Methodology for determining runoff coefficients based on rainfall depth and sensitivity analysis of influencing factors

The roof runoff coefficient (RC) is a critical design parameter for roof rainwater harvesting systems (RRHS), traditionally derived from standard norms with limited accuracy. This study investigates the influencing factors and determination methodology of RC through simulated rainfall experiments and analysis of 116 actual rainfall events. Pearson correlation and response surface analysis identified rainfall intensity as the primary factor affecting RC, followed by roof slope, while roof material and area showed negligible impact. A linear regression model (RC = 0.86 − 0.48 × 0.94 ^H ) was developed to estimate RC based on rainfall depth ( H ), demonstrating high reliability with a Nash-Sutcliffe efficiency coefficient of 0.927. Experimental results revealed that higher rainfall intensity and steeper slopes significantly increase RC by reducing runoff generation and confluence time. Application of the model in RRHS design optimized storage volume by 50.79% compared to traditional methods, although it resulted in a moderate decrease in cumulative water supply and reliability. The study provides a data-driven approach for accurately determining RC, enhancing the efficiency and economic viability of rainwater resource utilization in urban water management.