Collisional Excavation of Asteroid (596) Scheila

Collisional  Excavation  of  Asteroid  (596)   Scheila       D.  Bodewits1,  M.  S.  Kelley1,  J.-­‐Y.  Li1,  W.  B.  Landsman2,  S.  Besse1,  and  M.  F.  A’Hearn1           Accepted  for  publication  by  the  Astrophysical  Journal           1   Department   of   Astronomy,   U.   Maryland,   College   Park,   MD   20742.   [email protected];   [email protected];  [email protected];  [email protected];  [email protected]   2  NASA  GSFC,  Adnet  Systems,  Mailstop  667,  Greenbelt,  MD  20771.  [email protected]     4  figures,  1  table     Key   words:   minor   planets,   asteroids:   individual   (596)   Scheila   –   ultraviolet:   planetary  systems       Abstract   We  observed  asteroid  (596)  Scheila  and  its  ejecta  cloud  using  the  Swift  UV-­‐optical   telescope.   We   obtained   photometry   of   the   nucleus   and   the   ejecta,   and   for   the   first   time   measured   the   asteroid’s   reflection   spectrum   between   290   –   500   nm.   Our   measurements   indicate   significant   reddening   at   UV   wavelengths   (13%   per   103  Å)   and   a   possible   broad,   unidentified   absorption   feature   around   380   nm.   Our   measurements   indicate   that   the   outburst   has   not   permanently   increased   the   asteroid’s   brightness.   We did not detect any of the gases that are typically associated with either hypervolatile activity thought responsible for cometary outbursts (CO+, CO2+), or for any volatiles excavated with the dust (OH, NH, CN, C2, C3).   We   estimate   that   6 × 108   kg   of   dust   was   released   with   a   high   ejection   velocity   of   57   m/s   (assuming   1   µm   sized   particles).   While   the   asteroid   is   red   in   color   and   the   ejecta   have  the  same  color  as  the  Sun,  we  suggest  that  the  dust  does  not  contain  any  ice.   Based  on  our  observations,  we  conclude  that  (596)  Scheila  was  most  likely  impacted   by  another  main  belt  asteroid  less  than  100  meters  in  diameter.   1.  Introduction   Early   December   2010,   an   unexpected   dust   cloud   was   discovered   around   the   asteroid  (596)  Scheila  (Larson  et  al.  2010).  We  report  on  observations  using  the  UV-­‐ Optical  Telescope    (UVOT)  on  board  Swift.     Asteroids   have   long   been   believed   to   be   currently   geologically   inactive.   Instead,   their   shapes,   sizes,   and   surface   geology   are   dominated   by   impacts.   However,  several  asteroids  have  been  observed  to  develop  a  coma  and  or  dust  tail   (Hsieh  &  Jewitt  2006).  Some  asteroids  are  repeatedly  active  over  periods  as  long  as   months,  most  famously  133P/Elst-­‐Pizzaro  (Hsieh  et  al.  2004).  Others  are  only  active   in  short-­‐lived  bursts.  Two  main  ejecta  processes  have  been  suggested:  cometary-­‐like   outgassing   of   volatiles   driven   by   an   exoergic   interior   process   (Jewitt   2009),   and   inter-­‐asteroidal   collisions   (Jewitt   et   al.   2010;   Snodgrass   et   al.   2010).   Alternatively,   collisions  may  expose  icy  content  buried  beneath  the  surface  (Hsieh  et  al.  2004).   (596)  Scheila  is  a  T-­‐type  main  belt  asteroid  with  a  diameter  of  113  km1.  On   2010  Dec.  11.44  UT  the  asteroid  had  increased  by  mv  =  0.8  and  archival  Catalina  Sky   Survey  observations  showed  that  the  activity  was  triggered  between  2010  Nov.  11   and   Dec.   3   (Larson   et   al.   2010).   The   asteroid   orbits   the   Sun   in   5.0   yr   with   an   eccentricity  of  0.16  and  a  semi-­‐major  axis  of  2.93  AU.  Its  outburst  occurred  at  rh  =   3.1  AU,  close  to  its  aphelion  at  3.4  AU.  Scheila  offered  a  unique  target  as  we  observed   it  only  weeks  after  its  outburst;  the  outburst  of  P/2010  A2  (Linear)  was  discovered   in  January  2010  (Birtwhistle  et  al.  2010),  almost  a  year  after  it  was  triggered  (Jewitt   et  al.  2010;  Moreno  et  al.  2010,  Snodgrass  et  al.  2010).     2.  Observations   Swift  is  a  multi-­‐wavelength  observatory  equipped  for  rapid  follow-­‐up  of  gamma-­‐ray   bursts   (Gehrels   et   al.   2004).   Our   observations   use   the   UVOT   that   provides   a   17  ×  17   arcminute   field   of   view   with   a   plate   scale   of   1   arcsec/pixel   and   a   point   spread   function   of   2.5”   FWHM   (Mason   et   al.   2004).   Seven   broadband   filters   allow   color   discrimination,   and   two   grisms   provide   low-­‐resolution   (λ/δλ  =  100)   spectroscopy   between  170–650  nm.   Swift  observed  (596)  Scheila  on  2010  December  14  and  15  UT  (Table  1).  On   both   days   we   used   the   V   (λc   546.8  nm,   FWHM   75.0  nm)   and   UVW1   (λc   260.0  nm,   FWHM   70.0  nm)   filters.   On   Dec   15th   we   used   the   UV   grism   to   search   for   gaseous   emission   lines.   The   grism   was   operated   in   ‘clocked   mode’   to   suppress   background   stars  and  the  dispersion  axis  was  oriented  at  a  position  angle  of  ~260°.  The  asteroid   was  not  tracked,  and  the  proper  motion  is  6.2"  over  our  ~1300  s  exposures.   The   results   are   subject   to   several   possible   systematic   uncertainties.   The   absolute   calibration   of   UVOT2   is   accurate   to   within   0.01   magnitudes   (V)   and   0.03   magnitudes  (UVW1).  The  brightness  of  the  nucleus  results  in  significant  coincidence   loss   which   we   estimate   to   be   0.3  ±  0.05  mag   at   V   and   0.02  ±  0.005   mag   at   UVW1.   Shot   noise   and   background   subtraction   uncertainty,   as   well   as   the   aperture   and   coincidence  loss  correction  errors  are  all  included  in  the  listed  uncertainties.   1 http://ssd.jpl.nasa.gov/sbdb.cgi 2  http://swift.gsfc.nasa.gov/docs/swift/analysis/uvot_digest/zeropts.html      3.  Results   3.1  Imaging   The   images   obtained   with   the   V   and   UVW1   broadband   filters   were   registered,   binned,  and  logarithmically  stretched  to  allow  inspection  of  the  morphology  of  the   ejecta   (Fig.   1).   Both   panels   span   about   215”  x  204”,   corresponding   to   390,000  x  370,000   km.   We   used   archival   Digitized   Sky   Survey3   images   to   identify   and  remove  background  sources.  The  projected  directions  to  the  Sun  and  the  orbital   motion  almost  coincide  and  are  indicated  by  arrows.   We   have   clearly   resolved   detections   of   the   ejecta   in   both   filters.   In   the   V-­‐ band,  the  morphology  seems  to  consist  of  a  broad  ‘northern’  fan  from  Position  Angle   (PA)  =  330  to  60°,  and  a  curved  ‘southern’  feature  at  PA  =  180°.     Aperture   photometry   on   the   images   yields   an   average   mv   =   14.1  ±  0.07   and   muv  =  16.4  ±  0.09  for  the  asteroid  in  a  10  pixel  radius  aperture.  We  made  no  attempt   to  remove  the  dust  emission  within  our  asteroid  aperture.  Subtracting  the  nucleus’   brightness,   the   total   dust   brightness   in   a   160”  ×  100”   aperture   is   mv   =   14.4  ±  0.08.   The   northern   plume   is   brighter   than   the   southern   plume   (65   vs.   35%   of   the   total   ejecta   flux).   The   ejecta   are   faint   in   UVW1   (Fig.   1)   and   best   measured   in   a   smaller   80”  x  80”  aperture  to  minimize  uncertainties  from  the  background  subtraction.  We   find  an  average  mv  =  14.8  ±  0.08  and  muv  =  16.8  ±  0.11  for  the  ejecta  in  the  smaller   aperture   (Table   1   and   Figure   1).   Our   measurements   are   close   to   the   asteroid’s   predicted   brightness   (mv  =  14.214,   with   an   estimated   maximum   amplitude   of   0.095).   Within   the   uncertainties,   the   ejecta   (UVW1–V   =   1.91  ±  0.1)   have   solar   colors   (UVW1–V  =  1.94),  while  the  asteroid  (UVW1–V  =  2.36  ±  0.11)  is  redder  than  the  Sun.       3.2  Spectroscopy   Figure   2   shows   the   asteroid’s   spectrum   extracted   from   a   rectangular   region   13   pixels   wide   along   with   a   solar   spectrum   obtained   on   2010   Dec.   15   (SORCE;   McClintock   et   al.   20056).   The   spectrum   below   280   nm   was   contaminated   by   a   background   star.   The   data   were   binned   by   a   factor   of   8   to   improve   the   signal   to   noise,   and   to   suppress   the   Mod8   pattern   in   the   grism   image.   For   comparison,   the   grism   spectrum   of   comet   C/2007  N3   (Lulin)   is   shown   in   Fig.   2   (Bodewits   et   al.   2011).   While   Lulin’s   spectrum   is   dominated   by   the   emission   features   of   several   gaseous  species  (e.g.  OH,  CN,  C2,  C3,  NH,  CO2+),  there  is  no  evidence  of  any  gaseous   emission   feature   in   the   asteroid’s   spectrum.   Assuming   an   OH   fluorescence   feature   similar   to   that   in   Lulin’s   spectrum   (50  nm   wide   due   to   the   gas   distribution),   a   1σ   flux  upper  limit  of  1  ×  10-­‐13  erg  s-­‐1  cm-­‐2   A-­‐1,  and  a  total  OH  emission  fluorescence  of   2.2  ×  10-­‐16  erg  s-­‐1  cm-­‐2   molecule-­‐1   (Schleicher   &   A’Hearn   1988)   yields   a   an   upper   limit   3  ×  1030   OH   molecules   around   the   nucleus.   Considering   the   long   lifetime   of   OH   at   3.1  AU   from   the   Sun   (~106   s,   Huebner   et   al.   1992),   the   water   production   rate   3  http://archive.stsci.edu/cgi-­‐bin/dss_form   4  http://ssd.jpl.nasa.gov/?horizons 5 6 http://sbn.psi.edu/ferret/ http://lasp.colorado.edu/lisird/sorce/sorce_ssi/index.html would   be   <1025   molecules  s-­‐1,   1000   times   lower   than   typically   observed   in   Jupiter   family   comets   at   1   AU.   Although   we   cannot   constrain   the   gaseous   output   at   the   onset  of  the  anomaly,  we  conclude  that  the  asteroid  did  not  produce  significant  OH   or  water  gas  during  the  observations.       Scaling   the   solar   spectrum   yields   a   best   fit   of   mv  =  14.3  ±  0.1,   in   agreement   with  our  broadband  observations.  We  derived  the  asteroid’s  albedo  using  Equation   2,  assuming  Φ(α)  =  0.45,  a  diameter  of  113  km  (Dunham  &  Herald,  20055),  and  the   solar   irradiance   from   SORCE   (Fig   3).   We   combined   this   with   a   scaled   reflectance   spectrum   (Bus   &   Binzel   20024).   Our   data   are   in   excellent   agreement   with   albedo   measurements   from   other   observations   (Dunham   &   Herald,   20055;   Ryan   &   Woodward  2010).  Scheila’s  spectrum  is  significantly  reddened  compared  to  the  Sun.     There   appears   to   be   a   broad   absorption   feature   between   320  –  420  nm,   centered  at  about  380  nm,  superimposed  on  the  red  slope.  It  is  difficult  to  interpret   the   UV   spectrum   of   Scheila   due   to   both   the   limited   number   of   asteroids   observed   at   this  wavelength  (Butterworth  and  Meadows,  1985;  Roettger  and  Buratti,  1994),  and   due   to   the   lack   of   laboratory   reflectance   spectra   of   possible   compositional   minerals.   UV  absorptions  at  slightly  shorter  wavelengths  are  present  with  various  intensities   for   possibly   hydrated   asteroids   such   as   (1)   Ceres,   (2)   Pallas,   and   (102)   Hygiea   (Butterworth   and   Meadows,   1985;   Roettger   and   Buratti,   1994;   Li   et   al.,   2006;   2009;   Rivkin  et  al.,  2006;  Milliken  and  Rivkin,  2009).  UV  spectra  of  S-­‐types  show  a  strong   red   slope   and   sharp   drop   at   300   to   400   nm,   and   very   weak   UV   absorptions.   As   a   primitive   type   asteroid   in   the   outer   main   asteroid   belt   with   ~4%   albedo,   Scheila   probably   has   a   surface   composition   that   is   closer   to   carbonaceous   meteorites.   In   order   to   better   understand   the   UV   spectra   of   asteroids,   observations   of   various   spectral   types   and   laboratory   measurements   of   carbonaceous   meteorites   are   needed.       4.  Discussion   4.1  Ejecta  Dynamics   The   sunward   extend   of   the   cloud   is   ~15”,   corresponding   to   2.6  ×  104   km   at   the   asteroid.  This  is  the  distance  XR  where  solar  radiation  turns  dust  particles,  and  it  can   be   used   to   estimate   the   grain   ejection   velocity   (Hsieh   et   al   2004,   Jewitt   &   Meech   1987):     1   v d2 rh2 XR ~   2 "d gsun where   vd   is   the   relative   ejection   velocity,   rh   the   heliocentric   distance   in   AU,   β   the   dimensionless   ratio   between   radiation   pressure   acceleration   and   the   local   gravity,   and  gsun  =  0.006  m  s-­‐2  is  the  gravitational  acceleration  to  the  Sun  at  1  AU.  Using  this   we  find  vd  =  57√β  m  s-­‐1.     This   velocity   is   an   order   of   magnitude   larger   than   those   reported   for   133P   (1.5√β  m  s-­‐1;   Hsieh   et   al.   2004)   and   P/2010   A2   (1.1√β   m   s-­‐1   Moreno   et   al.   2010).   This  difference  is  partially  explained  by  the  different  escape  velocities.  Assuming  a   density   of   2500  kg  m-­‐3,   and   the   estimated   diameters   (113  km   vs.   0.12  km),   the   escape  velocities  are  75  m  s-­‐1  for  596  and  0.06  m  s-­‐1  for  P/2010  A2.  Additionally,  the   collision  disrupting  A2  may  have  produced  a  lot  of  high  velocity  dust,  but  this  had   left  the  vicinity  of  A2  over  the  year  between  the  impact  and  its  discovery.     We   simulated   the   ejection   of   dust   from   596   to   better   understand   the   dynamics   of   the   observed   ejecta.     The   simulation   accounts   for   the   gravitational   acceleration   of   the   Sun   and   planets   and   solar   radiation   pressure   (Kelley   2006;   Kelley   et   al.   2009).     The   model   was   designed   to   simulate   observations   of   comet   comae,  and  does  not  account  for  the  mass  of  the  parent  body.    Since  596  is  a  large   asteroid,   our   simulation   is   only   a   rough   approximation   to   the   true   dust   ejection   dynamics.   We   ejected   106   grains   isotropically   from   the   surface   in   a   single   event,   with  a  distribution  of  sizes  (0.1  –  104  µm)  and  velocities  (0  –  100  m  s-­‐1)  on  2010  Dec.   3,   and   integrated   their   positions   forward   in   time   to   2010   Dec.   15.     The   full   extent   of   the   dust   in   our   UVOT   images   can   be   encompassed   in   a   circle   centered   on   the   asteroid  with  a  radius  of  76”  (140,000  km).    Our  simulations  show  that  12  days  after   the   outburst,   only   grains   >1   µm   in   radius   remain   within   this   aperture.     Therefore,   we  adopt   1   µm  as   a   lower-­‐limit  to  the  ejected  grain  size.    A  stricter  limit  to  the  grain   radii   can   be   estimated   if   the   outburst   date   can   be   better   constrained,   and   if   the   dynamics   of   the   ejecta   could   be   reproduced,   neither   of   which   have   we   attempted.   Independent   of   these   refinements,   all   grains   smaller   than   ~1   µm   would   be   found   outside  the  observed  extent  of  the  dust  in  our  V-­‐band  image.  Therefore,  we  conclude   that  sub-­‐micron  sized  grains  are  few  or  absent  in  the  ejecta.     We  continued  our  simulation  forward  in  time  with  10-­‐day  steps  to  estimate   the   lifetime   of   the   ejecta   within   our   76”   radius   aperture.   We   have   weighted   the   dust   grain   cross   sections   by   v-­‐1   to   approximate   crater   formation   dynamics   (more   mass   is   ejected  at  lower  velocities),  and  converted  cross  section  to  V-­‐band  magnitudes  using   Eq.   2   (Figure   4).     The   evolution   of   the   dust   results   in   a   rapid   decline   of   mv   of   approximately  3  magnitudes  over  the  first  month.    If  the  ejecta  are  devoid  of  large   grains   (i.e.,   the   grain   size   upper-­‐limit   is   10   μm,   rather   than   our   initial   assumption   of   104   μm),   then   the   dust   will   be   even   fainter,   falling   5   magnitudes   by   mid-­‐January.     According   to   our   simulations,   unless   the   grain   radii   extend   at   least   into   the   millimeter   size   range,   we   expect   that   the   surface   brightness   of   the   ejecta   cloud   around  Scheila  will  be  too  faint  to  observe  in  early  2011.     4.2  Dust  Mass The  total  reflection  cross  section  Cd  is  related  to  the  difference  between  the  V-­‐band   magnitudes   of   dust   and   the   apparent   solar   magnitude   (msun  =  -­‐26.74)   by   the   following  relation:     2 2   rh Δ2 10 −0.4( md − msun ) 22   C d = 2.25 × 10 π pV Φ (α ) where   the   heliocentric   distance   rh   =   3.1   AU,   the   geocentric   distance  Δ = 2.5 AU,   pV   is   the  geometric  albedo,  and  Φ(α)  the  phase  darkening.  Assuming  a  phase  correction   of   0.75   for   a   solar   phase   angle   of   16   degrees   (Schleicher   et   al.   1998),   and   a   dust   albedo   of   0.1   (Jewitt   2009),   we   find   Cd  ~  2  ×  109   m2   for   the   ejecta.   Further   assuming   a   power   law   distribution   of   spherical   particles   with   radii   between   10-­‐6  –  10-­‐2   m   (section   4.1),   with   a   slope   of   -­‐3.5,   and   a   density   of   2500   kg   m-­‐3,   we   can   use   the   relation   between   mass   and   scattering   cross   section   derived   by   Jewitt   (2009).   We   then  find  that  there  was  approximately  6  ×  108  kg  of  dust  around  the  asteroid.     To   put   this   in   perspective,   this   is   a   much   larger   mass   than   found   around   active  centaurs  (104  –  106  kg,  Jewitt  2009),  comparable  to  the  mass  released  by  the   destruction  of  P/2010  A2  (LINEAR)  (5  –  60  ×  107  kg;  Moreno  et  al.  2010,  Snodgrass   et  al.  2010,  Jewitt  et  al.  2010),  and  comparable  to  the  total  amount  of  dust  and  gas   excavated   by   the   Deep   Impact   probe   colliding   with   9P/Tempel   1   (106  –  109   kg,   Küppers  et  al.  2005;  A’Hearn  et  al.  2008).     4.3  Asteroid  Impact  Scenario   We  did  not  detect  any  evidence  of  volatiles  that  are  typically  associated  with  the   hypervolatile   activity   (CO+,   CO2+),   or   of   any   volatiles   that   were   possibly   released   from   the   ejected   dust   (OH,   NH,   CN,   C2,   C3).   The   reddening   of   the   asteroid  (13% per 103 Å) and   the   neutral   color   of   the   ejecta   are   comparable   to   those   of   comet   surfaces   and   comae,   which   are   generally   considered   to   be   ice-­‐free.   Comets   9P/Tempel-­‐1,   21P/Giacobini-­‐Zinner,   and   67P/Churyumov-­‐Gerasimenko   have   nucleus   spectral   reflectivity   gradients   of   12.5  ±  1%,   12.8  ±  2.7%,   and   11  ±  2%,   and   comae   gradients   of   about   8%,   8  ±  3%,   and   -­‐1   to   1%,   respectively   (Li   et   al.   2007;   Luu   et   al.   1993;   Tubiana  et  al.  2008;  Schleicher  et  al.  2007;  Kolokolova  et  al.  1997;  Weiler  et  al.  2004;   Lara  et  al.  2005).  We  therefore  conclude  that  there  were  no  detectable  amounts  of   gas   or   ice   around   the   asteroid   during   our   observations.   However,   most   fragment   species  observed  in  optical/UV  wavelengths  have  lifetimes  corresponding  to  about   11.1   days   at   the   heliocentric   distance   of   the   asteroid   (Huebner   et   al.   1992).   Icy   grains  have  even  shorter  lifetimes  (Bockelée-­‐Morvan  et  al.  2001).  While  we  cannot   rule  out  that  gas  or  ice  was  released  at  the  onset  of  the  outburst  (before  Dec  3rd),  our   results  do  imply  that  the  excavation  event  did  not  trigger  any  continuous  activity.       The  plumes  are  consistent  with  ejecta  moving  away  from  the  asteroid  and   then   being   pushed   along   the   anti-­‐solar   direction   by   solar   radiation   pressure.   The   observed   morphology   could   be   well   explained   by   an   oblique   impact   where   one   plume   would   be   associated   with   downrange   ejecta   and   the   other   with   material   ejected  vertically  due  to  the  mechanical  excavation  by  shockwaves  generated  by  the   impact  (see  e.g.  Schultz  et  al.  2007).       We  estimated  above  that  about  6 × 108  kg  of  dust  was  released.  Assuming  a   density  of  2500  kg  m-­‐3,  this  corresponds  to  a  sphere  with  a  diameter  of  80  meters.  If   we  assume  that  most  of  the  momentum  input  from  the  impact  is  carried  out  by  the   ejecta,   and   an   impact   velocity   of   5  km  s-1,   we   find   an   impactor   diameter   of   20   m,   consistent   with   our   first   estimate.   If   the   observed   dust   was   indeed   ejected   by   an   impact,   then   the   ejecta   probably   constitute   only   10   –  20%   of   the   total   excavated   mass.  The  mass  of  the  material  excavated  by  hypervelocity  impacts  usually  exceeds   the   mass   of   the   projectile,   depending   on   many   parameters   such   as   the   impact   velocity   and   geometry,   porosity,   and   the   strength   of   the   target   (Holsapple   et   al.   2002,   Schultz   et   al.   2005).   We   conclude   that   the   projectile   was   therefore   probably   <100  m  in  diameter.   How   likely   is   this   collision   scenario?   Asteroid   collisions   have   been   extensively   studied   for   evolution   studies   (Bottke   et   al.   1994,   Davis   et   al.   2002).   In   the   main   belt,   the  most  likely  collision  partners  are  other  main  belt  asteroids.  Bottke  et  al  (1994)   give  an  intrinsic  collision  probability  of  3  ×  10-­‐18  yr-­‐1  km-­‐2  at  a  relative  velocity  of  5.3   km  s-­‐1.  Multiplying  this  with  the  square  of  the  radius  of  Scheila  and  N,  the  number  of   projectiles   sized   10  –  100   meter   yields   the   impact   frequency.   A   compilation   of   different  distribution  models  suggests  that  the  number  of  projectiles  is  in  the  range   1010  –  1011   (Davis   et   al.   2002).   Based   on   this   estimate,   an   object   the   size   of   (596)   Scheila   would   be   impacted   by   a   small   asteroid   once   every   1,000   years.     There   are   ~200  asteroids  the  size  of  (596)  Scheila  (Tedesco  et  al.  2004),  thus  an  event  like  this   would  occur  once  every  5  years,  and  collisions  with  asteroids  smaller  than  10  meter   should  occur  even  more  often.  Based  on  statistics,  a  collisional  impact  is  thus  very   likely.   More   tentatively,   it   is   possible   to   estimate   what   effect   the   impact   had   on   the   surface   of   Scheila.   Craters   on   (243)   Ida,   (951)   Gaspra,   (25143)   Itokawa,     (2867)   Steins,   and   (443)   Eros   have   different   depth-­‐to-­‐diameter   ratios   that   are   linked   to   the   history  of  the  asteroid  itself  (Sullivan  et  al  1996,  Carr  et  al.  1994,  Hirata  et  al.  2009,   Besse  et  al.  2011,  Robinson  et  al.  2002).  Assuming   spherical   craters,  the  diameter  D   is  related  to  the  volume  V  ~  0.06  D3  (Scheeres  et  al.  2002;  depth-­‐to-­‐diameter  ratio  of   0.15).   Based   on   the   equivalent   volume   of   the   ejecta   (3  ×  105   m3),   and   considering   that   the   ejected   mass   may   only   constitute   10   –   20%   of   the   total   excavated   mass,   the   impact  may  have  resulted  in  a  crater  with  a  diameter  of  300  meters.  Fresh  craters   often   have   a   higher   depth-­‐to-­‐diameter   ratio   than   aged   craters,   consequently   the   impact   may   well   have   created   a   crater   with   a   diameter   much   larger   than   300   meters.  An  impact  of  that  size  is  not  unexpected  on  the  surface  of  asteroids.  During   its   recent   flyby   of   the   asteroid   (21)   Lutetia,   which   has   a   size   comparable   to   (596)   Scheila,   the   Rosetta   spacecraft   discovered   numerous   craters   with   diameters   larger   than   500   meters,   likely   the   scars   of   impacts   similar   to   the   one   that   occurred   on   (596)  Scheila.     5.  Conclusions   We  obtained  optical  and  UV  images  and  spectroscopy  of  asteroid  (596)  Scheila  using   the   Swift   UVOT   within   at   most   6   weeks   after   it   first   started   to   show   a   dramatic   increase   in   brightness.   We   obtained   the   first   measurement   of   its   reflectance   spectra   below  450  nm,  a  wavelength  regime  in  which  asteroids  are  not  very  well  explored.   Whatever   caused   the   ejection   of   dust   from   Scheila   has   not   permanently   increased   the   asteroid’s   brightness.   We   did   not   detect   any   of   the   gases   that   are   typically   associated   with   either   hypervolatile   activity   thought   responsible   for   cometary   outbursts  (CO+,  CO2+),  or  for  any  volatiles  excavated  with  the  dust  (OH,  NH,  CN,  C2,   C3).  We  estimate  that  6  ×  108   kg  of  dust  was  released  with  an  ejection  velocity  of  57   m   s-­‐1     (assuming   1   micron   sized   particles).   The   ejecta   have   the   same   color   as   the   Sun,   while   the   asteroid   appears   reddened.   The   two   colors   are   consistent   with   observations   of   comet   comae   and   nuclei,   both   generally   considered   to   be   ice-­‐free.   Based   on   our   observations,   we   conclude   that   Scheila   was   most   likely   impacted   by   another  main  belt  asteroid  with  a  diameter  of  at  most  100  m.     Acknowledgements   We  thank  the  Swift  team  for  use  of  Director’s  Discretionary  Time  and  for  the  careful   and   successful   planning   of   these   observations.   We   thank   Pete   Schultz   for   valuable   discussions  on  this  study.     References     A'Hearn, M. F. 2008. Space Sci. Rev. 138, 237 Besse,  S.,  et  al.  2011,  Icarus  submitted   Birtwhistle,  P.,  Ryan,  W.  H.,  Sato,  H.,  Beshore,  E.  C.,  &  Kadota,  K.  2010,  IAUC,  9105,  1   Bodewits,  D.,  Villanueva,  G.L.,  Mumma,  M.J.,  Landsman,  W.B.,  Carter,  J.A.,  and  Read,   A.M.  2011,  AJ,  141,  12.   Bockelée-­‐Morvan  et  al.  2001,  Science  292,  1339   Bottke,  W.F.,  Nolan,  M.C.,  Greenberg,  R.,  Kolvoord,  R.A.  1994,  Icarus,  107,  255   Bus,  S.J.,  and  Binzel,  R.P.,  2002,  Icarus  158,  1   Butterworth,  P.S.,  and  Meadows,  A.J.,  1985,  Icarus  62,  305-­‐318   Carr,  M.H.,  et  al.  1994,  Icarus  107,  61   Davis,  D.R.,  Durda,  D.D.,  Marzari,  F.,  Campo  Bagatin,  and  Gil-­‐Hutton,  R.  2002,  in   Asteroids  III.  Ed.  Bottke  Jr.,  W.F.,  Cellino,  A.,  Paolichhi,  P.,  and  Binzel,  R.P.    (Tuscon,   AZ:  Univ.  Arizona  Press),  545   Dunham,  D.  W.  and  Herald,  D.,  2005,  Asteroid  Occultation  List.  EAR-­‐A-­‐3-­‐RDR-­‐ OCCULTATIONS-­‐V3.0:OCCLIST_TAB.  NASA  Planetary  Data  System   Gehrels,  N.,  et  al.  2004,  ApJ,  611,  1005   Hirata,  N.  et  al.  2009,  Icarus  200,  486-­‐502   Holsapple,  K.,  Giblin,  I.,  Housen,  K.,  Nakamura,  A.,  and  Ryan,  E.2002,  in  Asteroids  III.   Ed.  Bottke  Jr.,  W.F.,  Cellino,  A.,  Paolichhi,  P.,  and  Binzel,  R.P.    (Tuscon,  AZ:  Univ.   Arizona  Press),  443   Hsieh,  H.H.,  &  Jewitt,  D.C.  2006,  Science,  312,  561   Hsieh,  H.H.,  Jewitt,  D.C.,  and  Fernandez,  Y.R.  2004,  ApJ,  127,  2997     Huebner,  W.F.,  Keady,  J.J.,  and  Lyon,  S.P.  1992,  Astroph.  Sp.  Sci.,  195,  1   Jewitt,  D.C.,  &  Meech,  K.J.  1987,  ApJ  317,  992   Jewitt,  D.C.  2009,  AJ,  137,  4296.     Jewitt,  D.,  Waever,  H.A.,  Agarwal,  J.,  Mutchker,  M.,  Drahus,  M.,    Nature  (2010)  vol.  467   (7317)  pp.  817-­‐819   Kelley,  M.  S.  2006,  Ph.D.  Thesis,  University  of  Minnesota,  Minneapolis,  MN   Kelley,  M.  S.,  Wooden,  D.  H.,  Tubiana,  C.,  Boehnhardt,  H.,  Woodward,  C.  E.,  &  Harker,   D.  E.  2009,  AJ,  137,  4633   Kolokolova,  L.,  Jockers,  K.,  Chernova,  G.,  Kiselev,  N.,  1997,  Icarus  126,  351.   Küppers,  M.,  et  al.  2005,  Nature,  437,  987.   Lara,  L.M.,  de  Leon,  J.,  Licandro,  J.,  Gutierrez,  P.J.,  2005,  Earth,  Moon,  and  Planets  97,   165.   Larson,  S.,  et  al.  2010,  CBET,  2583   Li,  J.-­‐Y.,  McFadden,  L.A.,  Parker,  J.Wm.,  Young,  E.F.,  Stern,  S.A.,  Thomas,  P.C.,  Russell,   C.T.,  Sykes,  M.V.,  2006.    Icarus  182,  143-­‐160.   Li,  J.-­‐Y.,  et  al.,  2007,  Icarus  187,  41.   Li,  J.-­‐Y.,  McFadden,  L.A.,  A’Hearn,  M.F.,  Feaga,  L.M.,  Russell,  C.T.,  Coradini,  A.,  De   Sanctis,  C.,  Ammannito,  E.,  2009.    40th  LPSC,  abstract  #2101.   Luu,  J.  X.,  1993,  Icarus  104,  138.   Mason,  K.  O.,  Breeveld,  A.,  Hunsberger,  S.  D.,  James,  C.,  Kennedy,  T.  E.,  Roming,  P.  W.   A.,  &  Stock,  J.  2004,  in  UV,  X-­‐Ray,  and  Gamma-­‐Ray  Space  Instrumentation  for   Astronomy  XIII,  Vol.  5165,  ed.  K.  A.  Flanagan  &  O.  H.,W.  Siegmund  (Bellingham,  WA:   SPIE),  277   McClintock,  W.  E.,  Rottmann,  G.  J.,  and  Woods,  T.N.  2005,  Solar  Physics,  230,  225   Milliken,  R.E.,  Rivkin,  A.S.,  2009.    Nature  Geosci  2,  258-­‐261.   Moreno,  F.,  et  al.  2010,  ApJ  718,  L132.     Rivkin,  A.S.,  Volquardsen,  E.L.,  Clark,  B.E.,  2006.    Icarus  185,  563-­‐567.   Robinson,  M.S.,  et  al.  2002,  Meteoritics  and  Planetary  Science,  37,  1651   Roettger,  E.E.,  Buratti,  B.J.,  1994.    Icarus  112,  496-­‐512.   Ryan,  E.  L.,  &  Woodward,  C.  E.  2010,  AJ,  140,  933   Scheeres,  D.J.,  Durda,  D.D.,  and  Geissler,  P.E.  2002,  in  Asteroids  III.  Ed.  Bottke  Jr.,   W.F.,  Cellino,  A.,  Paolichhi,  P.,  and  Binzel,  R.P.    (Tuscon,  AZ:  Univ.  Arizona  Press),  527   Schleicher,  D.G.,  &  A’Hearn,  M.F.,  1988,  ApJ  331,  1058   Schleicher,  D.G.,  Millis,  R.L.,  and  Birch,  P.V.  1998,  Icarus  132,  397   Schleicher,  D.G.,  2007,  Icarus  190,  406.   Schultz,  P.H.,  Ernst,  C.M.,  and  Anderson,  J.L.B.  2005,  Space  Sci.  Rev.,  117,  207   Schultz,  P.H.,  et  al.,  2007,  Icarus  190,  295   Snodgrass,  C.,  et  al.  2010,  Nature,  467,  814   Sullivan  et  al.  1996,  Icarus,  120,  119   Tedesco,  E.F.,  Noah,  P.V.,  Noah,  M.,  &  Price,  S.D.  2004,  NASA  Planetary  Data  System,   IRAS-­‐A-­‐FPA-­‐3-­‐RDR-­‐IMPS-­‐V6.0   Tubiana,  C.,  Barrera,  L.,  Drahus,  M.,  Boehnhardt,  H.,  2008,  A&A  490,  377.   Weiler,  M.,  Rauer,  H.,  Helbert,  J.,  2004,  A&A  414,  749.      Figures  &  Tables       Figure  1.  UVOT  images  obtained  with  the  V  (left)  and  UVW1  (right)  filters  on  Dec   15th  2010  UT.  The  physical  scale  and  orientation  is  the  same  in  both  frames.  Both   images   are   scaled   logarithmically,   but   stretched   and   smoothed   separately   to   enhance   the   ejecta.   The   untracked   asteroid   appears   slightly   elongated   in   the   northward   direction   (µ).   Two   plumes   can   be   seen,   and   their   morphology   is   consistent  with  ejecta  moving  away  from  the  asteroid  and  then  being  pushed  in  the   anti-­‐solar  direction  by  solar  radiation  pressure.  The  yellow  boxes  indicate  the  area   from  which  we  obtained  the  color  (small)  and  total  (large)  photometry.       OH NH CN C3 C2   Figure  2.  Spectra  of  Scheila  (black),  comet  C/2007  N3  (Lulin)  (blue;  Bodewits  et  al.   2011)  normalized  at  450  nm)  and  the  Sun  (red;  McClintock  et  al.  2005),  scaled  to  mv   =   14.3.   Scheila’s   spectrum   is   reddened   compared   to   the   Sun.   There   is   no   evidence   of   volatile   emission   lines   that   dominate   Lulin’s   spectrum   (labeled).   Error   bars   indicate   1σ  stochastical  errors;  the  systematic  uncertainty  is  approximately  25%.              Figure   3.   The   albedo   of   Scheila.   (•)   Swift   measurements;   (+)   Reflectance   measurements   by   Bus   &   Binzel   (2002),   scaled   to   match   our   albedo   measurement.   Error  bars  indicate  1σ  stochastical  errors.  Our  data  are  in  excellent  agreement  with   other   albedo   measurements.   We   tentatively   identify   a   broad   absorption   feature   between  320–420  nm,  indicated  by  the  question  mark.            Figure  4:    Simulated  apparent  V-­‐band  magnitude  of  dust  within  76”.    Light  curves   are   given   for   grain   radius   upper-­‐limits   of   10   to   104  μm.   The   larger   radiation   pressure   efficiency   of   smaller   grains   removes   them   from   the   field   of   view   quickly,   reducing   the   reflection   cross   section   of   the   dust.   We   expect   that   the   surface   brightness  of  the  ejecta  cloud  will  soon  be  too  faint  to  observe.              Table  1:  Observing  log  and  photometry  results.       Start  Observing  Time   (UT)   2010-­‐12-­‐14T06:47   2010-­‐12-­‐14T07:09   2010-­‐12-­‐15T06:56   2010-­‐12-­‐15T08:29   2010-­‐12-­‐15T08:50   Filter     V   UVW1   UV  grism   V   UVW1   Exp.  Time   (seconds)   1285   701   1072   1300   1087   m  Nucleusa     14.1  ±  0.05   16.4  ±  0.07   14.3  ±  0.1   14.0  ±  0.05   16.3  ±  0.04   m  Ejectaa,b   (160”×100”)   14.4  ±  0.06   -­‐   -­‐   14.3  ±  0.06   -­‐   m  Ejectaa,b   (80”×80”)   14.8  ±  0.06   16.7  ±  0.08   -­‐   14.9  ±  0.06   16.8  ±  0.08       Notes:     a  Errors  include  both  stochastic  and  absolute  uncertainties.   b  The  UVW1  ejecta  is  faint  and  best  measured  in  the  smaller  80”  x  80”  aperture  to   minimize  background  stars  and  the  systematic  uncertainty.