Abstract: Hydrological simulation for ungauged regions is a challenge and a focal point in hydrological research. While reanalysis data has greatly enriched meteorological information in these regions, its performance remains debatable. In this study, using the measured meteorological dataset in the Honghe River Basin, we evaluated the temporal consistency of precipitation in reanalysis data using metrics such as correlation coefficient, root mean square error, and standard deviation ratio. Our findings indicate that the annual distribution of CMADS precipitation is consistent with the measured values, with a correlation coefficient up to 0.89. While CMADS dataset reflects the actual precipitation condition, it has underestimated precipitation, especially within the range of 0 to 10 mm. In terms of spatial consistency, Kappa consistency, spatial correlation, and relative error metrics illustrate that CMADS can effectively reflect the spatial distribution of precipitation. Finally, using corrected CMADS data and measured data to drive the SWAT hydrological model, we observed significant improvements in simulation results using CMADS at most stations compared to those using measured data. Therefore, CMADS is a vital source of hydrological data in the Honghe River Basin that can offer data support for water resources management and utilization in the area.

Key words: precipitation evaluation, hydrological simulation, unguaged area, CMADS, Honghe River Basin