The Io GIS Database 1.0: A Proto-Io Planetary Spatial Data Infrastructure

We collected a set of published, higher-order data products of Jupiter's volcanic moon Io and assembled them in an ArcGIS tm database we are calling the Io GIS Database, version 1.0. The purpose of this database is to collect image, topographic, geologic, and thermal emission data of Io in one geospatially registered location to form the data component of an Io planetary spatial data infrastructure (PSDI). The goals of an Io PSDI are (1) to make higher-order data products more accessible and usable to the broader planetary science community, particularly to new scientists that were not associated with the projects that obtained the data; (2) to enable new scientific studies with the data; and (3) to create a tool to support observation planning for future Io-focused planetary missions. In this paper we describe the motivation behind our project, discuss the data sets acquired for this first version of the database, and demonstrate how they can be used. We conclude with a discussion of how our database relates to other PSDIs, our plans for future updates, and a request for additional Io data sets.


Introduction
Over the last decade there has been great interest within the United States' National Aeronautics and Space Administrationʼs (NASAʼs) Planetary Science Division regarding the long-term accessibility and usability of planetary data, particularly geospatial image data of planetary surfaces, and particularly the higherorder data products (e.g., regional to global image mosaics, digital terrain models (DTMs), geologic maps, etc.) derived from NASAʼs robotic planetary missions (see https://www.lpi.usra. edu/mapsit/about/ and https://science.nasa.gov/science-pink/s3fspublic/atoms/files/PDE%20IRB%20Final%20Report.pdf (NASA Planetary Data Ecosystem Independent Review Board Final Report)). NASAʼs desire to maximize its investment in its planetary missions and their accumulated data is motivated by the need to enable future generations of planetary scientists to utilize the data for research projects, long after the creators of those data are gone, and where many spacecraft data sets are not described well enough to enable ease of use. Likewise, NASA is looking to ensure that data from past missions is usable in tools that will support the planning of future missions. This is particularly desirable for geologically active worlds, such as Jupiterʼs volcanic moon Io, where multiple, ongoing volcanic eruptions produce thermal anomalies related to its interior processes, and where active eruptions emplace effusive and explosive volcanic materials and gases that regularly modify its surface at weeks to months timescales (see, e.g., Lopes & Williams 2005;Lopes & Spencer 2007).
In 2019 we were funded through a 1 yr grant from NASAʼs Planetary Data Archiving, Restoration, and Tools (PDART) program to collect a discrete selection of mostly image-based, higher-order data products of Io produced from NASAʼs Voyager, Galileo, and New Horizons missions, as well as thermal emission observations from the last two decades from spacecraft and Earth-based telescopes utilizing adaptive optics (AO), and assemble them in a geospatially referenced format. This database would become the initial data component of an Io Planetary Spatial Data Infrastructure (PSDI), which would serve as a tool for future research and future mission planning. In this paper we discuss the results of our work, which we call the Io Geographic Information Systems (GIS) Database version 1.0, and include further details on the driving motivation of this project, the formats used, examples of the data sets that were included, the availability of our database after NASA Planetary Data System (PDS) review, and how it would enable future work.

Background: Data Accessibility, Usability, and PSDIs
Over the last decade NASA and members of the US planetary science community have expressed concern about the long-term usability and accessibility of NASA-acquired planetary data, particularly those higher-order data products derived from past planetary missions that are usually not archived in NASAʼs PDS. The Mapping and Planetary Spatial Infrastructure Team (MAPSIT), the NASA assessment group tasked with cartography and spatial data issues across the Planetary Science Division (see MAPSIT Roadmap: https:// www.lpi.usra.edu/mapsit/roadmap/), has focused on developing the concept of PSDIs (Laura et al. 2017) for all objects in our solar system with solid surfaces, which have been visited (or will soon be visited) by planetary spacecraft. A PSDI is derived from a similar concept in terrestrial geoscience called a spatial data infrastructure (SDI; Rajabifard et al. 2002). A PSDI is an enabling collection of planetary spatial data for a specific planetary object (planet, moon, asteroid, or comet), which also includes data access mechanisms, data interoperability agreements, data policies, and standards, as well as the spatial data users themselves (Figure 1). The goal of a PSDI is to collect data in a form that "just works" (i.e., is usable for science analysis without extensive calibration and processing from the original raw data), so that planetary scientists do not have to go back to the PDS and regenerate desired higher-order data products from scratch, a task that may be impossible for many planetary researchers decades after the data were acquired. PSDIs would thus more easily enable future research on the planetary object of interest and/or would support observation planning of that object for any future missions (Laura et al. 2017).
PSDIs are especially important for planetary objects that have been visited by past missions that are probable targets of future missions. For example, the Moon and Mars are probable targets for human missions in upcoming decades, whereas Venus, Mercury, Jupiterʼs icy moons Europa and Ganymede, and Saturnʼs moon Titan are definite targets of new robotic missions within this decade or the next. Europa especially has been considered an ideal target to develop one of the first PSDIs (Laura et al. 2018), given (1) the limited existing image data, and (2) that the NASA Europa Clipper Flagship mission is in development for launch later this decade.
Jupiterʼs innermost large moon Io is the most volcanically active object in the solar system (see, e.g., Lopes & Williams 2005, and references therein). Io was visited by NASAʼs Voyager (Smith et al. 1979a(Smith et al. , 1979b, Galileo (McEwen et al. 1998), Cassini (Porco et al. 2003), and New Horizons ) spacecraft; has been observed by the NASA Hubble Space Telescope; and is the target of ongoing observations by several Earth-based telescopes applying AO techniques to compensate for the distortion caused by the Earthʼs atmosphere and to improve the spatial resolution of thermal emission from active volcanoes. Because of Ioʼs ongoing geologic activity, with eruptions producing thermal anomalies indicative of different eruption styles (Carr 1986;Davies 1996;Davies et al. 2001Davies et al. , 2010Davies et al. , 2018, it has been the target of multiple mission proposals over the last decade (e.g., McEwen et al. 2014McEwen et al. , 2019Suer et al. 2017) to better understand its tidally induced volcanism. Thus, given Ioʼs wide interest and limited available data (relative to the Moon or Mars), Io is also a prime candidate to develop one of the first PSDIs.

Data and Methods
Our concept was to collect a subset of the accessible and usable, higher-order image-based data products of Io that have been peerreviewed and published over the last two decades, and assemble them in a geospatially controlled and registered format to enable future work (Figure 2). The primary software we chose to use is ArcGIS TM (ESRI 2020). We chose ArcGIS TM for two reasons. First, we have prior experience with that software, which was used by D.A.W. to produce the first complete global geological map of     (Becker & Geissler 2005), in which the mosaicked images were geodetically controlled using a triaxial ellipsoid shape model and the best available Galileo control point network (Archinal et al. 2001). Reported horizontal accuracy is nominally 1 pixel, translating to 1 km in low-latitude regions with good coverage. These mosaics were recommended to be used as foundational data products for Io (Laura & Beyer 2021) and thus are the best available prepared and controlled data set on which to build an Io PSDI. Table 1 lists the published Io data sets we chose to include in this first version of the database, which were listed in our selected PDART proposal. In addition to the combined Galileo-Voyager mosaics, we wanted to include all other regional-scale Galileo Solid State Imager (SSI) mosaics (∼150-900 m pixel −1 spatial resolution; Keszthelyi et al. 2001;Turtle et al. 2004) as well as other SSI-derived products, including the digital elevation model (DEM) of Io made using stereo photoclinometry (White et al. 2014), global geologic maps from Voyager (Crown et al. 1992) and Galileo ) data, derived geologic maps from the regional SSI mosaics (Williams et al. 2002(Williams et al. , 2004(Williams et al. , 2007Bunte et al. 2008Bunte et al. , 2010Leone et al. 2009), as well as global mosaics from the New Horizons Long Range Reconnaissance Imager (LORRI) and the Linear Etalon Infrared Spectral Array (LEISA) instruments, obtained during the 2007 February flyby Tsang et al. 2014). Some thermal emission data of Io exists both in image format, and as tables of hot spot area, power, and temperature information. We included thermal emission data as derived from the Galileo Near Infrared Mapping Spectrometer (NIMS) Thermal Emission Database (NITED: Veeder et al. 2009Veeder et al. , 2011Veeder et al. , 2012Veeder et al. , 2015Davies et al. 2012aDavies et al. , 2012bDavies et al. , 2015, New Horizons' LEISA instrument (Tsang et al. 2014), and several Earth-based telescopes such as Keck and Gemini using AO, as reported in multiple papers (Marchis et al. 2005;de Pater et al. 2014;Cantrall et al. 2018;de Kleer et al. 2019). Finally, a study by Hamilton et al. (2013) used the location and clustering of subsets of hot spots that were then compared with underlying geology and were subsequently compared to models of heat flux assuming heat derived from different Ionian interior layers (i.e., asthenosphere, deep mantle, various combinations thereof (Ross et al. 1990). A reviewer of this paper noted that the Hamilton et al. (2013) cluster analyses were not weighted by individual hot spot thermal emission. Instead, there was a comparison of hot spot number and geology with predicted models of global heat flow, not measurements of heat flux. (1) Compare to Figure 3. Note that there are additional thermal emission data for Io in the community, which we hope to add to future versions of this database. Please contact David Williams (David.Williams@asu.edu) to contribute new data sets to future versions of this database.
(2) Galileo SSI I24 observations were damaged by radiation exposure to the camera electronics and were only partially recoverable. There were insufficient resources to include them in this project.
(3) Co-I Milazzo thinks Io mosaics better than 200 m px −1 would require too much time to tie to this database, so they are not included in this first version of the database.
3 NASA now requires all planetary geologic maps to be produced using GIS software, and ArcGIS TM is favored by the US Geological Survey, who oversees NASAʼs Planetary Geologic Mapping Program.
Thus, we included those interior heat flux model maps in this database to enable future comparisons of additional thermal activity to heat flux source regions. Note: There are many more data sets (e.g., Galileo Photo Polarimeter-Radiometer: Rathbun et al. 2004), particularly thermal emission data sets, of Io that were not included. This is not because we did not like them but because we chose the data sets we were most familiar with so as to keep this 1 yr PDART project manageable. If there are Io data sets that the reader thinks would enhance the value of this database, then please contact David Williams at David.Williams@asu.edu and they will be assessed for inclusion in the next version of the database.

Results: Database Version 1.0
We assembled version 1.0 of the Io GIS Database using ArcGIS TM version 10, which is readable by ArcGIS TM Pro TM . In addition to the data sets described (Table 1, Figure 3), we included the latest file of named surface features from the USGS Planetary Nomenclature website, as well as a graticule displaying a 30°latitude-longitude grid (Figure 4). Data are presented using a simple cylindrical projection centered on the anti-Jovian point (0°, 180°W), as the Galileo mission obtained its best imaging data over the anti-Jovian hemisphere. 4 To enhance usage and availability of our database materials, we have reproduced them in Arizona State Universityʼs Java Mission-planning and Analysis for Remote Sensing platform (JMARS: Christensen et al. 2009), which is another GIS-based application with widespread usage in the planetary community for planetary mission data analysis.
Figures 5-6 demonstrate the functionality of the Io GIS Database. In Figures 5(a)-(c), we show the various regional and global geologic maps derived from the Voyager and Galileo images. Having geologic maps from the 1990s, 2000s, and 2010s in this database enables comparisons and can show the evolution in interpretation of Ioʼs geologic features, particularly between the Voyager and Galileo eras. Figures 6(a) and (b) show two examples of thermal imaging of Io, first by passing spacecraft (New Horizons, 2007) and second by Earth-based telescopes using AO (2001AO ( -2018. Importantly, the anomalous thermal emission hot spot data sets include attribute tables, which contain details on the recorded thermal activity at every location on Io, covering a time period between 1996 and 2018. By checking the power, area, and temperature variations at hot spots of interest, it is possible to investigate the waxing and waning of volcanic activity over this 20 yr time period. We note that the data from Veeder et al.  (Figure 6(c)), it is possible to correlate the spatial location of volcanic activity with models of specific distribution of heat in Ioʼs interior. An early attempt at this was done using spacecraft-only thermal data (Hamilton et al. 2013), but they did not include the telescopic observations. With these data now included in the Io Database, more robust studies could be done.
The primary advantage to collecting these Io data sets in one geospatial database, for both the ArcGIS TM and JMARS versions, is that it allows users to pick and choose which data sets to visualize, enabling comparison of geologic, thermal emission, topographic, and image data for any location on Io. Because of Ioʼs active volcanism, and repeated observations by both spacecraft and telescopes, our database contains a history of Ioʼs activity, enabling studies in both space and time. Our database thus provides a powerful tool both to enable further scientific research of Io and to identify imaging targets of repeated   Crown et al. (1992) rendered over the combined Galileo-Voyager monochrome global mosaic (Becker & Geissler 2005). Map unit names are included in the ArcMap Table of Contents for this layer. (b) Output from the Io GIS Database v. 1.0, showing the eight regional geologic maps made from regional Galileo SSI mosaics over Ioʼs anti-Jovian hemisphere (Williams et al. 2002(Williams et al. , 2004(Williams et al. , 2007Bunte et al. 2008Bunte et al. , 2010Leone et al. 2009). Rendered over the combined Galileo-Voyager monochrome global mosaic (Becker & Geissler 2005).  activity for future Io-dedicated missions. JMARS in particular was developed to aid observation planning by the NASA Mars Odyssey orbiter Thermal Emission Imaging System (THEMIS) camera (Christensen et al. 2009), and thus we hope that this database (in either ArcGIS or JMARS) could be equally helpful for planning observations for future Io missions.

Policies and Standards
Every PSDI must address data policy issues and standards, as well as the community of data users, access, and the data themselves. Our approach for the Io GIS Database as a proto-Io PSDI is outlined in Figure 7. In terms of standards, we chose ArcGIS TM and JMARS because of the wide usage of these software in the planetary science community, and because of their demonstrated performance in handling planetary geospatial data. We chose to use the combined Galileo-Voyager global mosaics of Becker & Geissler (2005) as our foundational data products, as recommended by Laura & Beyer (2021), because they represent the best global image products produced for Io since the end of the Galileo mission and were produced using the best geodetic control (Archinal et al. 2001). All other data are registered to these products. Finally, all data included in version 1.0 of the Io GIS Database have already been previously peer-reviewed and published, and thus have passed the scrutiny of the planetary science community.
In terms of policy, our goal is to update the database about every two years, as new Io data sets are published and made available to us, up to the start of the next Io-dedicated planetary mission. A NASA Discovery mission proposal for an Iofocused mission, the Io Volcano Observer (IVO; McEwen et al. 2019), was under consideration by NASA along with three other missions in 2021 but was not selected for this Discovery round. We think eventually IVO or another Io-dedicated mission will be selected and return new high spatial resolution data of Io, after which the onus would fall on members of the Io community and/or the mission Science Team to continue to update the database with new data products or create a new Io PSDI.

Peer Review and Access
Io GIS Database version 1.0 will undergo peer-review for formatting and operability considerations by the NASA PDS Imaging and Cartography Node during fall 2021. Upon approval, it will be considered open access, and it will be made freely available for download both at the PDS and at our university website, and announcements will be made to the planetary science community through various listservers and newsletters. A pre-PDS review copy of the whole Zipped Arc project can be downloaded from https://rpif.asu.edu/downloads/ PDART_Io_DB_GIS_data_1.0_v2.zip. Accessing the data in  Tsang et al. (2014). These hot spots (green dots) correlate with temperature and power data in a corresponding attribute table in ArcMap. By switching basemaps, users can compare hot spot measurements to imaged features and interpreted geology, and they can compare with data from other instruments to track thermal evolution at any given location on Io over the last 20 yr. JMARS via a web browser will require a free JMARS account. To sign up for an account, go to https://jmars.asu.edu/.

Conclusions and Future Directions
We constructed a GIS database containing a variety of higher-order data products of Jupiterʼs volcanic moon Io, derived from NASAʼs Voyager, Galileo, and New Horizons missions, and from Earth-based telescopic observations using AO. Contained in both ArcGIS TM and JMARS software, Io GIS Database version 1.0 is designed to serve as the initial data component of an Io PSDI. The goal of this database is to collect as many peer-reviewed and published, higher-order data products of Io as possible in a geospatial format to enable easy comparison of the data, as a tool to enhance future Io research and to enable observation planning for future Io missions. We previewed the content of the database, and we discussed the policies, standards, and accessibility of the database, which we hope will be useful to the community of Io scientists until new high-resolution spacecraft data are obtained. Until then, we are hoping to revise the database every few years, as we acquire new published Io data sets. We especially want to add new thermal emission data sets that track the variability of Ioʼs volcanic activity, including from AO telescopes and the Juno mission. We hope this database serves as a benchmark for the planetary science community to develop PSDIs for other objects in our solar system.