Applying Hilbert-Huang Transform to Develop a Novel Rainfall Generator

Climate change impacts on hydrology and water resources have drawn many attentions. More and more research have identified that not only human-induced climate change but also natural long-term trend should be also considered. The generation of rainfall is a useful tool for the management of water resources systems to climate change. Most of the generation models used previously only take into account daily and monthly variations in the model parameters. Some long-term patterns, such as decadal cycles or trends, are usually lost in the stochastic generation. The purpose of this study is to develop a novel rainfall generation. This research applies Hilbert-Huang Transform (HHT) to decompose the historical rainfall data into intrinsic mod functions (IMF) and trends. Each IMF represents a generally simple component of the rainfall time series. The generated base rainfall series, which is generated from the arranged historical rainfall series by stochastic simulation, is added to the trends and IMFs with more than one year periods derived from the result of HHT. Thus the newly simulated rainfall series with long-term properties is formed, and can be further applied to the impact of climate change on water resources management.