Causal effects of dams and land cover changes on flood changes in mainland China

Abstract. Quantifying the effects of human activities on floods is challenging because limited knowledge and observations. Many previous methods fail to isolate different reduce uncertainty caused by small samples. We use panel regressions derive sensitivity annual maximum discharges (Q) changing values three factors: urban areas, cropland reservoir indexes for large medium dams. also test whether increase or decrease with increasing initial factors. This method applied in 757 non-nested catchments China. Results show that a 1 % point areas causes around 3.9 Q confidence interval CI = [1.9 %, 5.7 %]. Cropland have no significant effect Q. Reservoir index has decreasing effect: unit from 21.4 (with [11.4 29.9 %]) 6.2 [3.2 9.1 0 3. Among 61 increases observed 1992–2017, cause more than 10 only five (8.2 61) catchments. 234 at least one dam decreases 1960–2017, 138 (59.8 234) 1249 impacts indexes, 403 (32.3 %) during 46.7 are located middle downstream Yellow River Basin upper streams Hai He Basin. study extends regression hydrology sheds light attribution flood changes national scale.