Forecasting Characteristic Water Levels of Taihu Lake Based on Time Series Decomposing Method

Based on the daily average water levels of the five gauges around the Taihu Lake observed from 1956to 2006,the annual characteristic water levels(average,highest and lowest)were obtained.Time series analysis method was adopted,and the decomposition model of trend,period and random components was applied to simulate and forecast the annual characteristic water levels of Taihu Lake.The annual characteristic water levels were forecasted for the next 15years and the future flood scenarios of Taihu Lake basin were also addressed.The results indicate that the highest water level of Taihu Lake may reach 4.05min 2021,which is significantly higher than the warning water level of Taihu Lake and thus worthy of attention for relevant department.