%0 Dataset
%T Long-term series kilometer-level active layer thickness data set in Northern Hemisphere (1850-2100)
%J National Cryosphere Desert Data Center
%I National Cryosphere Desert Data Center(www.ncdc.ac.cn)
%U http://www.ncdc.ac.cn/portal/metadata/483c1a8f-b73c-41c7-9d9d-406b458f3d39
%W NCDC
%R 10.12072/ncdc.nieer.db6392.2024
%A Peng Xiaoqing
%A jinhaodong
%A Zhao guohui
%K Permafrost;remote sensing;surface temperature
%X The active layer thickness is one important indicators for the permafrost study . Under the current global warming background, the rate of temperature warming is higher than that in other areas, especially in high altitude and high latitude areas. This will inevitably have a great impact on the change of the active layer thickness in the permafrost area. In previous studies, based on field measured data and remote sensing inversion algorithms, many discussions have been made on the changes in the active layer thickness in historical periods, but their spatial resolution is low. Based on the CMIP6 data, the surface air temperature in the permafrost area with a resolution of 1 km is obtained by downscaling; then the data is used as a high-precision and high-resolution input variable for the future changes in the active layer thickness, so as to further obtain the active layer thickness of the Northern Hemisphere with a resolution of 1 km. By using a variety of machine learning methods and obtaining multi-method multi-mode averages, the accuracy and resolution of the active layer thickness dataset are improved. The RMSE of the ensemble average model output result is 67.39 cm, the MAE is 44.39 cm, and the R is 0.74.