The application discloses a hydrological forecasting technical field and relates to an
estuary gate
discharge period
water level forecasting method based on a T_TIDE and MTR
hybrid model, which comprises the following steps: collecting astronomical tide forecasting
water level at a target time, real-time monitoring or forecasting gate dam
discharge and upstream flow, inputting a pre-constructed T_TIDE and MTR
hybrid model, and obtaining a final
water level forecasting value; and the construction process of the T_TIDE and MTR
hybrid model comprises the following steps: obtaining historical long-sequence tidal
level data of a target region, using a T_TIDE model to calculate the astronomical tide forecasting water level at the target time, taking the astronomical tide forecasting water level, historical gate dam
discharge and upstream flow as forecasting factors, and determining threshold variables; using a K-means clustering
algorithm and combining AIC / BIC criteria to determine an optimal threshold segment number and threshold value, and constructing an MTR model; obtaining an initial T_TIDE and MTR
hybrid model, calibrating and verifying the initial T_TIDE and MTR
hybrid model according to historical tidal
level data, and obtaining a final T_TIDE and MTR
hybrid model. The application improves the precision, stability and timeliness of the
estuary water level forecasting during the discharge period.