摘要： In order to improve the accuracy of flood calculation in the small watershed of tailings reser⁃ voir，optimize the safety of flood control design，solve the problems of insufficient suitability of rainstorm at⁃ tenuation index in traditional methods，unreasonable simplification of flood process line，and large calcula⁃ tion error of runoff and convolution parameters in the small watershed，the parameter system optimization and multi-peak flood process line reconstruction based on reasoning formula method were carried out. By es⁃ tablishing a three-stage dynamic correction model for rainstorm attenuation index，a double-source interac⁃ tive verification algorithm for loss rate and a regional correction method for confluence parameters，a five�stage multi-modal triangular superposition model was innovatively proposed，and a Python automatic calcu⁃ lation module was developed. The results show that the calculation error of flood peak flow can be controlled within 5% by dividing the periods of main rain peak and secondary rain peak and introducing the iterative tri⁃ al calculation mechanism. After using Python programming to realize the algorithm modularization，the com⁃ puting efficiency is increased by more than 80%. The improved reasoning formula method can effectively solve the problem of parameter adaptation in extra small watershed，and the multi-peak triangular superpo⁃ sition method simplifies the flood process line construction process，which has significant practical value for tailings pond projects lacking measured data. The flood process line calculation method has strong applica⁃ bility to the flood calculation project of extra small watershed，and can be used in the flood control review of tailings reservoir.