Aggravation of debris flow disaster by extreme climate and engineering: a case study of the Tongzilin Gully, Southwestern Sichuan Province, China

A large-scale debris flow occurred in Tongzilin Gully in Puge County, Sichuan Province in southwestern China from 4:00 a.m. to 4:30 a.m. on August 8, 2017. Although early warning measures were taken before the occurrence of the debris flow, the disaster resulted in 25 fatalities, the destruction of 71 houses, and economic losses amounting to 160 million RMB. This study aimed to analyze the movement characteristics of the debris flow, reveal the cause and mechanism of the catastrophe, and assist in avoiding similar future disasters. To this end, we calculated the debris flow density, velocity, and peak discharge and analyzed the dynamic characteristics of the debris flow. The distributions of regional precipitation and solid materials were also determined to analyze the formation conditions. Moreover, the debris flow blockage and disaster amplification mechanisms were investigated through field surveys, remote sensing images, and parameter calculation. The results show that the debris flow was a low-frequency turbulent event with a peak flow rate of once in a hundred years and an average velocity of 6.34 m/s. The debris flow developed under unique conditions, corresponding to the combined action of natural and anthropogenic factors. First, the low-frequency debris flow was initiated by antecedent drought and short-term heavy rainfall. The antecedent drought increased the amount of loose materials in the watershed (141.5 × 10 4 m 3 ). The antecedent drought and short-term heavy rainfall nearly doubled the scale of the debris flow. Second, coarse particles blocked road culverts in the study area, leading to overflow and exacerbating the disaster. Consequently, the affected area and scale of the debris flow have been enlarged by 25% and by two times, respectively. The study findings provide deeper insight into the initiation and development mechanisms of debris flow disasters in mountainous areas, which may help improve the monitoring, early warning, and forecasting system.