Soil thermal conductivity is the main parameter to determine the heat transfer performance of the soil layer, which affects the ground temperature distribution, soil environment, and crop growth. The composition of organic matter directly affects the thermal conductivity of high organic soil. However, the current soil thermal conductivity model does not consider organic matter content and decomposition degree. That study analyzed the influence of the properties of undisturbed turfy soil in different layers on thermal conductivity. Additionally, we compared the applicability of more than 10 soil thermal conductivity calculation models to turfy soil and proposed an improved model. 1) The findings indicate that the thermal conductivity of each layer of unfrozen turfy soil is similar (0.51~0.66 W/(m?K)). However, after freezing, the thermal conductivity differs significantly between layers (1.00~1.62 W/(m?K)), suggesting that freezing alters the soil's composition. Due to more organic matter components and pores in turfy soil, the proportion of components with low heat transfer performance is higher, and the thermal conductivity of unfrozen turfy soil is lower than that of other organic soil with higher dry density. After freezing, most of the water in the soil becomes ice, and the thermal conductivity is greatly improved. The high water content makes the thermal conductivity of frozen turfy soil close to that of other organic soil. Furthermore, the correlation analysis between the fundamental physical properties of turfy soil and the thermal conductivity shows that the soil particle size distribution, organic matter content and decomposition degree will significantly affect the thermal conductivity of unfrozen turfy soil. 2) Most of the soil thermal conductivity prediction models (Campbell, Johansen and their derived models) fail to directly consider the influence of the proportion of organic matter components in the turfy soil, and have an overestimation effect on the thermal conductivity of organic matter in the solid phase. Relatively, the soil thermal conductivity model considering dry density (Nikoosokhan model) and component weight (Tian model) have good applicability for predicting turfy soil because they indirectly quantify the low heat transfer performance of organic matter components and pores through density differences into the calculation of soil thermal conductivity. However, it is still difficult to achieve the desired level of accuracy (RMSE>0.07 W/(m?K) for unfrozen soil; RMSE>0.28 W/(m?K) for frozen soil). 3) Based on the influence of soil properties and the principle of the calculation model, the parameters that can characterize the characteristics of turfy soil, including organic matter content (Oc) and decomposition degree (Dd), were introduced to improve the calculation model of thermal conductivity. The improved model comprehensively considers the characteristics of low dry density, high water content and high organic matter of turfy soil. Modifying the calculation parameters reduces the overestimation effect on the thermal conductivity of organic matter components. Furthermore, the improved model proposed in this study achieved a better prediction effect (R2 > 0.75) for both unfrozen and frozen turfy soil and improved the applicability and accuracy of the soil thermal conductivity calculation model for high organic matter turfy soil. The research results can provide a basis for agricultural cultivation and engineering construction in areas with seasonal frozen turfy soil. They can also serve as a theoretical reference for studying the thermophysical properties of soil with high organic matter.