4. DISCUSSIONS
To investigate the relationship between the potential maximum retention and SY in the Chenyulan River catchment area in typhoon events during 1999–2009, the rating curves of the runoff and SY were first established. Rating curve analysis results showed that the two parameters were positively correlated (Y = 0.0002x² + 0.0046x), and the correlation coefficient reached 0.95. Therefore, the rating curve is applicable for estimating SY in rainfall events.
The relationship between the potential maximum retention and SY indicated that first, the relationship of the two parameters conformed to a power function, which was consistent with the aforementioned study on rainfall events (Mishra, Tyagi, Singh, & Singh, 2006). Second, the event-based cumulative rainfall was the primary cause of the difference in functions, which was different from relevant studies, and thus this study categorized cumulative rainfall into three scenarios. Type I cumulative rainfall was less than 197.2 mm, which was smaller than the invoked value of SY; Type II was 197.2–754.4 mm, and the corresponding function was SY = 473.64S–0.866; and Type III was 754.4–1419.7 mm, and the corresponding function was SY = 13764S–1.133.
The difference in S–SY curves of Types II and III was primarily caused by cumulative rainfall. Different cumulative rainfall levels affected the relationship between the potential maximum retention and actual SY. For Type II, the cumulative rainfall was slightly higher than the invoked SY value; thus, the S–SY curve was relatively flat. For Type III, because of higher cumulative rainfall level, according to the concept of energy accumulation, the same condition for S would cause a relatively large SY; therefore, the S–SY curve was relatively steep. In addition, the difference in S–SY curves can also be explained by the infiltration rate decreasing over time. When the duration of rainfall is long, high cumulative rainfall would easily cause soil saturation, thereby inducing more surface runoff and soil erosion.
To facilitate sediment management in the catchment area and effectively determine key locations, this study took each subdivision as a unit. The average potential maximum erosion and average slope of each subdivision were considered to determine the subdivisions with top management priority. If the average potential maximum erosion was larger and the corresponding slope was steeper, sediment disasters would be more likely to occur in such an area. Therefore, such regions should be prioritized for management. The average potential maximum erosion of the subdivisions could be estimated by the S–SY function (Figure 13(a)), whereas the subdivision slope could be extracted using the DEM (Figure 13(b)). Through the normalization of the two parameters, the risk of sediment disasters in each subdivision could be estimated (Figure 13(c)), and the higher the risk value, the higher the priority that should be assigned to that subdivision.