The time series of turbidity in a coastal area were analyzed using the Hilbert-Huang Transform. The Empirical Mode Decomposition decomposed the original data to only eight intrinsic mode functions (IMFs) and a residue. Applying the Hilbert Transform to IMFs, the Hilbert spectrum was calculated, which is the time-frequency distribution of the instantaneous frequency and energy. The Hilbert spectrum showed the instantaneous frequencies change corresponding to the high turbidity events. The investigation of instantaneous frequency variations can be used to grasp the transition of the state of turbidity. The comparison between the Fourier spectrum and the Hilbert spectrum integrated in time showed the integrated Hilbert spectrum could make possible to grasp the cycle of the locally repeated events