Permafrost degradation will be started by fires and concluded by climate change.
C12A-07 12:05h
Interactions of Multiple Factors in Creating Small Patterned-Ground Features Across the Arctic Bioclimate Gradient
Small patterned-ground landforms are described along a bioclimate gradient in northern Canada and Alaska and summarized in tables and figures showing strength of influence of contraction cracking, differential frost heave, and vegetation - within five bioclimate subzones and four major soil texture classes. In the coldest parts of the Arctic (bioclimate subzones A and B), contraction cracking at small scales (10-30 cm between cracks) is the dominant process and contributes to the formation of hummocky terrain; differential frost heave has a small role here except in course rocky terrain where sorted circles are common. The presence of contraction cracks on all surfaces, wet and dry, and on all soil types indicate that the majority of the contraction cracks are caused by thermal processes and not desiccation. Larger mounds, apparently the result of differential frost heave, occur in some areas of Subzone B where there is more vegetation and peat. In the Middle Arctic (bioclimate subzone C), both small turf hummocks and well-developed non-sorted circles occur. Turf hummocks are dominant on hill slopes; erosion of the inter-hummock areas and accumulation of eolian material on the hummock tops creates taller hummocks. Non-sorted stripes occur on many slopes. In the northern Low Arctic (Subzone D), non-sorted circles are the most common features; and turf hummocks are restricted to small areas - generally steep snow beds. The centers of most frost boils are barren or partially vegetated in Subzone D. In the sourthern Low Arctic (Subzone E), the vegetation is very active and able to colonize and totally cover frost boils. Large vegetated mounds are apparently the remnants of once active frost boils. In areas with more clayey soils of subzones D and E, well-developed tightly packed mounds are common, and frost boils often occur on the tops of the mounds. The spacing of the mound centers is often 2-3 m. Mounds are also common south of treeline. Soil texture affects frost boil morphology and heave characteristics. In silty areas of northern Alaska non-sorted circles have annual differential heave in the order of 20 cm - apparently contributing to the strong patterning in many areas (spotted tundra in the Russian literature). Areas with sandy soil have little differential heave and no frost boils in areas of pure sand; whereas, areas with clayey soils have mound shaped frost boils with little annual heave. Vegetation plays a major role in defining the boundaries of the patterned-ground features, possibly affecting differential frost heave by decreasing the soil temperature and thickness of the active layer in the inter-circle areas; however, at two sites on sandy soils with well-developed non-sorted circles only minor differential soil heave was measured. The cause of the barren centers at these sites is probably unrelated to heave and may be due to the accumulation of salts within the frost-boils. Needle ice is another major contributing cause of barrenness on frost boils and appears to develop most strongly on saturated silts.