A Robust Generative Adversarial Network Approach for Climate Downscaling and Weather Generation

Dataset Description for "A Robust Generative Adversarial Network Approach for Climate Downscaling and Weather Generation"

This dataset accompanies the research paper titled "A Robust Generative Adversarial Network Approach for Climate Downscaling and Weather Generation", currently under review for the AGU Journal JAMES. The study introduces a novel Regional Climate Model (RCM) emulator focusing on high-resolution climate downscaling for the New Zealand region. For additional insights and access to the codebase utilized in this research, please refer to our GitHub repository.

Aims

Our study's overarching goal was to assess the effectiveness of Generative Adversarial Networks (GANs) in a climate downscaling context and is structured around two aims. The first aim of our study is to examine whether GANs can overcome several important limitations of regression-based climate downscaling algorithms (i.e. underestimating the magnitude of extreme events). The second and most important aim of our study is to assess the robustness GAN performance to different training hyperparameters. Our robustness assessment thoroughly scrutinizes GANs for their application in climate downscaling contexts, ensuring that they can learn and capture regional climate processes

Geographic Focus

Our research focuses only on the New Zealand Region (165°E-184°W, 33°S-51°S).

Data Overview

Training and Evaluation Data

The training data used in this study (for our RCM emulator) only spans the historical period of simulation. It comprises daily accumulated precipitation as the primary target variable, alongside large-scale predictor variables. 

Resolution: The target variable is presented at a 12km resolution, reflecting the highest resolution face of RCM for the New Zealand region. Predictor variables are coarsened to a 1.5-degree resolution from original CCAM outputs using conservative interpolation. 

Period Coverage:

Training Data: 1960-2014

Validation Data: 1986-2005

Models:

Training on: ACCESS-CM2

Validated on: EC-Earth3, NorESM2-MM

File Structure

Training Data:

Target/Ground Truth (Y): predictor_ACCESS-CM2_hist.nc

Predictor (X): pr_ACCESS-CM2_hist.nc

Evaluation Data:

NorESM2-MM:

Target (Y): NorESM2-MM_historical_precip_compressed.nc

Predictor (X): NorESM2-MM_histupdated_compressed.nc

EC-Earth3:

Target: EC-Earth3_historical_precip_compressed.nc

Predictor: EC-Earth3_histupdated_compressed.nc

Methodological Insights

Regional Climate Model, Our Regional Climate Model training data is from the Conformal Cubic Atmospheric Model (CCAM) which is a global non-hydrostatic atmospheric model renowned for its variable-resolution cubic grid. . For more information about CCAM, please see the following paper.

Predictor and Target Variables: Daily-averaged large-scale prognostic variables, including zonal wind, meridional wind, temperature, and specific humidity, are employed as predictors at the 500mb and 850mb pressure levels. These are normalized (see the GitHub repository for the mean and standard deviation fields). Precipitation is taken as is from CCAM and accumulated for each given day. Static predictors are also used in our model, which is stored in a GitHub repository.

Training Framework: Our dataset benefits from the "perfect framework" training strategy, which uses CCAM-coarsened predictor variables. For more information about the perfect and imperfect training frameworks, see the following review