recovering more and better data from altimetry in the coastal...
TRANSCRIPT
Recovering more and better data from altimetry in the coastal zone:
a community effort
Paolo Cipollini, National Oceanography Centre, UK, [email protected]
& Coastal Altimetry WS OrgComm: Jérôme Benveniste, Hans Bonekamp,
Laury Miller, Nicolas Picot, P. Ted Strub, Doug Vandemark. Stefano Vignudelli
Pilot ARCOM workshop Lisbon, 1/2 Sept 2015
The Coastal Altimetry Community
• Coastal Altimetry Workshops (CAW) since 2008 – large community: 100+, COASTALT-SWT mailing list
– like a big splinter of OSTST
• A Community Science Review – improving techniques
– refreshing diversity of applications
– recommendations (internal / external)
CAW-5 San Diego, 2011
One of our first recommendations …
• Coastal Altimetry is inherently interdisciplinary and a pretty legitimate
(and very useful) component of coastal observing systems
• Need constant interaction with hydrographers, modellers, in situ
scientists on the application side, engineers on the technical side
RECOMMENDATION: we need to reach out to more (potential) users
CAW-2 Pisa, 2008
…and one of the latest!
• Coastal Altimetry is very relevant to GODAE
OceanView COSS-TT (Coastal and Shelf Seas
Task Team) so…
• …we need COSS-TT/CAW joint meeting to
promote uptake of Coastal Altimetry and expose
issues for SWOT etc.
CAW-8 Konstanz, 2014
Coastal altimetry continues to improve
• Technical Advances
• Continuing improvement in corrections
• Many datasets are now available…
• … and applications are numerous
Technical advances
• Knowledge on how to
handle SAR altimetry is
rapidly expanding
– stacking, improved
waveform models
• CryoSat-2 SARM working
very well and ideal
precursor to Sentinel-3
• AltiKa also extremely good
in coastal zone
• Envisat Individual Echoes
great testbed for new and
future missions
SAR Altimetry Training Course, 21-22 Oct 2014
Technical advances
• Knowledge on how to
handle SAR altimetry is
rapidly expanding
– stacking, improved
waveform models
• CryoSat-2 SARM working
very well and ideal
precursor to Sentinel-3
• AltiKa also extremely good
in coastal zone
• Envisat Individual Echoes
great testbed for new and
future missions
0 5 10 15 20 25 300
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
0.2
Distance from closest coastline (km)
abs(2
0H
z d
iffe
ren
ce
s)
(m
)
FBR ESRIN SAM R5; Jul12 & Jan13; abs(diff) of 20−Hz TWLE with misfit<4
samples
75th percentile
median
25th percentile
~4.5cm @ 2km
0 5 10 15 20 25 300
0.5
1
Distance from closest coastline (km)fraction
of va
lid p
oin
ts
FBR ESRIN SAM R5; Jul12 & Jan13; fraction of valid points passing misfit<4
~4.0cm @ 5km with ~80% valid
CryoSat-2
UK coasts
CP4O
Team
Technical advances
• Knowledge on how to
handle SAR altimetry is
rapidly expanding
– stacking, improved
waveform models
• CryoSat-2 SARM working
very well and ideal
precursor to Sentinel-3
• AltiKa also extremely good
in coastal zone
• Envisat Individual Echoes
great testbed for new and
future missions
10 km
20 km
AltiKa / Jason-2 West Mediterranean Sea
J2 AltiKa
G. Valladeau
PEACHI
Technical advances
• Knowledge on how to
handle SAR altimetry is
rapidly expanding
– stacking, improved
waveform models
• CryoSat-2 SARM working
very well and ideal
precursor to Sentinel-3
• AltiKa also extremely good
in coastal zone
• Envisat Individual Echoes
great testbed for new and
future missions 1-second of of 1800 Hz echoes in vicinity of
Rio Tigre, Peru; Amplitude in dB re noise
R. Abileah
Technical advances
• Avoiding land effects with
range gate selection
• Improvement in
retrackers/windowing
• Work on inland waters remains
very relevant A. Egido
Technical advances
• Avoiding land effects with
range gate selection
• Improvement in retrackers
– windowing/subwaveforms
• Work on inland waters remains
very relevant and promotes
better understanding
Stack of echos (after migration)
Multilooked echo
Lo
ok N
um
be
r N
orm
aliz
ed
Pow
er
SAR Mode
12 …33 ………………. 115 12 …33 ………..83 12 …33 … 63 12 . 33 . 48 range gates
Window Truncation 12-115 12-83 12-63 12-48
Radius of the WF footprint 7488 m 5848 m 4530 m 3203 m
P. Thibaut
NOC’s ALES retracker
Adaptive Leading-Edge Subwaveform (ALES) retracker
Passaro et al., Rem Sens Env., 2014
radar waveform
Pass 1: identification of leading edge and
initial estimate of SWH
ALES algorithm for width of sub-waveform
window (from Monte Carlo
Simulations)
Pass-2: subwaveform
retracking (Brown Model)
sub-waveform hs(1)
range, σ0, hs
The sub-waveform approach keeps the measurement focused
on the pulse-limited footprint of the instrument prevents any
non-homogeneities in the far-view field from degrading its
accuracy.
very good for coastal zone and fine scales
Examples of Jason-2 retracking by ALES
• Validated for SSH for Envisat, J-1, J-2, AltiKa
• Validated for SWH for Envisat, J-1/2 (Passaro et al 2015)
• J-2 data available now from PODAAC, Envisat coming
see Poster (Passaro et al.)
Open ocean
SWH = 0.75 m
Coastal ocean
SWH = 1.65 m
Open ocean
SWH = 9.5 m
Jason-2 p0044 cycle 252 - 7 May ‘15
44 44.1 44.2 44.3 44.4 44.541
41.5
42
42.5
43
43.5
44
44.5
45
45.5
46
latitude
US
SH
(m
)Uncorrected SSH (orbit minus range)
SGDR
SGDR MLE3
ALES
LAND
Coast
Ships?
ITALY
Ligurian
sea
Continuing improvement in corrections
• DComb Wet Tropo
• High-res MWR on the horizon
• MSS – and tides!!
– in coastal regions data editing for
MSS determination is critical
– SarIN! Example
• Significant differences shown
between the GDR-Global DAC
prediction and local models
within several regions.
– do not apply the global DAC to the
altimeter SSHA, or do so after
evaluation against wind-forced HF
signals along your coastline.
J. Fernandes
18.7 GHz QV
90 GHz
S. Reising
Continuing improvement in corrections
• DComb Wet Tropo
• High-res MWR on the horizon
• MSS – and tides!!
– in coastal regions data editing for
MSS determination is critical
– SARIn useful around fjords
• Significant differences shown
between the GDR-Global DAC
prediction and local models
within several regions.
– do not apply the global DAC to the
altimeter SSHA, or do so after
evaluation against wind-forced HF
signals along your coastline.
None of the global tide
models covers here !
O. Andersen
Continuing improvement in corrections
• DComb Wet Tropo
• High-res MWR on the horizon
• MSS – and tides!!
– in coastal regions data editing for
MSS determination is critical
– SARIn useful around fjords
• Significant differences shown
between the GDR-Global DAC
prediction and local models
within several regions.
– do not apply the global DAC to the
altimeter SSHA, or do so after
evaluation against wind-forced HF
signals along your coastline.
L Ruiz-Etcheverry
Variance HAMSOM
Variance Mog2D
2200 cm2
800 cm
Datasets for coastal altimetry
ID Produced
by
Missions Product
level
Posting
rate
Coverage Download
from
AVISO CLS
CNES
e1,tx,e2, en, j1,
j2, c2
(LRM/PRLM),
sa
L2, L3
also L4
1 Hz Global +
european regions
AVISO+
CMEMS CLS
CNES
e1,tx,e2, en, j1,
j2, c2
(LRM/PRLM),
sa
L3
L3 for assim
1 Hz Global +
european regions
marine.cope
rnicus.eu
PISTACH CLS
CNES
j2 L2 20 Hz Global AVISO+
PEACHI CLS
CNES
sa L2 40 Hz Global AVISO+
XTRACK LEGOS-
CTOH
tx, j1, j2, gfo, en L2, L3 1 Hz
20Hz
(test)
23 regions CTOH
AVISO+
RADS EUMETSAT,
NOAA,
TUDelft
gs, e1, tx, pn,
e2, gfo, j1, n1,
j2, c2, sa
1 Hz Global TUDelft
ALES NOC j2 20 Hz Global, <50 km
from coast
PODAAC
SARvatore ESA-ESRIN c2 (SAR only) 20 Hz SAR mode
regions
ESA GPOD
Cal/Val and Applications
• 30+ contributions at CAW-8
• from broad shelf to the coast
• strong complementarity with models
Mean surface geostrophic
current amplitude (in m/s)
Observations (MDT Rio 2014)
Curvilinear grid
Resolution: 0.6 to 6 km
Symphonie model (regional config in 2014)
F. Birol
Application example 1: coastal dynamics
Example at Ibiza, W Med:
SARAL/AltiKa derived
velocities reveal coherent
mesoscale features with
general good agreement with
HF radar fields
AltiKa
HF radar A. Pascual et al
Application example 2 : storm surges
Surge due to Hurricane Katrina,
29 August 2005
Hurricane Sandy storm surge flooding the World Trade Center construction site
Storm Surges
Predicted
Tides
Measured
Tides
Surge:
Measured - Predicted
Lillibridge et al -
Hurricane Sandy storm surge
HY-2A only altimeter to capture
surge < 12 hr duration
Montauk
HY-2A
Application example 2 : storm surges
Application example 2: storm surges
Present constellation insufficient to reliably capturing all surges, but the time/space information content can be extended using model dynamics (example: statistically altimetry/tide gauge blending by DMI)
ESA due eSurge / eSurge-Venice projects see poster by Bajo et al
Cyclone Xaver seen by Geodetic observations
L. Fenoglio et al
C2 SAR observations of Xaver storm - Dec2013
Friday 6 Dec 23:05
Data were available on eSurge Server by Saturday lunchtime
DMI model
Madsen et al eSurge: ESA DUE
(Data User Element)
Project for 2011-2015
www.storm-surge.info 24
Take home message – “the challenges”
• The quality of coastal altimetry – precision and
resolution – continues to improve, challenging
our understanding of the ocean at short scales
• how do we manage the extra resolution?
• can we use coastal altimetry data to prepare
the community to ingest high-resolution data
from SWOT?
• what applications are possible in Near Real
Time?
18-19 October 2015, Reston (Virginia), United States
www.coastalaltimetry.org
abstracts due Sept 1st, 2015
Extra material
18/10/2011 NOV-3900-SL-11654
© NOVELTIS 2011-This document is the property of Noveltis, no part of it shall be reproduced or transmitted without the express prior written authorisation of Noveltis
28
Comparison of the current anomalies
PISTACH RED3 DUACS
mg/m3
Mean geostrophic
current – HYCOM
model – 30/08/2009
Provided by
M. Le Hénaff,
RSMAS
Cancet et al
CAW-8 Recommendations
• [CAW8-REC1] Unfiltered along-track high-resolution
SSH should be open to public in both delayed and NRT
products
– Best filtering scales may differ regionally and seasonally
• [CAW8-REC2] A Coastal MSS and MDT recomputed
with high-resolution SSH would be a useful thing
– For understanding dynamics
– For cal/val with non-repeating tracks (although less accurate
than in repeating orbits)
• [CAW8-REC3] support R&D in development of more
accurate tidal models in the coast (including merging
regional models with global ones)
To Agencies and data providers
• Data assimilation (via variational methods)
integrates remotely sensed observations and
dynamical constraints to improve ocean state
estimation
• Using all data (SSH, SST, HF-radar) gives
greatest analysis skill; assimilating only one data
type degrades the analysis of unconstrained
variables.
• Adjoint-based methods quantify forecast
uncertainty without computing large ensembles
• Multi-scale hierarchy of models, using altimetry
processed to differing levels, acknowledges the
range of model and data covariance scales.
Coastal Altimetry and Modelling
Coastal Altimetry and Modelling
• Model-based analysis (assimilation; statistical
covariance analysis) are promising approaches
to infer subsurface circulation from surface-only
observations
• Coastal oceanographers need guidance on “best
practise” to match corrections appropriate to
dynamical processes.
• To utilize high sample rate SSH model
development required ~ multi-scale nesting and
data assimilation is recommended.
• Coastal modelers/assimilators need a test-
bed/comparison framework (complementary to
CAW process) to exchange developments in
Relevance to SWOT
US West Coast
80
0 k
m
600 km
*Proposed Mission – Pre-decisional – for planning & discussion purposes only P. Callahan et al