The dataset viewer is not available because its heuristics could not detect any supported data files. You can try uploading some data files, or configuring the data files location manually.

Hong Kong HLS phenology bundle (49QHE)

Path	Description
`meta.json`	Grid size (1830×1830), UTM EPSG:32649 transform, tile id
`imputator_hk/train.nc`	~37,484 pixels × 1342 time steps × 6 optical bands (int16 DN, 0=missing); `band_mean` / `band_std`; lon/lat
`imputator_hk/test.nc`	~4,685 held-out pixels, same schema
`phenology/veg_cube.nc`	Full-grid cube `(time=1342, band=6, y=1830, x=1830)`; `vegetation` mask (stable natural veg, no agriculture); `lum_code`

Bands (6)

Blue, Green, Red, NIR (L30 B05 / S30 B8A), SWIR1, SWIR2 — reflectance × 10000 as int16, 0 = nodata.

Time

3-day compositing steps from HLS granules (~2015–2025); time variable is %Y%j strings.

Vegetation mask

Natural vegetation codes 71–74; agriculture (61) removed; pixels kept only if natural in both 2018 and 2024 LUM and not urban in either year (~750k pixels).

Related model

ZiyunPOLYU/hk-hls-imputator — Transformer gap-filler trained on train.nc.

Usage

hf download ZiyunPOLYU/hk-hls-phenology --local-dir ./hk_hls_data

Point pipeline scripts at HK_DATA_ROOT or symlink to /intelnvme03/... layout:

hk_hls_data/meta.json
hk_hls_data/imputator_hk/train.nc
hk_hls_data/phenology/veg_cube.nc

Downloads last month: 18

Total file size:

3.85 GB