Description of data processing :

https://www.gmrt.org/about/index.php

How are swath data processed?

We use a combination of software developed in-house and open source software to process swath data for GMRT. MB-System provides the backbone of the code that we use for our data cleaning and processing. Raw multibeam sonar ping files available for each cruise are first run through a series of crude filters (e.g. ocean depth), and are then used to generate a development set of 100-m resolution tiled images and grids. These images and grids are reviewed and quantitatively interrogated using a custom version of GeoMapApp. This allows our team to assess data coverage and data quality, and to identify problems in the data that can be addressed using multibeam processing tools and techniques available through MB-System.

Data are typically processed to 100-m resolution. While our overall goal is to include only high-quality data, we sometimes include lower quality data in areas where data coverage is sparse (e.g. Southern Ocean). The data are reviewed, edited as necessary and re-gridded iteratively until the quality is suitable for inclusion in the tile set. During this process, we also determine the maximum data resolution for the cruise, the data are either gridded at 100m (default), 50m, and 25m resolution depending on sonar system capabilities and survey area. The chosen resolution is applied to the entire cruise and is therefore limited by maximum water depth. During QA/QC we also assign a quality value to the cruise on the range of 1-10 with 10 being applied to the highest quality cruises. This value is applied to the data and affects the degree of blending that happens when the data are merged into the compilation.

After extraction from the GMRT website, the grid is converted in the specified format of the EMODnet Bathymetry project before integration.

Multisources bathymetry

Data sources for GMRT grid is described in https://www.gmrt.org/contributors/index.php