WeatherForecast

English · 简体中文

Resolution variants

This public repository contains the existing low-resolution derivatives and their pre-downsampling source files. Source upload is complete (48 files; Hub paths and sizes verified); the low-resolution files remain available. Check Source/manifest.json for the intended inventory and source documentation for completion status.

Component	Pre-downsampling source	Existing low-resolution derivative
GlobalForecast	Source/GlobalForecast: 128 × 256 annual NetCDF, 2010–2018	GlobalForecast: 32 × 64 NPY
TyphoonTracking	Source/TyphoonTracking: 128 × 128 NPZ	TyphoonTracking: 32 × 32 NPZ

“Source” refers to the input of our preparation workflow, not native ERA5 model resolution. Source data files are mirrored without numerical changes. The 45 WeatherBench NetCDF files retain 13 relative-humidity levels, while GlobalForecast uses only 1000 hPa. The 3 TyphoonTracking NPZ files retain upstream data/statistics and lack independent time/coordinate metadata.

GlobalForecast: 128 × 256 → stride-2 sampling → 64 × 128 → non-overlapping 2 × 2 means → 32 × 64; the final hour of each concatenated split is omitted only in the derivative. TyphoonTracking: 128 × 128 → non-overlapping 4 × 4 means → 32 × 32. Low-resolution training statistics are recomputed, not reused from the source grid. See English source card / 中文源数据说明.

Component licenses

WeatherBench data are CC BY 4.0; the Met2Net source dataset declares Apache-2.0, with underlying ERA5/Copernicus attribution and applicable terms retained. These terms apply to the corresponding components and are not a blanket relicense. See LICENSE.md, NOTICE.md, and license texts. Neither the European Commission nor ECMWF is responsible for any use that may be made of the Copernicus information or data it contains.

Low-resolution data

Hourly ERA5-derived weather fields for lower-cost forecasting experiments. Both subsets store unnormalized float32 values and continuous sequences, not pre-extracted windows.

Datasets

	TyphoonTracking	GlobalForecast
Time between consecutive frames	1 hour (source setting)	1 hour (UTC timestamps included)
Grid	32 × 32, approximately 1°	32 × 64, 5.625° spacing
Coverage	110–142°E, 12.25–44°N¹	Global²
Array shape	(4, T, 32, 32) — C,T,H,W	(T, 5, 32, 64) — T,C,H,W
Channel order and units	msl (Pa), u10 (m/s), v10 (m/s), t2m (K)	u10 (m/s), v10 (m/s), t2m (K), tcc (fraction), r at 1000 hPa (%)
Data files	train_32_32.npz, val_32_32.npz, test_32_32.npz	train.npy, val.npy, test.npy
Statistics / coordinates	NPZ keys: data, stats; no coordinates	train_stats.json, coordinates.npz

Split	TyphoonTracking: years¹ / frames	GlobalForecast: years / frames
Train	2017–2021 / 43,824	2010–2015 / 52,583
Validation	2022 / 8,760	2016 / 8,783
Test	2023 / 8,760	2017–2018 / 17,519

¹ Regional bounds and years follow Met2Net's reported setup; source NPZ files contain no coordinates or timestamps for independent verification. Latitude direction is not established.

² Global cell centers: latitude −87.890625° to 86.484375°, south to north; longitude 1.40625° to 355.78125°E. coordinates.npz provides lat, lon, and time_train/val/test. Each split ends at 22:00 UTC on its final day because earlier preparation removed its last hour.

Preparation and use

• TyphoonTracking: 128 × 128 → 32 × 32 by non-overlapping 4 × 4 arithmetic means. Normalize with (raw - min) / (max - min + 1e-8) using the NPZ stats dictionary (min/max/mean/var per channel; reading it requires allow_pickle=True).
• GlobalForecast: WeatherBench's 1.40625° grid was first sampled every second point to 64 × 128, then reduced by 2 × 2 arithmetic means. Normalize with (raw - mean) / std using per-channel [mean, std] in train_stats.json (ddof=0). This is not the official WeatherBench 5.625° product.
• Both use float64 averaging followed by float32 storage, without area weighting or temporal downsampling. Statistics are recomputed from the downsampled training set only and reused for validation/test. Do not refit or automatically clip validation/test values.
• A 12-input / 12-target window covers hours −11…0 → +1…+12. Build windows within each split. Use NumPy to read the files; GlobalForecast NPY files support mmap_mode="r".

Limitations: these are lower-resolution derivatives, not the original Met2Net benchmarks. TyphoonTracking includes non-typhoon periods and has no track labels. Coarse fields smooth cyclone structure; field error does not measure track accuracy.

Sources and terms

ERA5 single levels · ERA5 pressure levels · WeatherBench (Rasp et al., 2020) · Met2Net (Li et al., ICCV 2025) · Authors' ERA5-Cropped release

Independently prepared on 2026-10-23. Contains modified Copernicus Climate Change Service information. Upstream data terms apply; this repository does not relicense the data.