The interpolation of sparse seismic data is important for seismic data processing, especially in the case of marine data acquisition. The conventional interferometric interpolation method interpolates missing data based on seismic interferometry in the time-space domain. However, missing traces and the limited acquisition aperture will reduce the quality and accuracy of virtual shot gathers, thereby affecting the final interpolation results. To improve the conventional method, we have adopted a four-step Radon domain interferometric interpolation method for sparse seismic data. First, we generate synthetic seismic data with a water-layer model and replace missing traces in the input data with synthetic traces. Second, we sort the input shot gathers into common-receiver-point (CRP) gathers and transform the CRP gathers into the Radon domain using a sparse Radon transform method. Third, we perform Radon domain seismic interferometric interpolation by cross-correlating traces with the same ray parameter and summing over all different ray parameters to obtain a virtual shot gather. Fourth, we apply a 1D least-squares matching filter to the virtual shot gather to obtain the interpolation result. Numerical and field examples demonstrate that our method can effectively interpolate sparse seismic data. Compared with the conventional time-space domain interferometric interpolation method, our method can provide more accurate interpolation results. Moreover, it is less sensitive to the depth of the water layer and an inaccurate impedance contrast of the water bottom.

中文翻译：

---

稀疏地震数据的氡域干涉插值

稀疏地震数据的插值对于地震数据处理很重要，特别是在海洋数据采集的情况下。传统的干涉插值方法基于时空域的地震干涉测量法对缺失数据进行插值。然而，缺失的轨迹和有限的采集孔径会降低虚拟炮点道集的质量和精度，从而影响最终的插值结果。为了改进传统方法，我们对稀疏地震数据采用了四步氡域干涉插值方法。首先，我们使用水层模型生成合成地震数据，并用合成道替换输入数据中缺失的道。第二，我们将输入炮点道集分类为公共接收点 (CRP) 道集，并使用稀疏氡变换方法将 CRP 道集转换为氡域。第三，我们通过将具有相同射线参数的轨迹相互关联并对所有不同的射线参数求和以获得虚拟炮点集来执行氡域地震干涉插值。第四，我们对虚拟镜头集应用一维最小二乘匹配滤波器以获得插值结果。数值和现场实例表明我们的方法可以有效地插入稀疏地震数据。与传统的时空域干涉插值方法相比，我们的方法可以提供更准确的插值结果。此外，它对水层深度和水底阻抗对比不准确。