---
language: en
license: mit
configs:
- config_name: default
  data_files:
  - split: bil
    path: bil/*.parquet
  - split: bon
    path: bon/*.parquet
  - split: bos
    path: bos/*.parquet
  - split: dra
    path: dra/*.parquet
  - split: e13
    path: e13/*.parquet
  - split: fpe
    path: fpe/*.parquet
  - split: gcr
    path: gcr/*.parquet
  - split: gim
    path: gim/*.parquet
tags:
- solar
- radiation
- nsrdb
- bsrn
pretty_name: BSRN NSRDB PSM v4 conus
---

<!-- This file is uploaded as the Hugging Face dataset README; keep content user-facing (no internal maintainer or upload-only notes). -->

# BSRN NSRDB PSM v4 conus

Point-extracted [NSRDB (National Solar Radiation Database)](https://nsrdb.nrel.gov/) PSM v4 (conus variant) data for [Baseline Surface Radiation Network (BSRN)](https://bsrn.awi.de/) stations.

## Dataset Description

Each file contains 5-minute NSRDB variables at a single BSRN station location. **Extraction is performed via the NREL NSRDB API (NLR)**. The `conus` variant (v4.0.0) covers the Contiguous United States using GOES-East and GOES-West satellite data.

### File Structure

Station folders use the **lowercase** BSRN three-letter code (e.g. `bil`).

```
{station}/
  {station}{MM}{YY}_nsrdb_conus.parquet   # One file per month
```

Example:

- `bil/bil0119_nsrdb_conus.parquet` — BIL, January 2019

### Variables

| Column | Description                      | Units         |
|--------|----------------------------------|---------------|
| ghi_nsrdb | Global Horizontal Irradiance  | W/m²          |
| bni_nsrdb | Beam Normal Irradiance (DNI)  | W/m²          |
| dhi_nsrdb | Diffuse Horizontal Irradiance | W/m²          |

- **Index**: UTC `DatetimeIndex` (5-minute resolution).
- **Time coverage**: NSRDB PSM v4 `conus` variant spans 2018–2024.

## Usage

### Load from Hugging Face

```python
from huggingface_hub import hf_hub_download
import pandas as pd

# Download a single file; path is {station}/{station}{MM}{YY}_nsrdb_conus.parquet
path = hf_hub_download(
    repo_id="dazhiyang/bsrn-nsrdb-conus",
    filename="bil/bil0119_nsrdb_conus.parquet",
    repo_type="dataset",
)
df = pd.read_parquet(path)
# df has DatetimeIndex (UTC) and columns: ghi_nsrdb, bni_nsrdb, dhi_nsrdb
```

### Use with bsrn package

The `bsrn` package can fetch NSRDB data from Hugging Face **into RAM** (no disk cache) automatically.

```python
from bsrn.io.nsrdb import add_nsrdb_columns

# Add NSRDB columns to an existing BSRN DataFrame
df = add_nsrdb_columns(df, station_code="BIL", variant="conus")
```

## Data Sources

- **NSRDB**: [NREL National Solar Radiation Database](https://nsrdb.nrel.gov/)
- **API**: [NREL Developer Network (NLR)](https://developer.nrel.gov/)
- **Station inventory**: BSRN FTP

## References

1. Sengupta, M., Xie, Y., Lopez, A., Habte, A., Maclaurin, G., & Shelby, J. (2018). The national solar radiation data base (NSRDB). *Renewable and Sustainable Energy Reviews*, 89, 51-60.

2. Maclaurin, G. J., et al. (2024). NSRDB Physical Solar Model (PSM) v4.0.0.

## Citation

If you use this dataset, please cite the references above and the **bsrn** package: [dazhiyang/bsrn](https://github.com/dazhiyang/bsrn).

## License

NSRDB data is provided by NREL and is subject to their [Data Usage Policy](https://nsrdb.nrel.gov/about/data-usage).