位场导数换算是重磁数据处理与解释的基本方法,但其不稳定性限制了在实际资料中的应用,且目前的改进措施并未能较好地解决导数计算精度和不稳定性之间的矛盾问题。针对于此,本文在波数域导数迭代法基础上,提出了迭代差分视导数。该方法是在导数迭代法的信号压制区内,通过对不同迭代次数计算结果之差进行合理修正实现的;同时对比分析了视导数与常规导数算子的滤波特性;并通过导数模型实验验证了视导数具有较强的计算稳定性和较高的异常分辨率。另外,本文还对三种常用边界识别方法进行了改进来进一步突出场源边界的异常特征,并将视导数应用到了这些边界识别方法之中。组合模型试验表明,本文提出的边界识别方法改进措施有效地提高了原方法对边界的精细检测能力,基于视导数的边界识别方法则不仅具有较强的抗干扰能力,而且更加精确地检测出了地质体的边界。最后,将视导数和基于视导数的边界识别方法应用到了黑龙江虎林盆地重力异常中,结果显示视导数不仅具有较强的计算稳定性,而且清晰地展示出了异常间的相互关系;同时,基于视导数的边界识别方法准确、清晰地识别出了地质体的边界,还展示出了更多的边界细节信息。
Derivative calculation is a basic method used for data processing and geological interpretation in potential fields,but all derivatives are high-pass filters and easily lead to unstable results in application. What 's more, almost all improved methods for derivative calculation have not resolved the contradiction problem between computational accuracy and instability at the same time.For this reason,this paper proposes apparent derivative calculated using the iterative differential which is based on iterative method in wave number domain in potential fields. The method is accomplished by the difference among results of different iterations in signal suppression zone of the iterative method. Comparative analysis of filtering characteristic between apparent derivative and conventional derivative shows that the apparent derivative has stronger stability in principle. Model test indicates that the apparent derivatives have higher resolution and less sensitivity to random noise than the conventional derivatives. Moreover, this paper presents three improved edge-detection methods which are usually used in potential-field and applies the apparent derivatives in these edge-detection methods. Combined model tests show that improved edge-detection methods presented by this paper effectively improve detectability of original methods, and edge recognition methods based on the apparent derivatives not only have stronger ability of anti-interference,but can precisely detect the edges of geological bodies. At last, we apply the apparent derivatives and edgedetection methods based on the apparent derivatives in gravity data of Hulin Basin,Heilongjiang province of China. The results show that the apparent vertical derivate has strong stability and clearly reflect the relationship among difference anomalies, and edgedetection methods based on apparent derivatives can accurately and clearly detect boundaries of geologic bodies,and can produce more detailed results.