To study the subsurface features and structures, the gravity effects of the surrounding topography should be reduced from acquired gravity field data. Several methods are available to calculate terrain effects on each gravity station. Some of these methods are tedious and time consuming due to a large number of calculations, long-distance terrain effects, and minimum acceptable errors. The other fast methods do not fulfill the accuracy requirement for local surveys. In rough topographies, using average elevation for sectors of the calculation area leads to overestimation or underestimation of the terrain effect. Since most of the terrain correction methods employ the pre-divided web or mesh for the survey area, the used sectors do not match geographical features with distinct mass centers. An Optimally Selecting Sectors (OSS) algorithm is proposed, which automatically partitions the surrounding area to a set of sectors in a way that separates different mass centers, finds optimum elevation of these sectors, and calculates terrain effect at the gravity stations. This new procedure improves the accuracy of calculated terrain effects. A proper tolerance inside the algorithm controls the accuracy of the method. Defining this tolerance relies on the application of gravity data. The application of this method on synthetic models with different geometrical shapes and real digital elevation data of a mountainous area at the Kurdistan region shows improvement in the accuracy of terrain correction. However, the proposed method OSS introduces extra calculation compared to some of the previous terrain correction methods.

为了研究地下特征和结构，应从采集的重力场数据中减少周围地形的重力影响。有几种方法可用于计算每个重力站上的地形影响。由于大量的计算，远距离地形影响以及最小的可接受误差，这些方法中的某些繁琐且耗时。其他快速方法不能满足本地调查的准确性要求。在粗糙的地形中，对计算区域的扇区使用平均高程会导致地形效应的高估或低估。由于大多数地形校正方法都针对调查区域采用了预先划分的网或网格，因此使用的扇区与具有不同质心的地理特征不匹配。提出了一种最优选择扇区（OSS）算法，会自动将周围区域划分为一组扇区，从而分隔不同的质心，找到这些扇区的最佳高度，并计算重力站的地形影响。此新过程提高了计算的地形效果的准确性。算法内部的适当容差控制方法的准确性。定义此公差取决于重力数据的应用。该方法在具有不同几何形状的合成模型和库尔德斯坦山区的真实数字高程数据的合成模型中的应用显示出地形校正精度的提高。但是，与某些先前的地形校正方法相比，提出的OSS方法引入了额外的计算。