frontpy 0.1.14

FrontPy is a Python package designed for the detection and analysis of atmospheric fronts. It uses the Thermal Front Parameter (TFP) method to identify cold and warm fronts. The package also includes visualization capabilities, allowing users to plot the identified fronts over GOES-16 satellite imagery.

FrontPy

FrontPy is a Python package designed for the detection and analysis of atmospheric fronts. It uses the Thermal Front Parameter (TFP) method to identify cold and warm fronts. The package also includes visualization capabilities, allowing users to plot the identified fronts over GOES-16 East satellite imagery.

Installation of FrontPy

Step 1: Create a new Conda environment

First, you should create a new environment with the specific version of Python and some necessary dependencies. Run the following command in your terminal or command prompt:

conda create -n your_environment_name python=3.9.15 gdal=3.3.3 poppler=21.09 -c conda-forge

Replace your_environment_name with your desired name for the environment. The conda will create an isolated environment with Python 3.9.15 and the gdal and poppler libraries, which are necessary for the functionality of FrontPy. Now, FrontPy will probably run smoothly without any errors.

Step 2: Activate your new Conda environment

Once the environment is created, activate it using the command:

conda activate your_environment_name

Step 3: Install FrontPy using pip

Now that you are in an isolated environment, install FrontPy using pip:

pip install frontpy

Usage

To use FrontPy, follow these steps:

Step 1: Import the main and plot_results functions from the package:

from frontpy import main, plot_results

Step 2: Set up the parameters required for downloading the data, calculating the Thermal Front Parameter (TFP), detecting cold and warm fronts, and final visualization.

The parameters must be provided in a dictionary, like this:

config = {
    "start_date": "2024-10-11 12",  # Start date and time for the analysis (format: YYYY-MM-DD HH) - string
    "end_date": "2024-10-12 12",    # End date and time for the analysis (format: YYYY-MM-DD HH) - string
    "lat_max": 20,                  # Maximum latitude for the analysis area (degrees, range: -90 to 90) - float
    "lat_min": -60,                 # Minimum latitude for the analysis area (degrees, range: -90 to 90) - float
    "lon_max": -20,                 # Maximum longitude for the analysis area (degrees, range: -180 to 180) - float
    "lon_min": -90,                 # Minimum longitude for the analysis area (degrees, range: -180 to 180) - float
    "model_name": None,            # Name of the dataset to be used. (Options available: ERA5, GFS). Leave it unchanged if you want to use your own files - string
    "filepath": "path_to_your_netcdf_file", #Path to the local NetCDF file containing both u, v, specific humidity and temperature (required if model_name is None)
    "pressure": 850,                # [Default: 850] Atmospheric pressure level (hPa) at which front identification is performed - int
    "thetaw_thresh": 3.0,           # [Default: 3.0] Threshold for the magnitude of the gradient of potential wet-bulb temperature (K/100km) for front identification -  float
    "vf_thresh": 1.0,               # [Default: 1.0] Threshold for wind velocity (m/s) for front classification as cold or warm front - float
    "smooth_sigma": 0.5,            # [Default: 0.5] Smoothing parameter (sigma) for Gaussian filtering - float
    "line_or_area": "area",         # Type of representation for the fronts (line or area) - string
    "min_points": 4,                # [Default: 4] Minimum number of frontal points required for a valid frontal line - int
    "min_length": 500.0,            # [Default: 500.0] Minimum length (km) for the frontal line to be considered - float
    "min_area": 5000.0,             # [Default: 5000.0] Minimum area (km²) for the frontal zone to be considered - float
    "frame_rate": 3.0,              # [Default: 3.0] Frame rate (frames per second) for the generated animation; higher rates result in a faster animation - float
    "output_directory_fronts": "path_to_your_chosen_output_directory"   # Path to your output directory
}

Currently, FrontPy allows you to automatically download ERA5 and GFS data (u, v, temperature, specific humidity) for calculating TFP and identifying fronts. To do this, enter "ERA5" or "GFS" in the model_name parameter.

If you want to use your own files, all variables must be contained in one netcdf file. In this case, do not assign any value for the model_name parameter. Instead, you must provide the path to the netcdf file in the filepath parameter, which should include u, v, specific humidity and temperature.

Note: The default values have only been tested for the Southern Hemisphere.

Step 3: Call the main function:

The main function will return two Pandas dataframes: one for cold fronts and one for warm fronts.

So, you can call the main function like this:

cold_fronts, warm_fronts = main(config)

Step 4: Plot the images and generate an animation:

Lastly, you can plot the results and generate an animation (GIF) using the plot_results function. The images and animation will be saved in your output directory.

plot_results(cold_fronts, warm_fronts, config)

Obs: For now, it is only possible to generate images within the coverage area and the available data period of GOES-16 East.

Command Line Interface (CLI)

FrontPy also supports command-line execution through the cli.py file and the fronts function. This allows users to specify the parameters directly from the command line.

To run FrontPy using the CLI, execute the following command in your terminal (without the <>):

fronts <start_date> <end_date> <lat_max> <lat_min> <lon_max> <lon_min> <line_or_area> <output_directory_fronts> <mode>

The above arguments are mandatory and must be provided in string format (in quotes) and in the specified order.

Mode Argument Explanation:

• To only identify and generate the files and dataframes relative to the TFP field and to cold and warm fronts, type "identify" as the <mode> argument.

• To only create the images and animation based on the files and dataframes that have already been generated, type "plot" as the <mode> argument.

• To perform both actions type "both" as the <mode> argument.

Optional arguments can also be included using the following prefixes before the values of your choice:

• --model-name
• --filepath
• --pressure
• --thetaw-thresh
• --vf-thresh
• --smooth-sigma
• --min-points
• --min-length
• --min-area
• --frame-rate

Real example using CLI:

1. Using GFS data to identify and plot the results:

fronts "2024-10-10 00" "2024-10-11 00" "20" "-60" "-20" "-90" "line" "/output_directory/" "both" --model-name "GFS"

2. Using my own data:

fronts "2024-10-10 00" "2024-10-11 00" "20" "-60" "-20" "-90" "line" "/output_directory/" "both" --filepath "/path_to_my_netcdf_file.nc"

License

This project is licensed under the Apache License 2.0. See the LICENSE file for more details.