Increasing the Depth of a Land Surface Model. Part II: Temperature Sensitivity to Improved Subsurface Thermodynamics and Associated Permafrost Response

Abstract

The impact of various modifications of the JSBACH Land Surface Model to represent soil temperature and cold-region hydro-thermodynamic processes in climate projections of the 21st century is examined. We explore the sensitivity of JSBACH to changes in the soil thermodynamics, energy balance and storage, and the effect of including freezing and thawing processes. The changes involve 1) the net effect of an improved soil physical representation and 2) the sensitivity of our results to changed soil parameter values and their contribution to the simulation of soil temperatures and soil moisture, both aspects being presented in the frame of an increased bottom boundary depth from 9.83 m to 1418.84 m. The implementation of water phase changes and supercooled water in the ground creates a coupling between the soil thermal and hydrological regimes through latent heat exchange. Momentous effects on subsurface temperature of up to ±3 K, together with soil drying in the high northern latitudes, can be found at regional scales when applying improved hydro-thermodynamic soil physics. The sensitivity of the model to different soil parameter datasets occurs to be low but shows important implications for the root zone soil moisture content. The evolution of permafrost under pre-industrial forcing conditions emerges in simulated trajectories of stable states that differ by 4 – 6 • 106 km2 and shows large differences in the spatial extent of 105 –106 km2 by 2100, depending on the model configuration.

Continental heat storage: Contributions from ground, inland waters, and permafrost thawing

“…(MacDougall et al, 2008(MacDougall et al, , 2010Cuesta-Valero et al, 2016). Furthermore, this deeper subsurface in LSMs has also improved the representation of permafrost dynamics, showing how the ground heat storage retrieved from measurements of subsurface temperature profiles have informed the development of climate models (Alexeev et al, 2007;Nicolsky et al, 2007;Hermoso de Mendoza et al, 2020;González-Rouco et al, 2021;Steinert et al, 2021). Another approach may be to use the retrieved estimates of continental heat storage as a reference to constraint projections of climate change (Tokarska et al, 2020;Ribes et al, 2021).…”

“…However, direct use of the simulated ice content and soil temperatures from the CMIP simulations is not currently possible because the representation of the soil in the models is too shallow to assess the evolution of the thermal state of the ground beyond the near-surface permafrost (Koven et al, 2013;Slater & Lawrence, 2013;Burke et al, 2020;Hermoso de Mendoza et al, 2020;Steinert et al, 2021). Furthermore, these land surface model components do not represent excess ice in the ground or ground subsidence, which bias the represented permafrost thawing (e.g., Lee et al, 2014;Rodenhizer et al, 2020).…”

“…Active layer thickness is also increasing at most measurement locations around the world (Smith et al, 2022b). Additionally, global climate simulations project a decrease in global permafrost extension from 10 % to more than 80 % due to thawing by the end of the 21 st century, depending on future greenhouse gas emissions (Koven et al, 2013;Slater & Lawrence, 2013;Burke et al, 2020;Hermoso de Mendoza et al;Steinert et al, 2021). Consequently, permafrost heat uptake is expected to increase in the following decades.…”

First Quantification of the Permafrost Heat Sink in the Earth's Climate System

“…(2022) to conduct distributed model simulations for the northern terrestrial permafrost region on a 1°‐latitude by 1°‐longitude grid. For our model setup, we used a subsurface domain down to a depth of z lb = 550 m, allowing for a much more realistic representation of the long‐term transient thermal regime and heat reservoirs than the shallow subsurface representation typical for LSMs (Hermoso de Mendoza et al., 2020; Steinert, González‐Rouco, Vrese, et al., 2021). At the bottom of the model domain, a geothermal heat flux ( Q geo [W m −2 ]) according to Davies (2013) was prescribed as the lower boundary condition.…”

Thermodynamic and hydrological drivers of the subsurface thermal regime in Central Spain

“…Moreover, the analysis might be hampered by potential deviations from the conductive regime in the soil near the surface due to biological, chemical, and hydrological processes (Gao et al, 2008;Tong et al, 2017). A deeper insight into the soil thermal structure is required to understand changes in its thermal properties with depth, relevant for Land Surface Models (LSMs) and therefore with implications on land-atmosphere interactions affecting Earth System (Smerdon and Stieglitz, 2006;Ekici et al, 2014;Hermoso de Mendoza et al, 2020;Steinert et al, 2021) and forecast models (Miralles et al, 2019). A better characterization of subsurface thermal properties can also contribute to derive more accurate land heat uptake estimates, which is the second component contributing the most to terrestrial energy partitioning (Cuesta-Valero et al, 2022b;von Schuckmann et al, 2022).…”