Abstract: To denoise microseismic noises,an EEMD_Hankel_SVD ( ensemble empirical mode decomposition_Hankel matrix_signular value decomposition) combined method is proposed. Firstly,the EEMD is used to obtain mode func- tions,then the correlation coefficient between each mode function and original signal is calculated,and the mode func- tions before the first local maximum correlation coefficient difference are deleted. The rest mode functions are used to construct Hankel matrixes,and the SVD is applied to decompose the Hankel matrixes. The microseismic signal and noises are divided by the curve of singular values and this is used to denoise mode functions,then the denoised mode functions are combined to obtain denoised microseismic signal. The simulated tests show that the proposed method can retain local features well and increase signal to noise ratio (SNR). The application to mine microseismic signals shows that the proposed method can effectively improve the P-phase picking results of the STA / LTA picker,PAI-K ( phase arrival identification-kurtosis) picker and AIC ( Akaike information criterion) picker. In addition,both the simulated tests and mine microseismic signal picking results show that the denoising performance of this method is better than that of wavelet reconstruction,empirical mode decomposition (EMD) reconstruction,and Hankel_SVD denoising. Furthermore,the combination with the AIC picker obtains a best picking result.