Abstract
Airborne gravity gradient (AGG) measurements offer an increased resolution and accuracy compared to terrestrial measurements. But interpretation and processing of AGG data are often challenging as levelling errors and survey noise affect the data, and these effects are not easily recognised in the gradient components. We adopted the classic method of upward continuation in the noise reduction using the noise level estimates by the AGG system. By iteratively projecting the survey data to a lower level and upward continuing the data back to the survey height, parts of the high-frequency signal are suppressed. The filter, which is defined by this approach, is directly dependent on the noise level of the AGG data, the maximum number of iterations and the iterative step. We demonstrate the method by applying it to both synthetic data and real AGG data over Karasjok, Norway, and compare the results to the directional filtering method. The results show that the iterative filter can effectively reduce high-frequency noise in the data.
A new noise reduction method that iteratively projects data to a lower height and upward continuing the data back to the survey height is described. This method can significantly improve the signal-to-noise ratio of noisy gravity gradient data, and has been successfully applied to both synthetic and real data.
Key words::
Acknowledgements
The Ph.D. project of the first author is funded by the China Scholarship Council. We thank the editor, Mark Pilkington as well as one anonymous reviewer for their helpful comments which resulted in an improvement of our manuscript. We thank the Geological Survey of Norway for making the Karasjok survey dataset available for our study. Data acquisition was funded by the Norske Gullkompaniet and the Geological Survey of Norway as part of the Mineral Exploration in Northern Norway (MINN) project.