Making a T-S Diagram: Python vs Interactive Tools

You can make a T-S diagram with Python or an interactive browser tool. Python suits custom calculations and batch processing; tools such as OceanGraph suit profile search and comparison.

If you want to understand how to read the finished plot, pair this article with T-S Diagrams in Oceanography Explained (With Examples).

What You Need Before Making a T-S Diagram

No matter which workflow you use, the basic ingredients are similar.

You need:

• Profile data with temperature and salinity information
• Enough context to know where and when the profile was collected
• A way to compare one profile or multiple profiles
• A clear reason for looking in T-S space rather than depth space alone

In practice, it also helps to understand the profile first. If you have not yet done that, How to Read Argo Float Data for Beginners (What to Look at First) and Ocean Temperature and Salinity Profiles Explained are useful preparation.

For interpretation, check the exact variables rather than relying on the title alone. T-S may use in-situ temperature and Practical Salinity; classical θ-S uses potential temperature and a stated salinity scale; TEOS-10 thermodynamic work generally uses Conservative Temperature and Absolute Salinity. OceanGraph’s θ-S Diagram specifically uses potential temperature and Absolute Salinity.

The Python Workflow

Python is a powerful way to make a T-S diagram, especially if you need full control.

Downloading the data

The first step is usually to obtain the relevant profile data.

That may mean:

• Downloading Argo profile files
• Selecting a float by WMO ID
• Narrowing the data by region or date
• Deciding whether one profile or many profiles should be included

Reading NetCDF files

Once you have the files, you need to open them and identify the variables you actually want to plot.

In a Python workflow, that usually means working with:

• NetCDF structures
• Array dimensions
• Metadata conventions
• Quality information

Cleaning temperature and salinity values

Before plotting, you often need to decide:

• Whether the Argo data mode requires raw or adjusted variables
• Which temperature and salinity definitions the axes will use
• Which matching quality flags and error estimates to apply
• How to handle missing values
• Whether to calculate potential or Conservative Temperature and Practical or Absolute Salinity

Converting Practical Salinity to Absolute Salinity requires pressure, longitude, and latitude; temperature conversions require pressure and a clearly stated reference or TEOS-10 convention.

Plotting the diagram

• Put salinity on the horizontal axis
• Put temperature or potential temperature on the vertical axis
• Plot one profile or many profiles
• Add potential-density contours calculated from variables consistent with the axes and a stated reference pressure, if needed
• Adjust labels, ranges, and styling

The Interactive Workflow

An interactive workflow flips the order. Instead of preparing code first, you inspect the data first.

Search for a float or profile

Start by narrowing the profiles you care about.

In OceanGraph, this means using the search workflow to filter by:

• Region
• Date range
• WMO ID
• Available profile characteristics

The search entry point is here:

• Search and Bookmark

Open the T-S view

Once the search results are in place, you can open the θ-S diagram view.

OceanGraph generates the diagram from the current search area, which means the T-S view stays connected to the actual profile set you are exploring rather than being detached from context.

The feature overview is here:

• θ-S Diagram

Compare profiles without writing code

In OceanGraph, you can:

• Search for relevant profiles
• Read the vertical profile first
• Open the θ-S view second
• Compare selected profiles against the broader background distribution

If you want the no-code exploration workflow in more detail, see Visualizing Argo Float Data Without Python (Step-by-Step Guide).

Python vs Interactive Tools: A Practical Comparison

Question	Python workflow	Interactive workflow
Time to first diagram	Slower	Faster
Setup required	Higher	Lower
Control over plotting details	Very high	Moderate
Best for screening profiles	Less convenient	Very convenient
Best for batch processing	Strong	Limited
Best for teaching or first interpretation	Often heavier	Strong
Best for final custom analysis	Strong	Usually a first step

When Python Is Still the Right Choice

Python remains the better option when you need to:

• Process many profiles in bulk
• Reproduce a figure exactly in a script
• Add custom derived calculations
• Merge Argo data with other datasets
• Generate publication-ready plots in a controlled pipeline

When an Interactive Tool Is the Better First Step

An interactive tool is often the better first step when:

• You are learning how T-S diagrams work
• You want to inspect real data quickly
• You are screening profiles before deeper analysis
• You need to compare observations without building a full code workflow
• You want to connect the T-S view directly to the vertical profile and search context

Example Workflow: From First Look to Deeper Analysis

A pragmatic workflow often looks like this:

1. Search for profiles in OceanGraph by region, date, or WMO ID.
2. Check the observation context.
3. Open the vertical profiles and identify which cases are interesting.
4. Open the θ-S diagram to examine the property-space structure.
5. Select profiles that show distinct layering, clustering, or mixing-like behavior.
6. Return to Python later only for the subset that deserves custom analysis.

OceanGraph’s related pages for this workflow are:

• Search and Bookmark
• Analysis Lab: Vertical Profiles (JSON upload)
• θ-S Diagram
• T-S Diagrams in Oceanography Explained (With Examples)