The Planet Pipeline: enabling data mining and citizen science with Hubble images of the Solar System

In 15 years of service, the Wide Field Planetary Camera 2 (WFPC2) onboard the Hubble Space Telescope (HST) obtained over 10,000 frames of Solar System data. Since standard data reduction pipelines are typically not optimized for movingtarget data, our “planet pipeline” will uniformly reprocess and catalog this WFPC2 image collection to make it more immediately science-ready. Some of our processing steps will utilize citizen scientists to perform visual inspections. Our corresponding database will enable robust queries which are more specific to planetary science, helping archival researchers quickly find and utilize the prepared images within our collection for a wide range of scientific analyses. We welcome suggestions (especially from veteran WFPC2 users) on the optimal treatment and organization of this data collection, and also to identify a broad range of analyses that might only be possible with visual inspections by citizen scientists. Our processed images and associated catalogs will be made available as High Level Science Products (HLSP) in the Multimission Archive at STScI (MAST): http://archive.stsci.edu/prepds/planetpipeline 1. Data processing and meta-data Our processing includes additional steps, beyond the basic calibrations performed by MAST, which are idiosyncratic enough that that they can be a barrier to faster and deeper analyses. Unlike fixed-target data, where rejection of artifacts is accomplished by combining multiple images, cleaning up moving target images requires an ability to distinguish real objects from artifacts in single images. For the initial rejection of bad pixels, cosmic rays, and star trails, we will use single-image rejection techniques based on Laplacian edge detection [3]. Figure 1: Drizzled WFPC2 mosaic of a Saturn observation (top), and corresponding satellite finder chart (bottom, from PDS Rings Node). Then each image will be visually inspected to identify missed rejections of artifacts, and also unintentional rejections of real objects or features. The inspections will also record secondary and serendipitous features in each image, to form a comprehensive catalog. Finder charts will be used to verify which satellites were detected in each frame, and possibly reveal the presence of known or unknown asteroids or Kuiper belt objects (Figure 1). Planetary surface features such as storms, vortices, satellite umbrae, volcanoes, and craters can also be recorded. Each of the objects and features recorded will be assessed for data quality. Secondary objects have a higher probability of being somehow nonoptimal: poorly placed, barely detected or overexposed (saturated). So it is important to indicate which of these “extra” observations may be useless. Using our refined rejection masks, we will utilize all four WFPC2 chips to produce clean and undistorted EPSC Abstracts Vol. 6, EPSC-DPS2011-1290, 2011 EPSC-DPS Joint Meeting 2011 c © Author(s) 2011 mosaic images. We will optimally resample both Planetary Camera (PC) and Wide Field (WF) data by drizzling them to a common pixel scale, and apply image deconvolution methods wherever possible. We will insert planetary meta-data into our output image headers, since interpretation of solar system observations relies on time-variable properties such as observing geometry, solar orientation, etc. For each image, we obtain ephemeris data from the JPL Horizons website (http://ssd.jpl.nasa.gov/?horizons), associating the keywords in Table 1 with the image data. These keywords are then used to calculate the conversion between counts/sec in the image data into reflectivity, or I/F, following Sromovsky & Fry [2]. Table 1: Ephemeris meta-data stored for each image, gathered from the JPL Horizons website. Horizons keyword Description 1 Astrometric R.A. and DEC (J2000) 6 Satellite offset and PA from primary 10 Illuminated fraction 13 Target angular diameter 14 HST sub-longitude/latitude 15 Sun sub-longitude/latitude 17 North pole position angle & distance 19 Heliocentric range (and rate of change) 20 HST range (and rate of change) 21 One-way light time 24 Phase angle 26 HST-primary-target angle 39 Uncertainty in range and range rate