Abstract: The observational data for more than 8 years have been accumulated since the reference stations of the Project of “China Crust Movement Measuring Network” run in 1999, now the analytical method of “point, line and surface” is generally proposed aiming at seismic prediction, observation object and the means of data processing. The practical application suggests that taking use of “line” maybe a kind of comparatively good method, whose best strongpoint is not only reducing noises in proper scale of space, but also giving prominence to the exceptional information. Based on the foundation of the relatively low power of measuring and surveying, and through the whole space scanning and phenomenon analyzing, we discover: ① Relative to Lasa station, Delingha station, the nearest one to the epicenter of Kunlun Mountain earthquake, appeared slow variation, even atresia (or no differential variation), in the EW direction early before the earthquake, and the atresia ended until the earthquake occurrence, the lasting time exceeded 2 years, the atresia magnitude was more than 10mm, while after the earthquake it resumed the normal status; ② Before several strong earthquakes in Yunnan region, Xiaguan station, relative to Luzhou station appeared atresia or variation that was lower than the normal level in the EW direction, after the strong earthquake in 2003, it resumed the normal status; In the latest Pu'er earthquake it has appeared relative atresia in both the EW and SN directions since the beginning of 2006, the magnitude of every direction approaches and reaches 10mm respectively; Besides, the lasting time of anomaly of every stage all exceed 1 year; ③ The EW direction of Tashi-Wushi appeared atresic activity before Bachu earthquake in Xinjiang, and the lasting time was more than 2 years, the magnitude was about 10mm, while after the earthquake it resumed the normal status; At the present time, the SN direction has the evidence of relative atresia, and which is inclining to be clear; ④ The EW direction of Dingxin-Xining has the evidence of relative atresia, but the lasting time is comparatively short; ⑤ The method of searching for section of relative “atresia” maybe become the main approach for determining the risk of strong earthquakes. Thus, it can be inferred that if the atresia was not relieved after Pu'er M_S6.4 earthquake, the tectonic zone of the west boundary of Sichuan and Yunnan diamond block and its surrounding regions is still the potential area for occurrence of strong earthquakes recently. The second risk area might be is the west area of Tianshan in Xinjiang, but the northeast boundary of Qingzang plateau is not clear at present.

Key words: GPS, Continuous observation, Atresia of tectonic activity, Strong earthquakes, Determining of the risk of strong earthquakes