Abstract: Multisource technology has significantly improved the acquisition efficiency of seismic data.However,the acquired data often suffer from severe aliasing and channel missing,necessitating effective reconstruction of missing channels during data separation.Since an individual separation and reconstruction algorithm cannot improve accuracy and computational efficiency simultaneously,this study proposed a method combining the iterative shrinkage-thresholding algorithm(ISTA) and the accelerated linearized Bregman method(ALBM) to leverage ISTA's high post-processing accuracy and ALBM's fast convergence speed for simultaneous separation and reconstruction of multisource data.The simultaneous separation and reconstruction process employed curvelet transform(as a sparse basis),hard threshold function,exponential threshold and acceleration factors,and a novel exponential weighting factor.Finally,the single-source data were separated and reconstructed.The results obtained from the combined method were compared with those obtained using ISTA and ALBM,respectively.Additionally,the noise robustness and denoising capability of the combined method were examined.Both theoretical simulation and practical application demonstrate that the combined method achieves higher accuracy and faster computational efficiency in separating and reconstructing complete single-source signals.