juliensimon/hst-observations

---
license: cc-by-4.0
pretty_name: "Hubble Space Telescope Observation Catalog"
language:
- en
description: "The Hubble Space Telescope Observation Catalog is a complete index of every observation obtained by NASA/ESA's Hubble Space Telescope since its launch on April 24, 1990, drawn from the Mikulski Archiv"
task_categories:
- tabular-classification
tags:
- space
- hubble
- hst
- nasa
- esa
- astronomy
- telescope
- open-data
- tabular-data
- parquet
size_categories:
- 1M<n<10M
configs:
- config_name: default
data_files:
- split: train
path: data/hst_observations.parquet
default: true
---

# Hubble Space Telescope Observation Catalog

<div align="center">
<img src="banner.jpg" alt="The Hubble Space Telescope being deployed from the Space Shuttle Discovery in 1990" width="400">
<p><em>Credit: NASA</em></p>
</div>

*Part of a [dataset collection](https://huggingface.co/collections/juliensimon/astronomy-datasets-69c24caf2f17e36128946743) on Hugging Face.*

## Dataset description

The Hubble Space Telescope Observation Catalog is a complete index of every observation obtained by NASA/ESA's Hubble Space Telescope since its launch on April 24, 1990, drawn from the Mikulski Archive for Space Telescopes (MAST). Hubble's 2.4-meter primary mirror and suite of instruments have produced one of the most scientifically productive archives in astronomy, with over 35 years of continuous operation in low Earth orbit.

Each row in this catalog is one HST observation — a unit of telescope time executing an exposure with a specific instrument, detector, filter, and target pointing. Rows include the proposal under which the observation was taken (proposal ID, PI, title, category: GO, SNAP, GTO, DDT, CAL), the target (name, coordinates, moving/fixed flag), the instrument and detector (ACS WFC/HRC/SBC, WFC3 UVIS/IR, WFPC2 WF/PC, STIS, COS, NICMOS, plus legacy FOC/FOS/HRS/FGS), and the observation intent.

This dataset is the canonical reference for answering questions like: what has Hubble observed near a given RA/Dec? Which proposals used STIS for UV spectroscopy? Which targets have the deepest imaging coverage? It is designed for cross-matching with target catalogs (galaxies, quasars, stars, solar system bodies), for program-level summaries, for planning parallel JWST follow-up, and as training data for observation-recommendation systems.

This v1 provides observation-level metadata only. Per-observation timing and exposure data — which require joining with MAST's `dbo.caomplane` table (4.6M rows) — will arrive in a v2 as we build an async batched pipeline. For now, detailed timing/filter information can be retrieved per observation via the MAST Portal or the `astroquery.mast` Python package.

The catalog is derived from MAST's CAOM (Common Archive Observation Model) table `dbo.caomobservation` and is refreshed weekly as HST observations enter the archive. Calibration and engineering observations are included but are distinguishable via the `intent` column.

This dataset is suitable for **tabular classification** tasks.

## Schema

| Column | Type | Description | Sample | Null % |
|--------|------|-------------|--------|--------|
| `obs_id` | string | MAST observation identifier (e.g., 'hst_05773_54_wfpc2_wf_pc_f547m'); encodes proposal, visit, instrument, aperture, and filter. Primary key. | f05i0201m | 0.0% |
| `obstype` | string | CAOM observation type code: 'S' (simple), 'C' (composite) | S | 0.0% |
| `intent` | string | Observation intent: 'science' or 'calibration' | science | 0.0% |
| `proposal_id` | string | HST proposal identifier (string, e.g., '5773'); groups related observations by Principal Investigator's program | 1538 | 3.3% |
| `proposal_pi` | string | Last name, first initial of the proposal Principal Investigator | BRADY | 3.3% |
| `proposal_title` | string | Full title of the HST observing proposal | PRE-ALIGNMENT OTA CALIBRATION | 34.1% |
| `proposal_project` | string | Proposal project code (e.g., 'GO', 'GTO', 'SNAP', 'DDT', 'CAL'); GO = General Observer, SNAP = snapshot survey, DDT = Director's Discretionary, CAL = calibration | OV/OTA | 3.5% |
| `target_name` | string | Target name as provided by the proposer (may include survey designations, coordinates, or informal names) | 1538_2 | 0.0% |
| `target_ra` | float64 | Target right ascension in decimal degrees (ICRS). May be 0 for moving or calibration targets. | 165.9390011752 | 0.3% |
| `target_dec` | float64 | Target declination in decimal degrees (ICRS). May be 0 for moving or calibration targets. | -58.28425312885 | 0.3% |
| `target_moving` | bool | True if the target is a moving solar system body (asteroid, comet, planet, moon); False for fixed celestial targets | False | 0.0% |
| `instrument` | string | Instrument name: ACS, WFC3, WFPC2, STIS, COS, NICMOS, FOC, FOS, HRS, FGS (from 'INSTRUMENT/DETECTOR' split) | FGS | 0.0% |
| `detector` | string | Detector or observing mode within the instrument (e.g., 'WFC', 'UVIS', 'IR', 'HRC', 'SBC', 'PC', 'WF'); from 'INSTRUMENT/DETECTOR' split | WFC | 4.8% |

## Quick stats

- **2,627,714** HST observations (1990–present)
- **2,165,235** science, **462,479** calibration
- **12,277** distinct proposals
- Top instruments: **WFC3** (896,172), **ACS** (542,987), **WFPC2** (377,101), **NICMOS** (345,395), **STIS** (244,163)

## Usage

```python
from datasets import load_dataset

ds = load_dataset("juliensimon/hst-observations", split="train")
df = ds.to_pandas()
```

```python
from datasets import load_dataset

ds = load_dataset("juliensimon/hst-observations", split="train")
df = ds.to_pandas()

# Science observations with WFC3 UVIS detector
import pandas as pd
wfc3_uvis = df[(df["intent"] == "science") & (df["instrument"] == "WFC3") & (df["detector"] == "UVIS")]
print(f"WFC3 UVIS science observations: {len(wfc3_uvis):,}")

# Proposals per decade
df["decade"] = df["proposal_id"].astype(str).str[:1].replace({
"5": "1990s", "6": "1990s-2000s", "7": "1990s-2000s",
"8": "2000s", "9": "2000s", "1": "2000s-2020s",
})
df.groupby("instrument")["proposal_id"].nunique().sort_values(ascending=False).head(15).plot.bar()
import matplotlib.pyplot as plt
plt.ylabel("Distinct proposals")
plt.title("HST proposal count by instrument")
plt.show()

# Cone search around a target (Hubble Deep Field)
import numpy as np
ra, dec = 189.139, 62.217
sep = np.hypot(df["target_ra"] - ra, df["target_dec"] - dec)
nearby = df[sep < 0.1]
print(f"HST observations within 0.1 deg of HDF: {len(nearby):,}")

# Instrument usage pie
df["instrument"].value_counts().head(10).plot.pie(autopct="%1.1f%%")
plt.title("HST instrument usage (observation count)")
plt.show()
```

## Data source

https://archive.stsci.edu/

## Update schedule

Weekly (Monday at 13:30 UTC) via [GitHub Actions](https://github.com/juliensimon/space-datasets).

## Related datasets

- [juliensimon/jwst-observations](https://huggingface.co/datasets/juliensimon/jwst-observations)

- [juliensimon/chandra-x-ray-sources](https://huggingface.co/datasets/juliensimon/chandra-x-ray-sources)

- [juliensimon/erosita-erass1-xray](https://huggingface.co/datasets/juliensimon/erosita-erass1-xray)

- [juliensimon/4xmm-dr14-xray-sources](https://huggingface.co/datasets/juliensimon/4xmm-dr14-xray-sources)

- [juliensimon/nasa-exoplanets](https://huggingface.co/datasets/juliensimon/nasa-exoplanets)

> If you find this dataset useful, please consider [giving it a like](https://huggingface.co/datasets/juliensimon/hst-observations) on Hugging Face. It helps others discover it.

## About the author

Created by [Julien Simon](https://julien.org) — AI Operating Partner at Fortino Capital. Part of the [Space Datasets](https://julien.org/datasets) collection.

## Citation

```bibtex
@dataset{hst_observations,
title = {Hubble Space Telescope Observation Catalog},
author = {juliensimon},
year = {2026},
url = {https://huggingface.co/datasets/juliensimon/hst-observations},
publisher = {Hugging Face}
}
```

## License

[CC-BY-4.0](https://creativecommons.org/licenses/by/4.0/)