The environmental background noise of a seismic station largely determines its signal detection capability. A systematic and objective assessment of noise levels is essential for station site selection and construction,as well as for network optimization and performance evaluation. the widely used New Low Noise Model(NLNM)was originally derived from data recorded at 75 stations worldwide during the 1980s and 1990s. In this study,the same noise power spectral density and noise probability density function methods used in the original modeling were employed to obtain noise samples from the same stations,thereby re-evaluating whether significant changes have occurred in the global background noise model over long time spans. the new results,based on the same set of stations,indicate that the current 90% noise level has increased by an average of 4.2dB in the 0.1～10Hz band,while the 50% noise level has increased by an average of 6.7dB. By comparing station noise levels with population distribution,and by considering the noise-reduction characteristics of borehole stations among the IMS primary stations established for global nuclear test monitoring,YKA0B,TX01,and NV01 were identified as ultra-low-noise stations located far from coastlines and sparsely populated areas. the common characteristics of these low-noise stations were then analyzed. Finally,statistical assessments of noise levels at stations in northwestern and northeastern China were conducted in combination with population distribution data,and the most suitable sites for detecting weak teleseismic signals were selected. This work provides a scientific basis for the construction of high-sensitivity seismic monitoring stations.