Mining the GPIES database

TL;DR

This paper presents the GPIES database, a comprehensive system for managing and analyzing data from the Gemini Planet Imager Exoplanet Survey, enabling insights into instrument performance and data quality.

Contribution

The paper introduces a detailed database system for GPIES, integrating raw and reduced data with metadata, and demonstrates its use in data exploration, visualization, and predictive modeling.

Findings

Correlations between instrument performance and environmental conditions identified.

Automated outlier rejection improves data analysis accuracy.

Supervised learning helps predict final data quality from early observations.

Abstract

The Gemini Planet Imager Exoplanet Survey (GPIES) is a direct imaging campaign designed to search for young, self-luminous, giant exoplanets. To date, GPIES has observed nearly 500 targets, and generated over 30,000 individual exposures using its integral field spectrograph (IFS) instrument. The GPIES team has developed a campaign data system with a database incorporating all of the metadata for all individual raw data products, including environmental conditions and instrument performance metrics. The same database also indexes metadata associated with multiple levels of reduced data products, including contrast measures for individual images and combined image sequences, which serve as the primary metric of performance for the final science products. The database is also used to track telemetry products from the adaptive optics subsystem, and associate these with corresponding IFS data. Here, we discuss several data exploration and visualization projects enabled by the GPIES database. Of particular interest are any correlations between instrument performance and environmental or operating conditions. We show single and multiple-parameter fits of single-image and observing sequence contrast as functions of various seeing measures, and discuss automated outlier rejection and other fitting concerns. Supervised learning techniques are employed in order to partition the space of raw (single image) to final (full sequence) contrast in order to better predict the value of the final data set from the first few completed observations. Finally, we discuss the particular features of the database design that aid in performing these analyses, and suggest potential future upgrades and refinements.