Science and Technology Production
2024 AGU Annual Meeting Preprint Collection on ESS Open Archive - Processing ?Bad Tracks? for Bathymetric Maps with GMT
Congress
Date:
2024Publishing House and Editing Place:
American Geophysical UnionSummary *
For the creation of global or regional-scale bathymetric grids (e.g., SRTM15_plus, GEBCO), different sources of direct data are used, including single-beam bathymetric profiles. Gridding methods use the position and depth values (XYZ) from the direct data and produce a grid through a gridding algorithm. If input data contains errors, these may be visible in the resulting grid. The standard procedure for these cases is to label these surveys as "bad tracks´´ and exclude them from the gridding algorithm. In this work, results are presented on a methodology to reuse these bad tracks. Instead of removing the data, we calculate the gradient and azimuth along the “bad track” and integrate them with the rest of good data with the greenspline module from The Generic Mapping Tools. First, some tests were carried out with theoretical surfaces. An inclined surface and a more complex surface (“hat”) were analyzed. For the inclined plane surface, this new method was able to accurately regenerate the original grid. In contrast, for the "hat" surface, the grid generated by this new method had some errors although these were less significant than those obtained by the standard method. In this presentation, we show the results of a real-world example of the Pacific Ocean (see figure). To test it, we took a track, and added a constant value to create a “bad track” (red dots in the figure). We created four grids. The first one uses the original data (without error) and can be considered as ground-truth. In the second grid, we used the standard method including the “bad track”. For the third grid, we applied the method described here. For the fourth grid we did the same but only retained 90% of the largest eigenvalues. The results show that with this new method, we obtained a better-looking result compared to the standard method. When all the eigenvalues were considered, there were some high-frequency features related to the “bad track”. However, these features were removed when only the largest eigenvalues were used. These results suggest that this methodology could allow the re-use of bad-tracks for the elaboration of bathymetric grids. These results are expected to be very useful for shelf areas where multibeam surveys are not efficient.https://agu.confex.com/agu/agu24/meetingapp.cgi/Paper/1706895 Information provided by the agent in SIGEVAKey Words
ProcessingBathymetryGridding