Searching for comets on the World Wide Web: The orbit of 17P/Holmes from the behavior of photographers

TL;DR

This study leverages web-sourced astronomical images to infer the orbit of comet 17P/Holmes, demonstrating the potential of heterogeneous online data for scientific discovery and measurement.

Contribution

The paper introduces a Bayesian method to determine comet orbits using publicly available, heterogeneous images from the Web, without relying on traditional observational data.

Findings

Strong probabilistic constraints on the comet's orbit.

Approximately 70% of image meta-data are accurate.

The comet is typically centered in the middle third of images.

Abstract

We performed an image search for "Comet Holmes," using the Yahoo Web search engine, on 2010 April 1. Thousands of images were returned. We astrometrically calibrated---and therefore vetted---the images using the Astrometry.net system. The calibrated image pointings form a set of data points to which we can fit a test-particle orbit in the Solar System, marginalizing over image dates and detecting outliers. The approach is Bayesian and the model is, in essence, a model of how comet astrophotographers point their instruments. In this work, we do not measure the position of the comet within each image, but rather use the celestial position of the whole image to infer the orbit. We find very strong probabilistic constraints on the orbit, although slightly off the JPL ephemeris, probably due to limitations of our model. Hyperparameters of the model constrain the reliability of date meta-data and where in the image astrophotographers place the comet; we find that ~70 percent of the meta-data are correct and that the comet typically appears in the central third of the image footprint. This project demonstrates that discoveries and measurements can be made using data of extreme heterogeneity and unknown provenance. As the size and diversity of astronomical data sets continues to grow, approaches like ours will become more essential. This project also demonstrates that the Web is an enormous repository of astronomical information; and that if an object has been given a name and photographed thousands of times by observers who post their images on the Web, we can (re-)discover it and infer its dynamical properties.