This repository contains a collection of Jupyter notebooks for hands-on atmospheric and climate data analysis using Python. Please report issues, if you find bugs.

Analysis

Basics

• linalg_matrix.ipynb: Basic linear algebra using numpy. Dot product, indentity matrix, inverse maxtrix, determinant, trace, eigenvalues/eigenvectors, Gram matrix

Linear regression

• regression.ipynb: one predictor to one predictand
• regression_multiple.ipynb: multiple predictors to one predictant
• regression_multiple_multiple.ipynb: multiple predictors to multiple predictants

Frequency analysis

• frequency_analysis.ipynb: FFT and power spectrum of a sinusoidal signal
• frequency_analysis_rmm.ipynb: FFT, power spectrum, power spectrum density, relationship with variance, power × frequency, and autocorrelation using RMM index

EOF

• EOF_simple.ipynb: EOF analysis using a simple data
• EOF_sst.ipynb: EOF analysis using 2D SST data containing missing values

Machine Learning

• fcnn.ipynb: Fully connected neural network (FCNN). Forecast Nino 3.4 index using the principal components of SST.

Data

• era5_monthly_sst_5x5.nc: ERA5 monthly sea surface temperature (SST) data from ECMWF, interpolated to a 5° × 5° grid. This dataset covers the tropical region (30°S to 30°N) from 1940 to 2023.
• era5_nino.csv: Nino indices calculated by nino_indices.ipynb.
• era5_monthly_sst_pc.nc: Principal components of ERA5 monthly SST calculated by EOF_sst.ipynb.
• era5_monthly_t2m_points.csv: Time series of ERA5 2m temperature data at 5 cities.
• rmm.csv: Real-time Multivariate MJO (RMM) Index from BoM.