Description: Palaeotemperature reconstructions play an important role as palaeoclimate records, for our understanding of the climate system behavior as such, as well as being the basis for models identifying the impact of these climate conditions on specific processes in the past and future. Temperature records reconstructed from borehole logs have a more direct relationship to the historic temperature history than other proxy-based reconstructions such as tree-rings, pollen or isotope ratios in ice cores, which can include influences from other independent factors on those proxies. At larger depths borehole temperatures are dominated by the geothermal heat flux and a rather uniform geothermal gradient in the profile. At shallower levels temperature variations at the surface propagate as heat waves into the ground. The further down, the more the temperature reflects influences of longer periods of surface variations due to the Earth’s damping higher angular frequency periods first. This study uses two inversion optimization methods previously applied to ice core sites (Roberts et al., 2013) to reconstruct the local surface temperature history at two shallow (100m and 65m deep) permafrost borehole sites: Sardakh Island in the Lena-Delta and Cape Mamontov Klyk in the Western Laptev Sea, Russia (Fig. 1). We employed a flux-conserving finite volume numerical soil model to calculate temperature-depth-profiles from surface temperature histories. Thermal properties of the sites were retrieved from either the observed temperature field or the sediment composition analysis of the borehole. Two inversion schemes that employ the forward soil model to optimize surface temperature history in a least square sense were used in the reconstruction: (i) the least square QR (LSQR) method and (ii) the particle swarm optimization (PSO) method. The latter resembles a Monte Carlo based approach (Ebbesen et al., 2012), the former is based on a generalized least-square solution of a linearized version of the problem as utilized by Orsi et al. (2012). Recoverable time length for the surface temperature histories for the two borehole sites were found to be well above 400 years in both cases by frequency-dependent heat wave damping analysis. The local surface soil temperature reconstructions for the two boreholes are discussed in comparison to other local as well as larger scale global temperature reconstructions to highlight important local and regional deviations. Additionally, the reconstructions of both sites are compared on the basis that one (Mamontov Klyk) is situated away from any major river systems and the other (Sardakh) is situated in the Lena River Delta, possibly showing thermal influence by the river. The local surface temperature history is important as a driving input factor in local permafrost models that assess the evolution, degradation and impact of permafrost in the high latitudes in the future climate system. References: Roberts JL, Moy AD, van Ommen TD, Curran MAJ, Worby AP, Goodwin ID, Inoue M. 2013. Borehole temperatures reveal a changed energy budget at Mill Island, East Antarctica, over recent decades. The Cryosphere 7: 263-273 Ebbesen S, Kiwitz P, Guzzella L. 2012. A generic particle swarm optimization Matlab function. American Control Conference (ACC); 1519-1524 Orsi A, Cornuelle B, Severinghaus J. 2012. Little Ice Age cold interval in West Antarctica: Evidence from borehole temperature at the West Antarctic Ice Sheet (WAIS) Divide. Geophys. Res. Lett. 39: L09710. DOI: 10.1029/2012GL051260

Description: Ice-rich permafrost coasts in the Arctic are susceptible to a variety of changing environmental factors, all of which currently point to increasing coastal erosion rates and mass fluxes of sediment and carbon to the shallow arctic shelf seas. Coastal erosion and flooding inundate terrestrial permafrost with seawater and create submarine permafrost. Permafrost begins to warm under marine conditions, which can destabilize the sea floor and may release greenhouse gases. The rate and spatial distribution of subsea permafrost degradation in the Laptev, East Siberian and Chukchi seas, which together comprise more than half of the Arctic Ocean continental shelf, remain poorly explored. We report on the transition of terrestrial to subsea permafrost at four coastal sites in the Laptev Sea: Cape Mamontov Klyk in the western Laptev Sea, and Buor Khaya Peninsula, Muostakh Island and the Bykovsky Peninsula in the central Laptev Sea. We use coastal erosion rates from about the last 70 years to estimate the period of inundation at these sites. Combined with direct (drilling and temperature) and indirect (geophysical) observations of thaw depths of ice-bonded permafrost, we estimate recent degradation rates of permafrost over the past centuries. Based on these observations, the unfrozen sediment layer overlying ice-bonded permafrost increased from less than a meter at the shoreline to over 30 m below seabed with increasing distance from the shoreline at our study sites, with high spatial variability between and within sites. Observed temperatures of the sediment ranged from -5 °C to positive temperatures. In coastal sediments, it is difficult to establish an age-depth model, making corroboration of estimated degradation rates a challenge. Nonetheless, as the thickness of the unfrozen sediment layer increases over time, the vertical thermal and salt concentration gradients decrease, slowing the downward heat and mass fluxes responsible for degradation. High sedimentation rates and ice contents probably stabilize subsea permafrost. We suggest that permafrost degradation relevant to gas flow is likely to have occurred where permafrost warmed prior to inundation.

Description: Permafrost inundated since the last glacial maximum is degrading, potentially releasing trapped or stabilized greenhouse gases, but few observations of the depth of ice-bonded permafrost (IBP) below the seafloor exist for most of the arctic continental shelf. We use spectral ratios of the ambient vibration seismic wavefield, together with estimated shear wave velocity from the dispersion curves of surface waves, for estimating the thickness of the sediment overlying the IBP. Peaks in spectral ratios modeled for three-layered 1-D systems correspond with varying thickness of the unfrozen sediment. Seismic receivers were deployed on the seabed around Muostakh Island in the central Laptev Sea, Siberia. We derive depths of the IBP between 3.7 and 20.7 m ± 15%, increasing with distance from the shoreline. Correspondence between expected permafrost distribution, modeled response, and observational data suggests that the method is promising for the determination of the thickness of unfrozen sediment.

Description: The degradation of permafrost beneath the seabed on the East Siberian Shelf has been implicated in the release of greenhouse gases, especially methane, and the potential de-stabilization of gas hydrates. We investigate the degradation of subsea permafrost using geophysical methods, drilling and temperature measurements. Recovered subsea sediments offer an opportunity to investigate sediment composition, reconstruct permafrost degradation processes, and to better understand the consequences of this degradation. A 52 m deep borehole was drilled about 800 m offshore to the west of the Buor Khaya Peninsula in the central Laptev Sea. Coastal exposures and an onshore borehole revealed “ice complex” stratigraphy with high ice and carbon contents. The landscape of the Buor Khaya Peninsula, however, has undergone substantial degradation, so that isolated islands of relatively intact ice complex cover about 15% of the area among a palimpsest of thermokarst basins. The subsea sediment was mostly sandy with spatially highly variable carbon contents and isolated layers of woody plant remains probably deposited in a fluvial environment before freezing. Ice-bonded permafrost was encountered at 28 m b.s.l. The western coast of the Buor Khaya Peninsula has been retreating at between 1 and 2 m per year. The position of the ice-bonded permafrost table with distance from shore suggests that subsea permafrost degrades at a mean rate of 3 to 4 cm a-1 following erosion. Methane was entrapped throughout the frozen sediment suggesting the mobilization of methane along with permafrost degradation at this site.

Description: In central Siberia, past temperature changes have been driving permafrost warming in a region with large organic carbon reserves stored in the perennially frozen ground. However, local arctic temperature histories in the ice-rich permafrost areas of the remote Russian Arctic are sparsely known or based on proxy data with potential seasonal biases and underrepresented in circum-Arctic reconstructions. This study employed two inversion schemes (particle swarm optimization and a least-square method) to reconstruct temperature histories for the past 200–300 years in the Laptev Sea region from two permafrost borehole temperature records. These data were evaluated against larger scale reconstructions from the region. Distinct differences between the western Laptev Sea and the Lena Delta sites were recognized, such as a transition to warmer temperatures a century later in the western Laptev Sea as well as a peak in warming 3 decades later. The local permafrost surface temperature history at Sardakh Island in the Lena Delta was reminiscent of the circum-Arctic regional average trends. However, Mamontov Klyk in the western Laptev Sea was consistent to Arctic trends only in the most recent decade and was more similar to northern hemispheric mean trends. Both sites are consistent with a rapid recent warming that is of synoptic scale. Different environmental influences such as synoptic atmospheric circulation and sea ice may be responsible for differences between the sites. The shallow permafrost boreholes provide missing well-resolved short-scale temperature information in the coastal permafrost tundra of the Arctic. As local differences from circum-Arctic reconstructions, such as later warming and higher warming magnitude, were shown to exist in this region, our results provide a basis for local surface temperature record parameterization of climate models, and in particular of permafrost models.

Description: Palaeotemperature reconstructions are valuable palaeoclimate indicators and important tools for the understanding of interactions in the climate system. They form a basis for models identifying the impact of various processes within the past and future climate system. Siberia is a region with large organic carbon reserves stored in permafrost (perennially frozen ground). Temperatures in this region are thus important as a driver for a positive feedback to the global climate. Local temperature histories in the ice-rich permafrost areas of the Russian Arctic are either sparse or based on proxy data with potential seasonal biases. Borehole temperature reconstructions are sensitive to the temperature signal throughout the year and available in regions for which no other records exist. This study used two inversion methods, particle swarm optimization and a least squares technique, to retrieve temperature histories of the last 200-300 years in the Laptev Sea region from two permafrost borehole temperature records. The retrieved histories were compared to larger scale reconstructions from the region. Distinct differences in the histories between the Lena Delta and western Laptev Sea sites were found, notably a one-century delay of warming and a three decade delay in peak warming in the western Laptev Sea. The local permafrost surface temperatures at Sardakh Island (central Lena Delta) resembled the circum-Arctic regional average trends. At Mamontov Klyk (western Laptev Sea) this was the case only for the most recent decade. In contrast, the Mamontov Klyk history was more similar to northern hemispheric mean trends. A rapid recent warming of synoptic scale was consistently observed at both sites. Differences in the past temperature trends between the sites may be caused by regionally differing environmental influences, such as atmospheric circulation and sea ice coverage. The re-constructed magnitude of temperature changes is consistent with warming greater than mean Arctic temperature trends. In conclusion, reconstruction from shallow permafrost boreholes provides short-scale temperature histories in the coastal tundra of the remote Arctic (resolved at annual to multi-decadal scale). As local differences from the circum-Arctic average – including later warming and higher warming magnitude – were shown to exist in this region, our results provide a basis for local surface temperature record parameterization of climate models and of permafrost models in particular-

Description: During lower sea levels in glacial periods, deep permafrost formed on large continental shelf areas of the Arctic Ocean. Subsequent sea level rise and coastal erosion created subsea permafrost, which generally degrades after inundation under the influence of a complex suite of marine, near-shore processes. Global warming is especially pronounced in the Arctic, and will increase the transition to and the degradation of subsea permafrost, with implications for atmospheric climate forcing, offshore infrastructure, and aquatic ecosystems. This thesis combines new geophysical, borehole observational and modelling approaches to enhance our understanding of subsea permafrost dynamics. Three specific areas for advancement were identified: (I) sparsity of observational data, (II) lacking implementation of salt infiltration mechanisms in models, and (III) poor understanding of the regional differences in key driving parameters. This study tested the combination of spectral ratios of the ambient vibration seismic wavefield, together with estimated shear wave velocity from seismic interferometry analysis, for estimating the thickness of the unfrozen sediment overlying the ice-bonded permafrost offshore. Mesoscale numerical calculations (10 1 to 10 2 m, thousands of years) were employed to develop and solve the coupled heat diffusion and salt transport equations including phase change effects. Model soil parameters were constrained by borehole data, and the impact of a variety of influences during the transgression was tested in modelling studies. In addition, two inversion schemes (particle swarm optimization and a least-square method) were used to reconstruct temperature histories for the past 200–300 years in the Laptev Sea region in Siberia from two permafrost borehole temperature records. These data were evaluated against larger scale reconstructions from the region. It was found (I) that peaks in spectral ratios modelled for three-layer, one-dimensional systems corresponded with thaw depths. Around Muostakh Island in the central Laptev Sea seismic receivers were deployed on the seabed. Derived depths of the ice-bonded permafrost table were between 3.7–20.7 m ± 15 %, increasing with distance from the coast. (II) Temperatures modelled during the transition to subsea permafrost resembled isothermal conditions after about 2000 years of inundation at Cape Mamontov Klyk, consistent with observations from offshore boreholes. Stratigraphic scenarios showed that salt distribution and infiltration had a large impact on the ice saturation in the sediments. Three key factors were identified that, when changed, shifted the modelled permafrost thaw depth most strongly: bottom water temperatures, shoreline retreat rate and initial temperature before inundation. Salt transport based on diffusion and contribution from arbitrary density-driven mechanisms only accounted for about 50 % of observed thaw depths at offshore sites hundreds to thousands of years after inundation. This bias was found consistently at all three sites in the Laptev Sea region. (III) In the temperature reconstructions, distinct differences in the local temperature histories between the western Laptev Sea and the Lena Delta sites were recognized, such as a transition to warmer temperatures a century later in the western Laptev Sea as well as a peak in warming three decades later. The local permafrost surface temperature history at Sardakh Island in the Lena Delta was reminiscent of the circum-Arctic regional average trends. However, Mamontov Klyk in the western Laptev Sea was consistent to Arctic trends only in the most recent decade and was more similar to northern hemispheric mean trends. Both sites were consistent with a rapid synoptic recent warming. In conclusion, the consistency between modelled response, expected permafrost distribution, and observational data suggests that the passive seismic method is promising for the determination of the thickness of unfrozen sediment on the continental Arctic shelf. The quantified gap between currently modelled and observed thaw depths means that the impact of degradation on climate forcing, ecosystems, and infrastructure is larger than current models predict. This discrepancy suggests the importance of further mechanisms of salt penetration and thaw that have not been considered – either pre-inundation or post-inundation, or both. In addition, any meaningful modelling of subsea permafrost would have to constrain the identified key factors and their regional differences well. The shallow permafrost boreholes provide missing well-resolved short-scale temperature information in the coastal permafrost tundra of the Arctic. As local differences from circum-Arctic reconstructions, such as later warming and higher warming magnitude, were shown to exist in this region, these results provide a basis for local surface temperature record parameterization of climate and, in particular, permafrost models. The results of this work bring us one step further to understanding the full picture of the transition from terrestrial to subsea permafrost.