The Auroral Planetary Imaging and Spectroscopy (APIS) service

9.5 MB·English

by  Laurent Lamy

The Auroral Planetary Imaging and Spectroscopy (APIS) service
L. Lamya,∗, R. Prang´ea, F. Henrya, P. Le Sidanerb
aLESIA, Observatoire de Paris, CNRS, UPMC, Univ. Paris Diderot, 92190 Meudon, France.
bDIO-VO, CNRS, Observatoire de Paris, 75014, Paris, France.
Abstract
The Auroral Planetary Imaging and Spectroscopy (APIS) service, accessible online, provides an open and interactive
access to processed auroral observations of the outer planets and their satellites. Such observations are of interest for
5
a wide community at the interface between planetology and magnetospheric and heliospheric physics. APIS consists
1
of (i) a high level database, built from planetary auroral observations acquired by the Hubble Space Telescope (HST)
0
2since 1997 with its mostly used Far-UltraViolet spectro-imagers, (ii) a dedicated search interface aimed at browsing
efficiently this database through relevant conditional search criteria and (iii) the ability to interactively work with the
n
dataonlinethroughplottingtoolsdevelopedbytheVirtualObservatory(VO)community,suchasAladinandSpecview.
a
JThis service is VO compliant and can therefore also been queried by external search tools of the VO community. The

6diversity of available data and the capability to sort them out by relevant physical criteria shall in particular facilitate
1statistical studies, on long-term scales and/or multi-instrumental multi-spectral combined analysis.

]Keywords: Aurorae, Magnetospheres, Planets, Database, Virtual Observatory, Observation service
M
I
h.1. Introduction Amongspace-basedUVobservatories,theHubbleSpace
p Telescope(HST)intensivelyobservedtheauroraeofouter
Ultraviolet(UV)planetaryastronomyprovidesawealth
- planetary systems (Jupiter, Saturn, Uranus) in the Far-
oofinformationonplanetaryenvironmentsofthesolarsys-
r UV (FUV, 116 to 200 nm) from 1993 up to now, provid-
tem and beyond [1, and refs therein] and benefits a good
t ing thousands of images and spectra, often in the frame
s
angularresolution. Planets,moonsandringsprimarilyre-
a of combined observations with spatial probes dedicated
[flect the (time variable) solar continuum [2] with various
to planetary exploration such as Galileo (in orbit around
albedos, depending on their composition and dynamics.
1 Jupiter over 1995-2003), Cassini (flyby of Jupiter in 2000,
TheUVdomainisalsoadaptedtomeasureintrinsicatmo-
v inorbitaroundSaturnsince2004)andNewHorizons(flyby
0spheric emissions such as airglow and aurorae, which are
of Jupiter in 2007) or with Earth-based observatories ob-
2produced by electronic transitions of neutral species pre-
servingintheradio,IRandX-raysdomains. Figure1dis-
9vailingintheupperatmosphereofgiantplanets(asHand
3 plays examples of HST-FUV observations of giant plane-
H )andtheirsatellites(asO)thatarecollisionallyexcited
0 2 tarysystems. Theseobservationsnowformarichdatabase,
by precipitating charged particles. Airglow is a weak ra-
. of interest for a wide community interested in planetary
1diation emitted over the whole atmospheric disc and pow-
0 magnetospheres, planetology and heliospheric physics in
ered by solar fluorescence and photoelectron excitation [3,
5 theoutersolarsystem. However,theirusebynon-specialists
and refs therein]. In contrast, aurorae are bright localized
1 remains limited by their complexity, the lack of high level
:emissions radiated from the auroral regions of magnetized
v data and associated Figures, and the difficulty to access
planets(orofconductivemoonsinteractingwiththem)by
i them.
Xenergeticchargedparticlesacceleratedfartherinthemag-
TheAuroralPlanetaryImagingandSpectroscopy(APIS)
rnetosphere [4, and refs therein]. Auroral observations, on
a service,accessibleathttp://apis.obspm.fr,aimsatpro-
which we focus hereafter, thus provide direct contraints
viding an open and easy access to a high-level auroral
on the electrodynamic interaction between planetary at-
database, built from public HST observations processed
mospheres, magnetospheres, moons and the solar wind as
in convenient formats, and compatible with virtual obser-
well as on the underlying plasma processes at work (par-
vatory(VO)facilities. Itthuscontributestocurrentefforts
ticle acceleration, energy and momentum transfert). Au-
oftheVOplanetologycommunitytodevelopstandardized
roral spectro-imaging in the UV provide key observables,
databases and tools to access them. It has been open to
such as the spatial topology and dynamics of the active
the community in July 2013 [7]. Fortuitously, the bull
magnetic field lines, the radiated and precipitated power,
APIS is also the ancient egyptian god of (data) fertiliza-
and even the energy of precipitating electrons [5, 6].
tion, wearing an (active) solar disc between its horns.
∗Correspondingauthor
Preprint submitted to Astronomy and Computing January 19, 2015
JUPITER SATURN URANUS
Reflected sunlight
Dayglow
Io Titan
Aurora
Figure 1: Images of outer planets and satellites obtained by the ESA/NASA Hubble Space Telescope in the Far-UV, available through the
APISdatabase.
2. A high level database means that they have undergone a standard processing
(correctionfordarkcurrent,flat-fielding,correctionforge-
APISprimarilyreliesonahighleveldatabase,archived ometric distortion) and are expressed in basic units, such
at Virtual Observatory - Paris Data Centre (VO-PDC) ascounts.pix−1 forSTISandelectrons.s−1.pix−1 forACS,
and based on HST FUV auroral observations of the outer withtheXandYaxisdisplayedinthetelescopegeometric
planets and their moons acquired since 1997. For con- frame.
venience, we restricted to the mostly used instruments, For our purpose, we have systematically reprocessed
namelytheSpaceTelescopeImagingSpectrograph(STIS) the STIS calibrated images (x2d extension) from raw data
and the Advanced Camera for Surveys (ACS), using the throughthestandardSTScIpipelinewiththemostrecent
using the FUV and the Solar Blind Chanel (SBC) Multi- calibration files, which are referenced in the header of up-
Anode Microchannel Array (MAMA) detectors, respec- dated fits files. Alternately, we used the calibrated ACS
tively. We refer the interested reader to the STIS and fits files (drz extension) directly as retrieved from MAST,
ACS handbooks provided by the Space Telescope Science as these are calibrated up-to-date by default. Before fur-
Institute (STScI) for further details. To date, the con- therprocessing,the(dark)lineofbadpixelsvisibleinACS
sidered observations consist of about 6000 individual im- images was replaced by values linearly interpolated from
ages and spectra, issued from 37 observational campaigns neighbor pixels.
of the Jupiter, Saturn and Uranus systems distributed as Plots of level 1 images display the telescope field-of-
illustrated in Figure 2, and obtained with very diverse in- view (FOV) in the telescope frame with a linear intensity
strumentalconfigurations(filtersforimaging,combination scale.
ofslitandgratingsforspectroscopy). Foreachofthesein-
dividual observations, APIS provides a set of value-added Level 2. Processed images - level 2 - are obtained as fol-
data levels briefly described below, each available in work lows. Level 1 images are first rotated to be displayed in
and graphical formats such as fits files [8] and jpeg, pdf a target-related physical frame. The Y axis points along
plots. the rotation axis for all bodies, except for Uranus where
it points along the celestial north, and the X axis points
2.1. Imaging Eastward.
Images are provided under three levels of data : orig- The images are then individually corrected from the
inal images as calibrated by STScI (level 1), spatially re- HST pointing inaccuracy, which can reach a few tens of
orientedandfittedimagesprovidedwithpixelplanetocen- pixels in translation and a few fractions of a degree in
triccoordinates(level2)andbackground-subtractedcylin- rotation. In practice, this essential step is achieved by fit-
drical/polarprojectionstransposedintophysicalunits(level ting the planetary limb at the 1-bar level, as well as the
3). At the bottom of Figure 3, the row labelled ”Images” rings when visible, by a reference ellipsoid. This is done
(in blue) illustrates these data levels with quick looks, be- automatically through 2D cross-correlation for small size
low which are provided various data formats. targets (Uranus and moons) and ACS images of Saturn,
withanaccuracyestimatedto1pixelinbothXandYdi-
Level 1. Original images - or level 1 images - originate rections. As STIS observations of Saturn display a noisier
from public data delivered by STScI through the Mikul- limb which can hardly be automatically fitted, the corre-
ski Archive for Space Telescopes (MAST), and mirrored sponding images (about 70 in total) were all individually
at the ESA Hubble Science Archive. They correspond to re-centered manually, with an accuracy estimated to 1-2
calibrated images in the sense intended by STScI, which pixels. The more complex fitting of Jupiter images, for
2
APIS : an interactive database of Far-UV planetary
auroral observations
L. Lamy, R. Prangé, F. Henry and P. Le Sidaner
http://lesia.obspm.fr/apis/
LESIA, VO-Paris Data Centre, Observatoire de Paris, CNRS, UPMC, Univ. Paris Diderot, Meudon, France
T SEARCH INTERFACE
C
A
R
T
S
B
A
THE APIS DATABASE! FIGURE 1
The STIS and ACS Far-UV instruments
of the Hubble Space Telescope1 (HST)
Reflected sunlight
acquired ~5000 individual spectra and
Dayglow Search
images of the aurorae of Jupiter, Saturn
criteria
and Uranus over 1997-2013 (Figure 2).
Io Titan
But their use remains generally limited, Aurora
owing to the difficulty to access and use
raw and value-added data. JUPITER SATURN URANUS
Figure 1 : Remote UV measurements of the outer planets.
A
APIS, the egyptian god of fertilization, is also the acronym of the new database Auroral Planetary Imaging
s
and Spectroscopy, aimed at facilitating the use of HST planetary auroral observations2. APIS is based at e
g
the Virtual Observatory (VO) of Paris and provides a free and interactive access to a variety of high level a
m
data through a simple research interface (Figure 3) and standard VO tools (Figure 4), presented hereafter.
I
Data
FIGURE 2
a
6 r
JUPITER ct
4 e
p
S
°) 2
(
e
d 0
atitu -2 Basic Raw data High level data
L
information
-4 Various
formats
-6
40
SATURN Interactive investigation
of fits data with VO tools
20
)
°
(
e
ud 0
atit
L
-20 Aladin (CDS) Specview (STSci)
-40
URANUS
40
) 20 sub-solar latitude
°
(
e
d 0
u
Latit-20 sub-earth latitude
-40
Figure 4 : Jupiter image and spectrum read with VO softwares.
REFERENCES!
1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
Years !
1 http://www.stsci.edu/hst/HST_overview/instruments
Figure 2 : HST-FUV observations over 1997-2013. 2 L. Lamy, R. Prangé, F. Henry and P. Le Sidaner, Magnetospheres of Outer Planets meeting, Athens, 11 July 2013!
Figure 2: Sub-Earth (solid line) and Sub-solar (dashed line) latitude of Jupiter, Saturn and Uranus as a function of time, together with 3 S. Erard, B. Cecconi, P. Le Sidaner et al., see http://voparis-europlanet.obspm.fr/docum.shtml
combinedHSTimagesindicatingthemainSTIS/ACSobservingcampaignssince1997.
which only part of the planetary disc is visible, is under based on SPICE kernels), the Saturn’s Kilometric Radi-
progress. ation (SKR) southern and northern phases [9] (source :
The level 2 images are then extracted from the re- Cassini/RPWS/HFR Kronos server, http://www.lesia.
centered images with a square box centered on the target obspm.fr/kronos/skr_periodicity.php), and resolved
and whose size varies with that of the observed body. secondary targets coincidently visible in the HST FOV
The fits files provided for level 2 images contain 7 sci- (such as Callisto for Jupiter, Titan for Saturn or Saturn
ence extensions together with an updated header. Exten- for Titan).
sion 0 corresponds to the re-centered image in unchanged Plots of level 2 images display the telescope FOV in
units. Extensions 1 to 6 provide the associated pixels co- the target frame with a logarithmic intensity scale, aimed
ordinates in a planetocentric frame, namely the latitude at reducing the contrast between bright and faint emis-
(extension 1), the local time (extension 2), the zenithal sions,andawhitegridofplanetocentriclatitudeandlocal
solar angle (extension 3) and the zenithal observing angle time at the limb altitude. The red meridian marks noon.
(extension 4) at the limb altitude, together with the lati- Basic information on the observation and the associated
tude (extension 5) and the local time (extension 6) at the ephemeris are reminded on the right.
auroralpeakaltitude,henceatthefootprintofactivemag- Level 2 data are intended to be useful to the broadest
netic field lines. Figure 4 shows plots of extensions 0 to 3 rangeofusers,asthelatterhavethepossibilitytotakead-
foralevel2imageofSaturn. Theheadercontainsbasicin- vantage of pixel coordinates (latitude, longitude, zenithal
formationontheoriginalobservationtogetherwithvalue- angles) to derived and remove any background model of
addedinformationastheplanetaryephemeris(source: In- their choice and/or to apply any preferred conversion fac-
stitut de M´ecanique C´eleste et de Calcul des Eph´em´erides tor to transpose basic units to physical ones.
orIMCCE,http://www.imcce.fr),theephemerisofmoons
andGalileo,CassinispacecraftforJupiterandSaturn(source Level 3. Cylindrical and polar projections at auroral al-
: UniversityofIowa,http://www-pw.physics.uiowa.edu, titudes expressed in physical units - level 3 - are finally
3
k
c
i
u
Q
Search
criteria
d
e
c
n
a
v
d
A
s
e
g
a
m
I
Data
a
r
t
c
e
p
S
Initial data Higher level data
Summary
Data Interactive use of fits
informations
formats data with VO tools
Figure3: Searchinterface.
derived for the images of planets through the following grounds have been employed and tested in the literature.
additional processing steps. The most accurate solution is generally to derive an em-
A model of reflected background at the limb altitude piricalbackgroundfromoneorseveralimageswithoutau-
isfirstsubtractedtolevel2images. Severaltypesofback- roral signal acquired close to the investigated observation.
4
calibrations of STIS and ACS with typical auroral spec-
tra of planets to convert instrumental units into auroral
(a) (b) brightnesses [13, 14, 15, 16, 17, 11]. The resulting con-
version factors yielded significant disparities, owing to the
considered radiative species (H and/or H-Lyα), the con-
2
sideration of absorption by hydrocarbons and the selected
bandpass (spectral domain of H emission, FUV domain
2
or bandpass restricted to that of HST imaging filters), as
discussed by [18, 19]. Here, we used the most recent esti-
mates of [18] derived for the two broadband filters of each
instrument, which are the clear and SRF2 filters for STIS
and the F115LP and F125LP ones for ACS, that we ex-
(c) (d)
trapolated to the other employed filters. The conversion
factors of filters outside of the spectral domain of auroral
H , as F165LP for ACS, were set to 0.
2
The obtained images are finally projected at the auro-
ralpeakaltitude,indicatedintheheaderoflevel2fitsfiles,
with a dedicated projection routine preserving the photon
fluxofagivenemittingsurfacebothincylindricalandpo-
larframes. Cylindricalandpolarprojectionsaresystemat-
Figure 4: Example of Aladin’s use with a level 2 image of Saturn. icallyderivedforbothnorthernandsouthernhemispheres,
Panels (a), (b), (c) and (d) respectively plots the extensions 0 (sci-
butthesearchinterface(describedinthenextsection)only
ence image), 1 (latitude at the limb altitude), 2 (local time at the
displays the projection of the main (summer) hemisphere.
limbaltitude)and3(solarzenithalangleatthelimbaltitude)ofthe
level 2 fits file. The plotted extension is highlighted in blue on the The projection of the secondary (winter) hemisphere, to-
rightbar. gether with all data levels are accessible when clicking on
the ”Detailed information” link, visible on the left of each
observation in Figure 3. Such projections were derived for
As this is not always possible to find relevant references
planets only, and not for moons, owing to the different ra-
images for a given campaign and because the amount of
diative species and the lack of reliable conversion factors.
observations requires an automated processing pipeline,
Plots of level 3 images display cylindrical and polar
we rather chose to subtract a Minnaert-type numerical
projections with a logarithmic intensity scale and a white
background model [10, and refs therein] fitted to the non-
grid of planetocentric latitude and local time at the peak
auroral regions of the observed planetary disc and then
auroral altitude. The red meridian marks noon.
convolvedbytheSTISorACSpointspreadfunction. The
To help the APIS user to track auroral dynamics, sev-
Minnaert model consists of a polynomial law of the nepe-
eral visual references were additionally added to Saturn’s
rian logarithm of the product of the cosines of the obser-
projections. Cylindricalprojectionsdisplaywhiteboxesin
vation and solar zenith angles. We tested different sets of
both hemispheres at the southern and northern expected
polynomial orders and latitudinal bins separately for each
footprints of Enceladus. The latter was indeed detected
planet. The best results were obtained with first order
by Cassini/UVIS [20], but never identified in HST images
polynoms and latitudinal bins of 2◦ and 30◦ for Saturn
so far [21]. Kronian cylindrical and polar projections also
and Uranus respectively. In the case of Jupiter, we note
display dashed-dotted (dashed) lines to mark the south-
that a first order Minnaert polynom with latitudinal bins
ern (northern) reference rotating meridian derived from
of 1◦ has been used [11]. The uncertainty of the model
the southern (northern) SKR phase system [19].
background in the auroral regions was estimated for Sat-
urn to be less than 10%, and less for Uranus. For the
2.2. Spectroscopy
sake of completeness, we also tested an alternate numeri-
cal background model, developed for FUV auroral images Spectra included in the APIS database so far restrict
inthecaseoftheEarthandbasedonthesumofthecosines to first-order long-slit, or slitless, spectroscopic measure-
of the observation and solar zenith angles [12]. This did ments acquired with STIS. They are also provided under
notprovidebackgroundmodelsofhigheraccuracythough. three levels of data : original 2D spectra as calibrated
An example of the applied processing is illustrated with a by STScI (level 1), extracted 2D spectra which are re-
STIS image of Saturn on Figure 5. calibrated in wavelength whenever needed and possible
Once the background subtracted, the resulting images (level2)and1Daveragespectrabuiltfromdifferentoccur-
are converted into auroral brightnesses (in units of kilo- rence levels (level 3). At the bottom of Figure 3, the row
rayleighs or kR, with 1 R = 106 photons.s−1.cm−2 in 4π labelled ”Spectra” (in blue) illustrates these data levels
sr) of total H emission over the 80-170nm spectral range. with quick looks and the available data formats.
2
Many past studies have performed successive photometric
5
Figure5: Exampleofsubtractionofanumericalbackgrounddiscmodel,herecompletedbyanempiricalringbackgroundmodel,toaSTIS
imageofSaturn.
Level 1. As for images, original 2D spectra - level 1 - been expressed in R.angstrom−1 to facilitate comparisons
are retrieved from the MAST archive and individually with the literature.
re-calibrated (x2d extension) along the STScI processing The fits files provided for the level 3 spectra contain
pipeline with the most recent calibration files. They are three science extensions, corresponding to the three 1D
expressedinphysicalunits,namelyerg.s−1.cm−2.angstrom−1.aarvcesreacg−e2spectradescribedabove,togetherwithanupdated
bydefault,anddisplayedinthetelescopegeometricframe header similar to that of level 2 spectra.
such that the X axis corresponds to wavelengths and the Plotsoflevel3spectradisplaythe1Dspectrasmoothed
Y axis corresponds to the spatial direction along the slit over 5 pixels with a linear intensity scale in two panels.
(or, equivalently, the Y axis of level 1 images). Theintensityrangeoftheupperpanelisvariable,inorder
Plotsoflevel1spectradisplay2Dspectrawithalinear toshowtheprofileofthemostintensesignal(generallythe
intensity scale, saturated at H-Lyα. As planetary targets H-Lyαline),whiletheintensityrangeofthelowerpanelif
are extended, the width of the H-Lyα line provides the fixed, to facilitate comparisons between weaker emissions
effectivespectralresolutionand,whenthetargetfullyfills as those of H .
2
intheslit,thewidthoftheH-Lyαlinecorrespondstothat
of the slit.
3. Search interface
Level 2. Processed 2D spectra - level 2 - are extracted
The above database can be queried by a dedicated
from original 2D spectra, and re-calibrated in wavelength
search interface available on the APIS web portal, as dis-
when not already performed by STScI. In practice, the
playedonFigure3. Adataqueryconsistsofaconditional
latter step was done by fitting the H-Lyα line whenever
search over a set of parameters (such as the date, the tar-
contained in the observed spectral range.
get, the instrument, the observation name etc.), known as
Thefitsfilesprovidedforlevel2spectracontainasingle
metadata which describe each single observation.
science extension, corresponding to the derived 2D spec-
trum, together with an updated header similar to that of
3.1. Quick search
level 2 images, which mentions the used spectral calibra-
The selection criteria located right to the blue label
tion (STScI, APIS or none).
”Quick” on Figure 3 are always visible. They consist of
Plots of level 2 spectra are displayed with a linear in-
basic criteria needed to perform a quick search. The first
tensity scale and axis coordinates. Basic information on
row lists fields dealing with instrumental specifications :
theobservationandtheassociatedephemerisaredisplayed
Target (name of the Planet or Satellite), Telescope (HST
on the right.
by default), Instrument (ACS or STIS), Observation type
Level 3. As 2D spectra are not necessarily a convenient (Imaging or Spectroscopy), Filter or Aperture (filters for
format to deal with, a set of typical 1D spectra - level 3 - imaging, any combination of slit and gratings for spec-
are provided for each long-slit spectral observation. They troscopy), the second row lists fields dealing with tempo-
consist of three average spectra respectively derived from ral selection : Data Interval, Observing Campaign (Data
the 50%, 10% and %1 most intense 1D spectra (which are groupedbyHSTobservingprogram)andDataset(unique
compared once integrated in wavelength) acquired along HST data identifier).
the slit. These average spectra thus provide an indication An example of data request (selection of the observing
ofthemean,highandpeakspectralactivityduringtheob- campaign Saturn - 2000 07-08 Dec) is illustrated at the
servation, regardless of the target size though. They have bottom of the Figure. The data are sorted out by time
6
vertically and the different data levels appear in succes- datalevels,APISisparticularlyadaptedtostatisticalstud-
sive horizontal columns. The first column, entitled ”Ob- ies, over long-term scales, and the interval between 1997
servation summary”, provides a list of basic parameters and 2014 now exceeds the duration of a solar cycle, a jo-
associated to each observation. More parameters are dis- vian revolution or half a kronian revolution around the
playedbyclickinginonthe”Detailedinformation”linkat sun.
the bottom of each list. For each data level, the available
data formats lie below a quick look. Clicking in on the
4. VO compatibility
latterdisplaysthefullresolutionplotandenablestheuser
to pass from one observation to the following or preceding 4.1. Interactive use
one for a chosen data level.
AlastfunctionalityoftheAPISserviceisthattheuser
can work with the data directly online, without needing
3.2. Advanced search
to download them. This is achieved by using the VO soft-
The selection criteria located right to the blue label wares Aladin (http://aladin.u-strasbg.fr, developed
”Advanced”onFigure3automaticallyappearwhenclick- bytheCentredeDonn´eesastronomiquesdeStrasbourgor
ing in on the ”Advanced Search” button, at the bottom CDS) for images and Specview (http://www.STScI.edu/
rightofthestandardquicksearchinterface. Thesesupple- institute/software_hardware/specview/,developedby
mentary criteria have been chosen according to the typ- STScI) for spectra. The user just has to click on the ”Al-
ical needs of the community. The available fields divide adin” or ”Specview” links visible under each data level
in five rows : (i) Data level, range of Integration time and providing a fits file. Tutorials for using APIS in general,
Mainhemisphere(NorthorSouth),(ii)rangesofSub-solar and these VO tools in particular, are available on the web
latitude and of Solar phase angle, (iii) Planetary longi- page (written document, movie).
tudesystem(SystemIIIforJupiter,NorthernorSouthern
SKR phase system for Saturn, IAU longitude system for Images. Aladin is capable of reading any 2D image or
Uranus) and range of Central meridian longitude (CML) spectrum, and enables the user to apply simple opera-
orPhase,(iv)Moonlongitudesystem(LocalTimeorLon- tions, such as plotting the various extensions of a fits file,
gitude for the main satellites of Jupiter and Saturn) and changing the contrast, mapping intensity levels, drawing
range of Local Time/Longitude and (v) Spacecraft Loca- an histogram of intensities, plotting a cut of the image,
tion (Galileo for Jupiter, Cassini for Saturn) and ranges identifying the intensity and coordinates of pixels of in-
of Distance, Latitude and Local time relative to the host terest etc. The user does not have to download Aladin
planet. before to use it, as APIS automatically launches a JAVA
Tofacilitateiterativesearches,therangesoffieldsTime applet version of the software, which opens in a new win-
interval and Integration time are automatically filled in dow. The data are then automatically sent toward Al-
whenperformingarequest. Theseparametersalsoappear adinthroughaSimpleApplicationMessagingProtocol(or
by default for each observation among the summary in- SAMP)hub. FurtherdocumentationonSAMPisprovided
formation displayed in the left column of Figure 3. When bytheInternationalVirtualObservatoryAlliance(IVOA)
performingaconditionalsearchonparametersnotpresent at http://www.ivoa.net/documents/SAMP/. Figure 4 il-
among these (say, for instance, the Local Time of Ence- lustratestheuseofAladinwithplotsofseveralextensions
ladus), they are automatically added to the list of sum- of the level 2 image visible in Figure 3. It is also possible
mary information. to upload several fits files together to compare them.
3.3. Science uses Spectra. Specview is capable of reading both 2D and 1D
spectra to display a 1D plot of intensity as a function
In the past, HST auroral observations have been most
of wavelength. It also enables the user to apply sim-
generally analyzed per observing campaign. The above
ple operations, such as coplotting different spectra, make
describeddatabaseandsearchinterfacefurtherenablethe
some simple measurements or add theoretical transitions
non-specialist, but also the specialist user, to easily find
linesretrievedfrominteroperablespectroscopycatalogsis-
individual observations of interest over a larger set of ob-
sued from the Virtual Atomic and Molecular Data Center
serving campaigns as a function of an extended space of
(VAMDC) program (http://www.vamdc.eu/). By wait-
parameters: thesub-solarlatitudetolookforseasonalef-
ingforanoperationalJAVAapplet, theuserhastodown-
fects (Figure 2), the satellite position to search for planet-
load the latest version of Specview and launch it before
satellite electrodynamic interactions, the planetary longi-
clicking on the ”Specview” link below the fits file. The
tude/phase to investigate any rotational control on mag-
data are then sent to the software through a SAMP hub.
netospheric dynamics, the date of arrival of interplane-
When reading a 2D spectrum, the user has to choose
tary shocks to study the influence of solar wind on au-
whichspatialpixelsoftheslitwillbesummeduptoderive
roral forcing or the location of a spacecraft performing
asingle1Dspectrum. Figure6displaysaSpecviewplotof
remote/in situ measurements to lead multi-instrumental
a1Dspectrumextractedfromthelevel1spectrumvisible
multi-spectral analysis. Furthermore, with standardized
in Figure 3, which reveals a typical spectrum of Saturn’s
7
auroralemissions. Thebutton”LineIDs”enablestheuser for investigating the influence of solar wind conditions on
toretrievetheoreticallinesfromasetofVAMDCcatalogs, kronian auroral emissions is presented in [24].
such as SESAM (http://sesam.obspm.fr) for H lines.
2
For instance, we superimposed the H-Lyα line (provided
5. Conclusions and perspectives
by default with Specview) and a few H lines retrieved
2
from the SESAM catalog to the observed spectrum. In this article, we have presented in detail the APIS
service. This primarily consists of a high level database,
made of several data levels derived for most of the HST-
FUV STIS and ACS spectro-imaging auroral observations
acquired since 1997, with very diverse instrumental con-
figurations. A dedicated search interface accessible on the
APISportalenablestheusertoefficientlybrowseandsort
out these data, which can be used interactively with VO
softwares such as Aladin and Specview. The database is
alsocompliantwiththeEPN-TAPprotocol,sothatitcan
alsobequeriedbyexternalVOtools. Wehavepresenteda
few examples of science studies easy to perform with this
service.
This service is in continuous evolution, with numerous
perspectives of development and improvement. First of
Figure 6: Example of 1D spectrum plotted with Specview. Several
all, although most of the data levels are already available
theoretical lines have been retrieved and plotted on the top of the
investigatedspectrumthroughthebutton”LinesIDs”: theH-Lyα for the considered targets, the level 2,3 images of Jupiter
line is provided by default with Specview, while the H2 lines have remain for instance to be derived (under progress). The
beenretrievedfromtheSESAMcatalog. datalevelsthemselvesmayalsobeupdatedtoimprovethe
existing plots (for instance by drawing the magnetic foot-
prints of Galileo and Cassini spacecraft, or the expected
4.2. Interoperability
footprintofGalileansatellites,ontopofprojections)orto
Finally,outofthesearchinterfacepresentedabove,the provide additional observables (such as the total auroral
APIS database can also be queried externally, by wider power radiated per image) and formats (such as movies).
search tools of the Virtual Observatory community such Thedateofcreationofeachdatasetbeingcontainedwithin
asTOPCAT(ToolforOperationsonTablesandCatalogs, theheaderofassociatedfitsfilesanddisplayedinthe”De-
http://www.star.bris.ac.uk/~mbt/topcat/),VESPA(Vir-tailedinformation”linkofthesearchinterface, updatesof
tual European Solar and Planetary Data Access, http: existing data levels can be traced by the user. This ser-
//vespa.obspm.fr) or the CDPP/AMDA tool (Centre vice shall also evolve as a function of the feedback of the
de Donn´ees de Physique des Plasmas/Automated Multi community. It has been and will be regularly presented
Dataset Analysis, http://amda.cdpp.eu/). at science and hands-on sessions of international meetings
To achieve such a VO interoperability, APIS has been to facilitate exchanges with the users and probe them on
built to be compliant with the Europlanet - Table Access desired further improvements.
protocol (EPN-TAP), whose aim is to enable a user to ac- Thedatabasewillthennaturallybeextendedwiththe
cess Planetary Science data in a standard manner. The observing campaigns of the Jupiter, Saturn and Uranus
EPN-TAP service relies on IVOA TAP specifications and systems to come by the end of the HST mission. But
a data model developed by the EuroPlaNet team. For it could also be extended to the other planets observed
details, the reader is referred to the description of the by STIS and ACS (such as Mars), and to the other HST
EPN-TAP protocol [22]. In practice, the set of meta- FUV spectro-imagers (such as the FOC and WFPC2, a
data attributed to each individual observation includes few images of which were already included to APIS as
the(mandatoryandoptional)parametersrequestedbythe a test). Comments or suggestions of improvement from
EPN-TAP data model to fully describe the data. These the users on these possible extensions are also welcome.
are divided in three categories : axis ranges, data descrip- In parallel, simple (blind) usage statistics can be used to
tion, and data access. The metadata are organized into a assesswhether,andatwhichdegree,APISrespondstothe
singletable(orview),whichcanbethenqueriedusingthe needs of the community.
TAP protocol between a user TAP client and the APIS Furthermore,nowthatthisserviceisfullyoperational,
TAP server. it could easily deal with much wider datasets such as the
The APIS EPN view has been officially registered by auroral spectro-imaging observations acquired by the in-
IVOA through VO-PDC. APIS was one of the first ser- strumentsonboardthemissionCassini,orbysimilarspectro-
vices connected to VESPA through EPN-TAP [23]. A imagers onboard the various past and future planetary
use-case of the CDPP/AMDA tool querying APIS data exploration spacecraft (such as Voyager, Galileo, Polar,
throughEPN-TAPandaSAMPhub(asdescribedabove)
8
Mars Express, JUNO, JUICE) or Earth-based observato- [4] S. V. Badman, G. Branduardi-Raymont, M. Galand, S. L. G.
ries (such as Chandra, XMM-Newton, IR ground-based Hess, N. Krupp, L. Lamy, H. Melin, C. Tao, Auroral Pro-
cessesattheGiantPlanets: EnergyDeposition,EmissionMech-
telescopes, JWST). Another axis of development is to in-
anisms, Morphology and Spectra, Sp. Sci. Rev.doi:10.1007/
tegrate databases of observed FUV solar spectra and/or
s11214-014-0042-x.
theoretical spectra of H2 bands (the SESAM catalog only [5] J. Gustin, J.-C. G´erard, D. Grodent, R. G. Gladstone, J. T.
providesindividuallinessofar). Finally,thisservicecould Clarke,W.R.Pryor,V.Dols,B.Bonfond,A.Radioti,L.Lamy,
J. M. Ajello, Effects of methane on giant planets UV emis-
be easily interfaced with other VO tools (such as HELIO,
sions and implications for the auroral characteristics, Journal
or the 3Dview and propagation tools at CDPP).
ofMolecularSpectroscopy)291(2013)108–117. doi:10.1016/
j.jms.2013.03.010.
[6] C. Tao, L. Lamy, R. Prang´e, The H-Lyman alpha/H2 bands
6. acknowledgements brightness ratio of FUV auroral emissions: a diagnosis for the
energyofprecipitatingelectronsandassociatedmagnetospheric
We thank CNRS and CNES for supporting the au- accelerationprocessesappliedtoSaturn,GeophysicalResearch
thors. We acknowledge all the persons involved in the Letters(2014)inpress.
[7] L. Lamy, R. Prang´e, F. Henry, P. Le Sidaner, APIS, an inter-
acquisitionofdatabytheESA/NASAHubbleSpaceTele-
active database of HST-UV observations of the outer planets,
scope (from the conceptors of the telescope and its in- Presented at the Magnetospheres of Outer Planets conference,
struments to the principal investigators of observing pro- Athens,11July,2013.
[8] W. D. Pence, L. Chiappetti, C. G. Page, R. A. Shaw, E. Sto-
grams),theSpaceTelescopeScienceInstitute(STSci)and
bie,DefinitionoftheFlexibleImageTransportSystem(FITS),
the ESA Hubble Science Archive (ESAC) for calibrating
version 3.0, Astronomy & Astrophysics 524 (2010) A42. doi:
and archiving raw data and developing interactive tools 10.1051/0004-6361/201015362.
for spectra (Specview, VOspec), the IMCCE for provid- [9] L. Lamy, Variability of southern and northern periodicities of
SaturnKilometricRadiation,Planetary,SolarandHeliospheric
ing the physical ephemeris of the outer planets and their
RadioEmissions(PREVII)(2011)38–50arXiv:1102.3099.
moons, and the Cassini, Galileo, and Voyager ephemeris
[10] M.B.Vincent,J.T.Clarke,G.E.Ballester,W.M.Harris,R.A.
toolmadeavailablebytheRadioandPlasmaWaveGroup West,J.T.Trauger,R.W.Evans,K.R.Stapelfeldt,D.Crisp,
from the Department of Physics and Astronomy of the C. J. Burrows, J. S. Gallagher, R. E. Griffiths, J. Jeff Hester,
J. G. Hoessel, J. A. Holtzman, J. R. Mould, P. A. Scowen,
University of Iowa. We acknowledge S. Cnudde and SI-
A. M. Watson, J. A. Westphal, Mapping Jupiter’s Latitudinal
GALfordesigningthewebpage,andS.Erard,B.Cecconi, BandsandGreatRedSpotUsingHST/WFPC2Far-Ultraviolet
J. Aboudarham and X. Bonnin for useful discussions on Imaging, Icarus 143 (2000) 189–204. doi:10.1006/icar.1999.
theAPISdevelopment. WethankN.MoreauandI.Busko 6232.
[11] B.Bonfond,MorphologyanddynamicsoftheIoUVfootprint,
fordevelopingaversionofSpecviewcompatiblewithAPIS
Ph.D. thesis, Universit´e de Li`ege, Acad´emie Wallonie-Europe
products, and E. Roueff and H. Abgrall for exchanges on (2010).
the SESAM database. [12] X. Li, R. Ramachandran, S. Movva, S. Graves, G. Germany,
W. Lyatsky, A. Tan, Dayglow Removal from FUV Auroral
Images, Proceedings of the Geoscience and Remote Sensing
7. Bibliography styles Symposium 6 (2004) 3774–3777. doi:10.1109/IGARSS.2004.
1369944.
[13] J.-C. G´erard, D. Grodent, J. Gustin, A. Saglam, J. T. Clarke,
References
J. T. Trauger, Characteristics of Saturn’s FUV aurora ob-
served with the Space Telescope Imaging Spectrograph, Jour-
References
nal of Geophysical Research (Space Physics) 109 (2004) 9207.
doi:10.1029/2004JA010513.
[1] A. I. Gomez de Castro, T. Appourchaux, M. A. Barstow, [14] J.-C. G´erard, D. Grodent, S. W. H. Cowley, D. G. Mitchell,
M. Barthelemy, F. Baudin, S. Benetti, P. Blay, N. Brosch, W. S. Kurth, J. T. Clarke, E. J. Bunce, J. D. Nichols, M. K.
E.Bunce,D.deMartillo,J.-M.Deharveng,R.Ferlet,K.France, Dougherty,F.J.Crary,A.J.Coates,Saturn’sauroralmorphol-
M. Garca, B. Gansicke, C. Gry, L. Hillenbrand, E. Jos- ogyandactivityduringquietmagnetosphericconditions,Jour-
selin, C. Kehrig, L. Lamy, J. Lapington, A. Lecavelier des nal of Geophysical Research (Space Physics) 111 (A10) (2006)
Etangs, F. LePetit, J. Lopez-Santiago, B. Milliard, R. Monier, 12210. doi:10.1029/2006JA011965.
G.Naletto,Y.Naze,C.Neiner,J.Nichols,M.Orio,I.Pagano, [15] D.Grodent,J.-C.G´eRard,S.W.H.Cowley,E.J.Bunce,J.T.
C. Peroux, G. Rauw, S. Shore, M. Spaans, G. Tovmassian, Clarke, Variable morphology of Saturn’s southern ultraviolet
A. ud-Doula, J. Vilchez, Building galaxies, stars, planets and aurora, Journal of Geophysical Research (Space Physics) 110
the ingredients for life between the stars. The science behind (2005)7215. doi:10.1029/2004JA010983.
theEuropeanUltraviolet-VisibleObservatory., Astrophys.and [16] J. T. Clarke, J. Nichols, J.-C. G´erard, D. Grodent, K. C.
Sp.Sci. Hansen, W. Kurth, G. R. Gladstone, J. Duval, S. Wan-
[2] G. Cessateur, T. Dudok de Wit, M. Kretzschmar, J. Lilen- nawichian,E.Bunce,S.W.H.Cowley,F.Crary,M.Dougherty,
sten, J.-F. Hochedez, M. Snow, Monitoring the solar UV irra- L. Lamy, D. Mitchell, W. Pryor, K. Retherford, T. Stallard,
diance spectrum from the observation of a few passbands, As- B.Zieger,P.Zarka,B.Cecconi,ResponseofJupiter’sandSat-
tron. & Astrophys. 528 (2011) A68. doi:10.1051/0004-6361/ urn’s auroral activity to the solar wind, Journal of Geophysi-
201015903. cal Research (Space Physics) 114 (2009) 5210. doi:10.1029/
[3] M.Barth´elemy,L.Lamy,H.Menager,M.Schulik,D.Bernard, 2008JA013694.
H. Abgrall, E. Roueff, G. Cessateur, R. Prange, J. Lilensten, [17] L. Lamy, Etude des ´emissions radio aurorales de Saturne,
Dayglow and auroral emissions of Uranus in H2 FUV bands, mod´elisation et aurores UV, Ph.D. thesis, Universit´e Paris VI
Icarus 239 (2014) 160–167. doi:10.1016/j.icarus.2014.05. (2008).
035. [18] J. Gustin, B. Bonfond, D. Grodent, J.-C. G´erard, Conversion
from HST ACS and STIS auroral counts into brightness, pre-
9
cipitatedpower,andradiatedpowerforH2 giantplanets,Jour-
nal of Geophysical Research (Space Physics) 117 (2012) 7316.
doi:10.1029/2012JA017607.
[19] L. Lamy, R. Prang´e, W. Pryor, J. Gustin, S. V. Badman,
H. Melin, T. Stallard, D.-G. Mitchell, P. C. Brandt, Multi-
spectral simultaneous diagnosis of Saturn’s aurorae through-
out a planetary rotation, Journal of Geophysical Research
(SpacePhysics)118(2013)4817–4843. arXiv:1307.4675,doi:
10.1002/jgra.50404.
[20] W. R. Pryor, A. M. Rymer, D. G. Mitchell, T. W. Hill, D. T.
Young, J. Saur, G. H. Jones, S. Jacobsen, S. W. H. Cowley,
B.H.Mauk,A.J.Coates,J.Gustin,D.Grodent,J.-C.G´erard,
L.Lamy,J.D.Nichols,S.M.Krimigis,L.W.Esposito,M.K.
Dougherty,A.J.Jouchoux,A.I.F.Stewart,W.E.McClintock,
G. M. Holsclaw, J. M. Ajello, J. E. Colwell, A. R. Hendrix,
F. J. Crary, J. T. Clarke, X. Zhou, The auroral footprint of
EnceladusonSaturn,Nature472(2011)331–333.doi:10.1038/
nature09928.
[21] S. Wannawichian, J. T. Clarke, D. H. Pontius, Interaction ev-
idence between Enceladus’ atmosphere and Saturn’s magneto-
sphere, Journal of Geophysical Research (Space Physics) 113
(2008)7217. doi:10.1029/2007JA012899.
[22] S. Erard, P. Le Sidaner, B. Cecconi, J. Berthier, F. Henry,
M.Molinaro,M.Giardino,N.Bourrel,N.Andr´e,M.Gangloff,
C.Jacquey,F.Topf,TheEPN-TAPprotocolforthePlanetary
ScienceVirtualObservatory,AstronomyandComputing(2014)
inpressarXiv:1407.5738.
[23] S. Erard, B. Cecconi, P. Le Sidaner, J. Berthier, F. Henry,
C. Chauvin, N. Andr´e, V. G´enot, C. Jacquey, M. Gangloff,
N.Bourrel,B.Schmitt,M.T.Capria,G.Chanteur,Planetary
ScienceVirtualObservatoryarchitecture,AstronomyandCom-
puting(2014)inpressarXiv:1407.4886,doi:10.1016/j.ascom.
2014.07.005.
[24] V. G´enot, N. Andr´e, B. Cecconi, M. Bouchemit, E. Budnik,
N. Bourrel, M. Gangloff, N. Dufourg, S. Hess, R. Modolo,
B. Renard, N. Lormant, L. Beigbeder, D. Popescu, J.-P. To-
niutti, Joining the yellow hub: uses of the Simple Applica-
tion Messaging Protocol in Space Physics analysis tools, As-
tronomy and Computing (2014) in pressarXiv:hal-01055883,
doi:10.1016/j.ascom.2014.07.007.
10