Both of the terrestrial water cycle and catchment characteristics can influence the propagation of meteorological to hydrological drought, and analyzing the propagation characteristics is crucial for the early warning and management of drought. In this work, the SWAT model was used to simulate the hydrological cycle in the Jinghe River Basin. Run theory, paired analysis, and correlation analysis were combined to reveal the spatial evolution patterns and driving factors of drought propagation across multiple time scales. Results show that drought propagation exhibits significant temporal dependence. As time scales increase, propagation time and propagation rate increase, while the propagation degree of duration and severity decrease. Spatially, there is a distinct upstream-downstream difference. The sensitivity of drought propagation in downstream areas is relatively low, but once the propagation starts, the response is faster, the duration is longer, and the severity is higher. Drought propagation at the monthly and seasonal scales are mainly driven by hydrological variables such as precipitation, actual evapotranspiration, soil water content, and lateral recharge. Catchment characteristics such as agricultural land, grassland, and slope of 0-15° or greater than 25° also play an important role in drought propagation. But at the annual scale, drought propagation is less affected by these factors.