This Python code is designed for constructing the Wheeler-Kiladis Space-Time Spectra (WK spectra) from gridded time series data, focusing on the analysis of tropical waves. It is based on various functions found in the following GitHub repositories:
This code is based on the functions provided in the following GitHub repositories:
The purpose of this implementation is to provide a Python-based solution for computing the WK Space-Time Spectra while optimizing computational performance. Adjustments have been made to improve the speed of the calculations, resulting in about a 2-second improvement compared to earlier versions. Although the results might slightly differ from those generated by the official NCL website or the original implementations, the extraction of symmetric and antisymmetric signals from the overall wave fluctuations remains consistent when compared to NCL results.
-
Speed Optimization: Adjustments to the original code have reduced computational time by approximately 2 seconds, making this version more efficient for large datasets.
-
Straightforward Implementation: This version focuses on providing a clear, easy-to-understand implementation of the WK Space-Time Spectra calculation without complex encapsulation. It is ideal for those wishing to understand the process of WK spectrum calculation.
- Python_wk_spacetime.ipynb: The main Python code used to compute the WK Space-Time Spectra.
The output consists of visual plots showing the space-time spectra of tropical wave fluctuations, extracted symmetrically and antisymmetrically.
Generated by the NCL script: wk_time4.ncl
pip install numpy matplotlib xarray scipy netCDF4Download the Uninterpolated OLR data from NOAA's website: https://psl.noaa.gov/data/gridded/data.uninterp_OLR.html
Load the Python_wk_spacetime.ipynb Jupyter notebook.
Follow the instructions in the notebook to load the dataset and run the analysis.
You can run the script in vscode by installing the jupyter extension