Based on the observation data of 211 Global Navigation Satellite System (GNSS) regional stations from 2011 to 2016, and 59 continuously operating GNSS stations in the Yunnan region of China since 2013, we analyzed the strain characteristics of Yunnan from the following two perspectives. On the one hand, we analyzed the strain characteristics and tried to determine the potentially abnormal areas of Yunnan based on a multi-scale spherical wavelet method, from an overall strain background view. On the other hand, we investigated the relationship between the strain time series and the potentially abnormal areas of Tonghai County through an earthquake case study, and we attempted to extract the potentially information before the earthquake from the strain time series based on Hilbert-Huang transform (HHT), from a local strain time-series view. The results showed that: (1) in the integrated solution processing of the regional strain patterns, the spherical wavelet method can adapt to the spatial variability in station density and estimate the strain patterns at multiple scales; and (2) by combining the long-term characteristics of the strain time series, the grid division, the two times standard deviation constraint, and the HHT dynamic analysis, it gives the some potential information details on the eve of an earthquake such as the Tonghai earthquake.