ClimX: A challenge for extreme-aware climate model emulation

ClimX is a competition focused on developing fast and accurate machine learning emulators for the NorESM2-MM Earth System Model, with evaluation centered on climate extremes.

What’s in this dataset?

This Hugging Face dataset hosts the ClimX full-resolution training data (historical + projections). A lightweight, $16\times 16\backslash times$ spatially coarsened variant is provided separately for rapid prototyping (hosted on Kaggle).

The full dataset is distributed in NetCDF-4 format to support broad compatibility with common climate tooling.

Problem summary

Participants train emulators that take forcing trajectories (greenhouse gases + aerosols) and optionally past predicted state to produce daily climate fields at the native NorESM2-MM grid (\(192 \times 288\)). The benchmark target is not the raw fields themselves, but 15 extreme indices derived from daily temperature and precipitation.

Primary leaderboard metric (mean standardized MAE over indices):

$$S=\frac{1}{15}\sum _{i=1}^{15}\frac{\mathrm{M}\mathrm{A}\mathrm{E}({\widehat{Y}}_i,Y_i)}{{\sigma }_i}S\; =\; \backslash frac\{1\}\{15\}\backslash sum\_\{i=1\}^\{15\}\backslash frac\{\backslash mathrm\{MAE\}(\backslash hat\{Y\}\_i,\; Y\_i)\}\{\backslash sigma\_i\}$$

Here ${\sigma }_i\backslash sigma\_i$ is computed from the ground-truth $Y_{i}Y\_i$ values for the evaluation split (public/private) and held fixed for all submissions on that split.

Links

• Kaggle competition
• Public code repository (challenge materials)
• Website

License and usage

This dataset is provided for the ClimX competition and associated research/education use. Please follow the competition rules regarding external data/model restrictions and data redistribution.