Quality control of stream water-stage using Hilbert-Huang Transform

Hydrological data, especially water stage and discharge, is very important for water resources planning and development, hydraulic structure design, and water resources management. Thus the hydrological data has to be observed and collected regularly and continuously. The hydrological data can be affected by many factors such as people, instruments, and climate. Therefore, the collected hydrological data still need to be subject to quality control and inspection to eliminate unreasonable data to ensure the accuracy and reliability. Traditionally, the quality control and inspection of stream water-stage is mainly manual. The verification of water stage data needs experienced hydrologists to judge the correctness of the data, and cannot be processed automatically. It is time consumed, costly, and labor intensive to process the quality control of stream water stage. Therefore, it is necessary to develop a feasible model to automatically check stream water-stage for providing reliable and accurate hydrological data.

This study applies Hilbert-Huang Transform (HHT) to process stream water-stage. The HHT is composed of Empirical Mode Decomposition (EEMD) and Hilbert transform (HT). The EEMD decomposes stream water-stage into many intrinsic mode functions (IMFs) and a residual. The first IMF component is used for Hilbert transform conversion to obtain the time amplitude energy relationship diagram. The amplitude fluctuation of the corresponding component of the stream water-stage, the amplitude value of the outliers can be revealed. When the amplitude value is larger than usual, there may be outliers, and vice versa. It depends on the threshold that is established in this study as the basis for filtering the incorrect water-stage. Therefore automatically inspecting the water-stage data can be achieved. The model for automatic inspecting procedure developed by this study will greatly reduce the manual quality control, not only shorten the checking time, save manpower, but also provide reliable and correct river water stage data.