Bathymetry

This repository holds the software for the batheymetry data editing software used by David Sandwell and his students, working in collaboration with Yoav Freund and his students.

Organization of the software

The raw data is in .cm files. There is a master index file called sid_filelist.txt

Initial filtering software

This software is called CM_EDIT. It takes as input a file corresponding to a single cruise. Display the data visually using apple xcode and lets the user flag (9999) the noisy measurements.

Assemble the information from a geographic area.

The information from the edited raw data is windowed and then used to create a full image using interpolation. This is performed by the following software:

maketile script runs the following steps.

1. GMT/block-median: remove outliers using a median filter.
2. GMT/grd_track: removes a model depth from the data (computes residual)
3. GMT/surface: Interpolation of the residual generates a grid of image. (netcdf files)
4. GMT/grdmath: Add the residual grid back the current map.
5. GMT/landmask: creates a grid of land masks. (is it land or ocean).
6. GMT/grdmath: Combines topography and bathymetry based on the land mask.
7. GMT/create the map: produce the geotiff.

Stereo toolkit from NASA

Takes as input the geotiff and produces a pyrmidal representation that can be used in google maps (kml files)

Human correction

A human goes through the google map and marks polygons at the places that are judged incorrect.

Software that takes the polygons and marks (9999) those locations in the measurements (cm files)

Machine Learning Software.

Organization of the repository.

databse

organization

organization	PRIMARY KEY(organization_id)
organization_id	int
name	varchar(255)
access_method	varchar(255)

file_paths

file_paths
source_id	int4
file_path	varchar

pings

pings	PRIMARY KEY(ping_id)
ping_id	int
time	int4
longitude	float8
latitude	float8
depth	float8
sigma_h	float8
sigma_d	float8
source_id	int4
predicted_depth	float8
predicted_bad	float8
organization_id	int

Indexes:

• "pings_pkey" PRIMARY KEY, btree (ping_id)
• "pings__depth_btree_index" btree (depth)
• "pings__source_id_btree_index" btree (source_id)
• "pings_organization_id_btree_index" btree (organization_id)
• "pings_predicted_bad_btree_index" btree (predicted_bad)
• "pings_latitude_btree_index" btree (latitude)
• "pings_longitude_btree_index" btree (longitude)

Human editing software

dev environment installation

First install the Anaconda Python distribution:

https://www.anaconda.com/download/#all

Then in a terminal run:

conda create --name pycmeditor
source activate pycmeditor
conda install -c clinicalgraphics vtk
conda install python.app=1.2
conda install wxpython=4.0.1
conda install folium

To run the app use:

pythonw Py-CMeditor.py

The there is a demo .cm file to load in the human_editing branch.

Notes

Dev branch is called: human_editing

Once software is completed, a pip package will be created for distributing.