We derive deformation rates of the fault applying wavelet transform and fault kinematics, by using the data of creepmeter and short baseline from Xianshuihe fault zone. Based on these results and combined with GPS observations, we analyze the segmental character and the spatial-temporal evolution of the fault activities. Our study shows that: (1) The fault activity is different along different segments. The segmental activity of Xianshuihe fault zone is significant. In detail, Luhuo and Daofu fault, extending northwest from Daofu County, are tensional and left-lateral, with comparatively larger strike-slip rates, and their motions are more. stable than Qianning and Zheduotang fault. While Qianning and Zheduotang fault to the south of Daofu County tend to be locked, and the states of movement are complex, where Qianning fault is compressive left-lateral and Zheduotang fault is right-lateral. The possible causes of this segmentation include fault bending and the differential movements of sub-block in Bayankala active block. (2) The slipping mode of fault alternates during different period of time. Although Xianshuihe fault is a left-lateral strike-slip active fault, nonetheless, there is reverse strike-slip along some segments before Wenchuan earthquake. The left-lateral creep rates of Luhuo and Daofu fault are decreased in the 2 years before M(s)8. 0 Wenchuan earthquake, and Qianning and Zheduotang fault exhibited right-lateral strike-slip abnormally in 2007. To the end of 2009, the range of reverse strike-slip is still growing. (3) The strike-slip rates observed by different-length measuring lines and the spatial distribution features of deformation belt are different. The strike-slip rates of Xianshuihe fault zone are not consistent at different segement and scale by different observing methods. The creepmeters are 0. 01 similar to 0. 78 mm/a (18. 7 m to 65. 1 m across fault), and short baselines are 0. 02 similar to 2. 46 mm/a (72m to 288 m across fault), however, GPS observations are 6 similar to 11 mm/a (dozens of kilometers across fault), and geological survey results are 5 similar to 15 mm/a. At increasing distances perpendicular to the fault, the average creep rate on one side of the fault grows following a logarithm function, and the deformation intensity attenuates following a power function. We infer that the far-field displacement includes the deformation or distributed offset of bilateral blocks of fault. At last, the difference between present fault deformation observations and geological survey results indicates that the fault slip rate is not a constant in time domain.
