dustmaps3d 2.1.0

An all-sky 3D dust map based on Gaia and LAMOST.

Read this in: 中文

dustmaps3d

🌌 An all-sky 3D dust extinction map based on Gaia and LAMOST

📄 Wang et al. (2025), An all-sky 3D dust map based on Gaia and LAMOST
📌 DOI: 10.12149/101620

📦 A Python package for easy access to the 3D dust map
📌 DOI: 10.12149/101619

---

📦 Installation

Install via pip:

pip install dustmaps3d

Note: Installing the package does not include the data file.
The ~350 MB model data will be automatically downloaded from GitHub on first use.
⚠️ If you experience network issues when downloading from GitHub,
you can manually download the data from NADC:
🔗 https://nadc.china-vo.org/res/r101619/

---

🚀 Usage

from dustmaps3d import dustmaps3d

l = [120.0]
b = [30.0]
d = [1.5]

EBV, dust, sigma, max_d = dustmaps3d(l, b, d)

print(f"EBV: {EBV[0]:.4f} [mag]")
print(f"Dust: {dust[0]:.4f} [mag/kpc]")
print(f"Sigma: {sigma[0]:.4f} [mag]")
print(f"Max distance: {max_d[0]:.4f} kpc")

Batch example with FITS:

import numpy as np
from astropy.io import fits
from astropy.table import Table
from dustmaps3d import dustmaps3d

data = Table.read('input.fits')
l = data['l'].astype(float)
b = data['b'].astype(float)
d = data['distance'].astype(float)

EBV, dust, sigma, max_d = dustmaps3d(l, b, d)

data['EBV_3d'] = EBV
data['dust'] = dust
data['sigma'] = sigma
data['max_distance'] = max_d

data.write('output.fits', overwrite=True)

---

🧠 Function Description

dustmaps3d(l, b, d, n_process=None)

Estimates 3D dust extinction and related quantities for given galactic coordinates and distances.

Input	Type	Description	Unit
l	np.ndarray	Galactic longitude	degrees
b	np.ndarray	Galactic latitude	degrees
d	np.ndarray	Distance	kpc
n_process	int, optional	Number of processes for parallel computing	–

Returns:

Output	Type	Description	Unit
EBV	np.ndarray	E(B–V) extinction	mag
dust	np.ndarray	Dust density (d(EBV)/dx)	mag/kpc
sigma	np.ndarray	Estimated uncertainty in E(B–V)	mag
max_d	np.ndarray	Maximum reliable distance	kpc

If d contains NaN, it will be automatically replaced by the maximum reliable distance along that line of sight.

---

⚡ Performance

• Fully vectorized and optimized with NumPy
• Supports multiprocessing with n_process argument
• On a modern personal computer, evaluating 100 million stars takes only ~100 seconds

---

📂 Data Version

This version uses data_v2.1.parquet, released under v2.1

---

📜 Citation

If you use this model or the Python package, please cite both:

• Wang, T. (2025), An all-sky 3D dust map based on Gaia and LAMOST. DOI: 10.12149/101620
• dustmaps3d: A Python package for easy access to the 3D dust map. DOI: 10.12149/101619

---

🛠️ License

MIT License