GNDC-HiGLASS-LS20: Gap-free 20 m 5-Day LAI/FAPAR over China, 2018–2023
# GNDC-HiGLASS-LS20
GNDC-compressed version of **HiGLASS-LS20** — the first gap-free, 20 m, 5-day
LAI / FAPAR product over China (2018–2024), derived from integrated
Landsat-8/9 and Sentinel-2 Analysis Ready Data.
Files are stored as `.gndc` (a neural data cube format; see GeoNDC reference
below) instead of GeoTIFF to reduce size. **A decoder is required to read them.**
## Decode
```bash
pip install pygndc
```
Documentation: https://github.com/jianboqi/pygndc/blob/main/TUTORIAL.md
### Minimal example
```python
import pygndc
# Decode a .gndc file. The output is already in PHYSICAL values
# (no scale factor needs to be applied).
data = pygndc.read("HiGLASS_LS20_n018e108_2020_v01_LAI.gndc")
print(data.shape) # (73, rows, cols) — 73 five-day composites
print(data[0]) # first composite (DOY 001), LAI in m² m⁻²
```
> Note: after decoding, values are already physical (LAI in m² m⁻²,
> FAPAR unitless). Do **not** multiply by the scale factors below — those
> apply only to the raw uint8 GeoTIFF source and are listed for reference.
## File naming
```
GNDC-HiGLASS_LS20_[lat][lon]_[year]_v01_LAI.gndc
GNDC-HiGLASS_LS20_[lat][lon]_[year]_v01_FAPAR.gndc
```
- Tiles are 1° × 1° (WGS84 / EPSG:4326), centred at integer degree + 0.5°
(e.g. `n018e108` = 18.5°N, 108.5°E, covering ~18–19°N, 108–109°E).
- 73 bands per file = 5-day composites: DOY 001, 006, 011, …, 361
(Jan 1, Jan 6, …, Dec 22).
## Physical values & ranges
| Variable | Range | Unit |
|----------|-------|------|
| LAI | 0 – 10.0 | m² m⁻² |
| FAPAR | 0 – 1.0 | unitless |
(For reference, the original uint8 GeoTIFF used scale factors LAI ×0.1 and
FAPAR ×0.004. pygndc returns decoded data already in the physical ranges above.)
## Extend to other regions (2019–2024)
Open-source pipeline (Google Colab):
https://colab.research.google.com/drive/1zGYbUkh4COoZqpWA2tfMLZHVfsGZ1nAo
## Citation
**For the GNDC compression format / decoder:**
Qi, J., M. Li, B. Jiang, Y. Chen, and Q. Wang (2026).
"GeoNDC: A Queryable Neural Data Cube for Planetary-Scale Earth Observation."
arXiv:2603.25037 [cs.CV]. https://arxiv.org/abs/2603.25037
```bibtex
@misc{qi2026geondc,
title={GeoNDC: A Queryable Neural Data Cube for Planetary-Scale Earth Observation},
author={Jianbo Qi and Mengyao Li and Baogui Jiang and Yidan Chen and Qiao Wang},
year={2026},
eprint={2603.25037},
archivePrefix={arXiv},
primaryClass={cs.CV},
url={https://arxiv.org/abs/2603.25037}
}
```
**For the underlying HiGLASS-LS20 product:**
Ma, H., Q. Wang, W. Li, Y. Chen, J. Xu, Y. Ma, J. Huang, and S. Liang (2025).
"The first gap-free 20 m 5-day LAI/FAPAR products over China (2018–2023) from
integrated Landsat-8/9 and Sentinel-2 Analysis Ready Data."
*Remote Sensing of Environment*, 331, 115048.
DOI: 10.1016/j.rse.2025.115048
## Contact
- Original product: Dr. Han Ma — mahan@hku.hk
- GNDC packaging & pygndc decoder: Jianbo Qi — jianboqi@bnu.edu.cn