Estimation Of Soil Thermal Conductivity And Its Influence On Permafrost Distribution Simulation In Qinghai-Tibet Plateau

The Qinghai-Tibet Plateau is home to the largest and widest permafrost regions in the middle and low latitudes of the world,and is an indicator of global climate change.With the global warming,the permafrost of the Qinghai-Tibet Plateau has degraded to varying degrees,which has had a significant impact on regional ecology,hydrology and climate.Soil thermal conductivity,as one of the key parameters for controlling surface heat transfer,is the key to studying soil surface energy exchange and water balance in frozen soil regions.However,most of the soil thermal conductivity parameterization schemes used in numerical simulations are empirical models.These models rely on prior knowledge.Therefore,in some areas with large areas and large soil heterogeneity,empirical models based on limited soil samples is usually impossible to accurately calculate the soil thermal conductivity of the entire area.Although some theoretical thermal conductivity models have also been proposed,they often include input parameters that are difficult to determine and their applicability in frozen soil is not verified.Therefore,most theoretical models for thawing soil face the challenge in frozen soil areas where freeze-thaw cycles frequently occur.In fact,in the numerical simulation of frozen soil,in order to improve the calculation accuracy of soil thermal conductivity,frozen soil and unfrozen soil often use piecewise functions with completely different mathematical forms.The piecewise function will bring greater uncertainty to the simulation of frozen soil distribution in the state of ice-water phase transition.Therefore,the development of a consistent physical-based unfrozen-frozen soil thermal conductivity model with a unified functional form helps reduce the uncertainty of numerical modeling,and is of great significance for accurately predicting the process of permafrost changes in the Qinghai-Tibet Plateau and the study of global climate change.Based on the comparative analysis of many thermal conductivity models based on theory and experience,this study screened a thermal conductivity model based on the Percolation Effective Medium Approximation Theory(PEMA),which can better express the relationship between thermal conductivity and saturation of thawing soil.In theory,it can be extended to frozen soil.This study proposed a consistent functional form of the thermal conductivity of thawing and frozen soil-Consistent Model.The input parameters of the model are the soil component's thermal conductivity and volume fraction.In addition,it also includes the scaling exponent t,the penetration threshold f_cand the proportional factor?.This study collected measured samples from existing literature,70%were used for model parameter determination,30%were used for model accuracy verification and comparison with the original PEMA model,the de Vries model(DV1963 model)often used in land surface process models and three Johansen series models.On this basis,the Consistent model was integrated into the Kudryavtsev model(K model)to simulate the permafrost distribution of the Qinghai-Tibet Plateau,and the station data,regional soil distribution survey data,and the existing frozen soil distribution map(Zou map)were used to evaluate the applicability of the Consistent model in the Qinghai-Tibet Plateau.The conclusions of this study are as follows:(1)The Consistent model has good accuracy in the prediction of thermal conductivity of thawing and frozen soils.The coefficient of determination R² is the largest,and the root mean square error RMSE and bias are the smallest.For thawing soil,the R²,RMSE and bias are 0.92,0.20W·m^-1K^-1,0.12W·m^-1K^-1,respectively;for frozen soil,they are 0.76,0.32W·m^-1K^-1 and 0.23 W·m^-1K^-1.It is suitable for calculating the thermal conductivity of various texture soils ranging from dry to saturated,from thawing to freezing.(2)The Consistent model uses the same functional form when calculating the thermal conductivity of thawing and frozen soils,which can reduce the uncertainty in the numerical simulation.The model introduced parameter t,f_c,and?all have physical meanings,and can be estimated by simple transfer function(PTF).(3)The permafrost distribution obtained by the K model under the two parameterization schemes of Consistent and DV1963 is more in line with the actual situation of the Qinghai-Tibet Plateau.The results show that the simulated temperature of the two models at seven stations is closer to the actual value(total error is less than-2.36?),the permafrost distribution results in the five survey areas are more similar to the actual spatial distribution.The permafrost distribution of the whole plateau is closer to the Zou diagram in the area of permafrost and spatial similarity.The distribution areas of permafrost in Zou diagram,Consistent,and DV1963 models are 1.03×10⁶,1.02×10⁶,and 0.94×10⁶ km²,respectively,the Kappa coefficients are all greater than 0.77,and the overall accuracy are all greater than 88.Other empirical models are quite different.(4)The thermal scheme has a significant impact on the simulation of the permafrost distribution on the Qinghai-Tibet Plateau.The colder season thermal conductivity calculated by the theoretical model and the empirical model are significantly different in value and spatial distribution.The theoretical model has the characteristic that the thermal conductivity of a large area in the colder season is smaller than that in the warmer season,while the normalized empirical model is just the opposite.The difference makes the final permafrost soil distribution is significantly different.