Diapositiva 1

1
Product validation Towards a best practice
guidelines IPG Paris Aline Peltier,
Jean-Christophe Komorowski Edouard Kaminski
2

• Product validation
• Towards a best practice guidelines
• New results on the validation
• of thermal products on effusive volcanoes

3
Validation of thermal results on Le Piton de la
Fournaise (La Réunion) Recent high resolution Hot
spots (Spot scenes) visualized on google Earth
09 April 2007
4
Validation of thermal results on Le Piton de la
Fournaise (La Réunion) Recent high resolution Hot
spots (Spot scenes) visualized on google Earth
30 April 2007
09 April 2007
5
09 April 2007
Lava fountain activity
30 April 2007
Lava tube activity
Lava flow mapping Field observations (Staudacher
et al., 2009)
6
Validation of thermal results on Le Piton de la
Fournaise (La Réunion) Recent high resolution Hot
spots (Spot scene) visualized on google
Earth Results the data correspond to the
progressive change of activity from lava
foutaining to lava tube. Implications for best
user guidelines provides criteria to "quantify"
data quality, even in the case of underground
lava tube (minimal thermal anomaly for
unambiguous signal) - no available ground thruth
thermal activity
7
Validation of thermal results on Le Piton de la
Fournaise (La Réunion) Recent high resolution Hot
spots (Aster image) visualized on google Earth
09 April 2007
8
Validation of thermal results on Le Piton de la
Fournaise (La Réunion) Recent high resolution Hot
spots (Aster image) visualized on google Earth
05-06 April 2007 Crater collapse
400
9
Validation of thermal results on Le Piton de la
Fournaise (La Réunion) Recent high resolution Hot
spots (Spot scene) visualized on google
Earth Results the data correspond to the
cooling of the caldera via fumerolic activity
after the collapse. Field temperatures were
measured in the range of 300-400 C a few days
after the collapse. On the day of the collapse
temperatures were likely higher due to the during
previous eruption phase few weeks before (see
photo) Implications for best user guidelines
provides criteria to "quantify" data quality, in
the case of low thermal anomaly (minimal thermal
anomaly for unambiguous signal)
10

• Product validation
• Towards a best practice guidelines
• (2) New results on the validation
• of thermal products on explosive volcanoes

11
Validation of thermal results on Montserrat
(Lesser Antilles) Recent low resolution Hot spots
(MODIS and ASTER scene) visualized on local IPGP
ArcGis project and SPOT 2006 background scene
Results the data correspond to the cooling of
recently emplaced pyroclastic flows, and/or to
residual thermal signature of active lava
dome. Implications for best user guidelines
provides criteria to "quantify" data quality for
an explosive volcano (minimal thermal anomaly
for unambiguous signal)
12
(No Transcript)
13
(No Transcript)
14
(No Transcript)
15
Conclusions Validation of thermal results on
Montserrat (Lesser Antilles)

• suggest development of google earth interface
  like for PDF to view the hot spots directly, have
  access to TC data as a table and overlay - local
  ArcGis is cumbersome and not widely accessible
  (need to load all the files manually)
• can higher resolution thermal data be available
  for Montserrat like for PDF (SPOT, ASTER) ?
• MOD and MYD hot spots still suffer from apparent
  geometric artefacts or do the HS identify grossly
  a hot spot enveloppe ?
• there are some systematic significant
  localization errors and HS are offset from their
  logical expected position in the field by several
  hundred meters or is this due to diffuse thermal
  in atmosphere associated with co-PF ash cloud ?

16

• Globvolcano products did not detect the renewal
  of activity with ash-venting on 4-7 Oct 2009 and
  new dome of 9 Oct, first GV HS detected is on
  2009-10-11 at 0215 UTC in the White River to the
  South despite visible incandescence on 10-10, PFs
  of 10-11 to N-NW (Tyers Ghaut) are not detected
• temperatures are compatible with pyroclastic flow
  deposits but no other thermal ground truth, need
  to have MVO validation (new staff member)
• prominent incandescence of dome is not detected
  systematically or it is not evident from the
  current location of hot spot that it is picking
  up dome incandescence
• problem of residual heat from previous PF
  deposits that shows as hot spots (HS with no
  associated contemporaneous activity) but other
  large PF well identified with HS and sometimes
  suggests that the Globvolcano product is
  complementary to MVO (blind views)
• what is the role of vigorous degassing and ash
  venting and co-PF ash cloud/surges at explosive
  volcanoes in masking the HS signal and diffusing
  it over wider area ?

17

• Product validation
• Towards a best practice guidelines
• (3) New results on the validation
• of deformation products
• on effusive volcanoes

18
Validation of PS Insar results on Le Piton de la
Fournaise (La Réunion) Bulk comparison of data at
the scale of the edifice (base of the volcano)
09 April 2007
19
Validation of PS Insar results on Le Piton de la
Fournaise (La Réunion) Bulk comparison of data at
the scale of the edifice (summit of the volcano)
09 April 2007
20
Validation of PS Insar results on Le Piton de la
Fournaise (La Réunion) Bulk comparison of data at
the scale of the edifice
Results the trend inferred from satellite
results are consistent with GPS data and their
associated error bars. Anomalous results are
related to highest deformation rates not well
recorded by PS Insar (lack of signal coherence).
In this case the PSinSAR data trend could be
re-aligned with the GPS data trend that is more
robust in case of high deformation rates.
However in these cases the difference between the
PSInSAR and the GPS data trends is at most 10-15
and the trends go in the same
direction. Implications for best user
guidelines provides criteria to "quantify" data
quality for an effusive volcano (maximal
deformation rate for unambiguous signal)
21
Validation of PS Insar results on Le Piton de la
Fournaise (La Réunion) Bulk comparison of data at
the scale of the edifice (subsidence of old lava
flow)
22
Validation of PS Insar results on Le Piton de la
Fournaise (La Réunion) Bulk comparison of data at
the scale of the edifice
Results the overall subsidence is well
constrained by the PS Insar technique Implicatio
ns for best user guidelines additional
constraints are provided by satellite data on the
structural evolution of the volcano at a larger
scale than GPS data
23

• Product validation
• Towards a best practice guidelines
• CONCLUSIONS
• comparison between ground data and products
  identify quality criteria
• New constraints on the evolution of the volcano
  are obtained from Globvolcano products