Project Details
Description
This research aims to improve flood forecasting by enhancing the entire state estimation of the hydrological
forecasting chain. To achieve the goal of this study, three strategies will be conducted as follows. The first
is to improve rainfall input, as it is crucial for hydrological simulation, by using data from weather radars.
Machine learning and empirical methods for radar rainfall estimation will be investigated to provide better
rainfall estimates. The second is a modeling approach, as the existing system uses a conceptual lumped
hydrologic model (NAM) to translate precipitation to runoff, which usually does not perform well when the
spatial variation of catchment properties is high. The distributed hydrological model wflow_sbm will be
set up and validated and used to take advantage of spatial data (from weather radar, numerical weather
prediction) and deal with the high spatial variation in, among other, rainfall, topography and land use in
Thailand. The third approach is to adopt data assimilation techniques to enhance the hydrological model
states, which also contributes some improvement to the forecast discharge. Furthermore, this study will
explore the possibility of improving the assimilation process using machine learning techniques in terms
of efficiency or performance when implemented in the operational system. To test the approach, this study
will focus on the Ping River basin, the largest tributary of the Chao Phraya River.
Status | Active |
---|---|
Effective start/end date | 1/05/24 → … |
Fingerprint
Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.