samosa - COASTALT

Transcription

samosa - COASTALT
The SAMOSA project
The SAMOSA project main
achievements and their contribution
to Coastal Altimetry
Cristina Martin-Puig; STARLAB
[email protected]
In collaboration with:
Philippa Berry – DMU
David Cotton - SATOC
Christine Gommenginger – NOC
Lars Stenseng - DTU
SAMOSA
Project funded by ESA under the
STSE programme:
SAMOSA
Motivation and Mission
MOTIVATION
•  CryoSat-2 à SIRAL
−  First SAR altimeter on board
a Satellite
MISSION
•  To quantify the improvements
of SAR altimetry compared to
conventional altimetry for
observations over ocean,
coastal zones and inland
waters
IMAGES COURTESY OF ESA
The SAMOSA project
26 July 2010
2
SAMOSA
Project description and
objectives overview
Full Name: Development of SAR Mode Altimetry studies and
Applications over Ocean, coastal zones and inland waters
Objectives:
1
2
To establish a state of the art review for SAR altimetry
capabilities to observe water surfaces
To reduce SARM data into LRM data and perform a scientific study of the
potential improved capability of SAR data compared to conventional
altimetry over water surfaces
3
To develop a theoretical model for the SAR altimetry mode echoes over
water surfaces
4
5
To define a new re-tracking method for the SAR altimetry mode
To perform a scientific study of the potential capabilities of SAR altimetry
data to characterise coastal zones, estuaries, rivers and lakes
6
To evaluate the SAMOSA re-tracker with ASIRAS data
The SAMOSA project
26 July 2010
3
SAMOSA
TEAM
Expert Advice and Support:
DTU-SPACE
SATOC
JHU - APL
R.K.Raney
ESA - ESTEC
R.Cullen
MSSL
CRYMPS
DMU
NOC
STARLAB
ESA ESRIN
The SAMOSA project
26 July 2010
4
O1: State of the art review
SAMOSA
•  The SAMOSA team defined a SAR Altimetry state of the
art review available at:
http://www.satoc.eu/projects/samosa/
Along-­‐track Ti
me del
ay The SAMOSA project
26 July 2010
5
SAMOSA
O2: RDSAR
By:
In collaboration:
•  The SIRAL modes are mutually exclusive
•  For the quantitative comparison of SARM and LRM data
the only way around is through the reduction of SAR
data such that it emulates conventional altimetry data
•  RDSAR allows for the retrieve of LRM data and SARM
data from a single operating mode
In addition to the previous, it also allow for the simplification of
hardware/mission planning and risks estimation of offset/gaps between
modes
The SAMOSA project
26 July 2010
6
SAMOSA
SIRAL processing modes
Low Resolution Mode LRM
Etc …
PRFLRM = 1970 Hz
Decorrelated echoes
SAR Mode - SARM
time
Burst period
11.7 ms
PRFSAR = 17.8 KHz
Etc…
Correlated echoes
Full
Bit
Rate
time
CryoSat-2 operating modes are mutually exclusive, thus for quantitative
comparison we will need to reduce SARM FBR data to emulate LRM L1b
data (step 1). We will refer hereafter to reduced SARM data as pseudo-LRM.
Once we achieve the pseudo-LRM sequence (power) we need to transform it
to L1b (~20Hz incoh. Integrated data; step 2), and compare performance
with L1b SARM data (step 3).
The SAMOSA project
26 July 2010
7
SAMOSA
Step 1: correlated vs
decorrelated pulses
•  The return echoes in SARM are correlated. This is an
effect from the high PRF of this mode. In LRM
correlation between pulses is not present, nor desired
•  We need to find a solution to combine, or intelligently
select, SARM echoes such that the resulting sequence of
this satisfies decorrelation between pulses
Pseudo-LRM
(at this stage we have a complex
sequence with I,Q components)
The SAMOSA project
26 July 2010
8
SAMOSA
Step 2: Transform pseudoLRM to LRM L1b
•  Apply tracking corrections
−  Rx correction
−  Gain compensation
•  Convert to IFFT
•  Power transformation
•  Incoherent integration
LRM
Block
diagram
SARM
Block
diagram
SARM
FBR data
The SAMOSA project
26 July 2010
9
SAMOSA
Step 3: Quantitative
comparison
The SAMOSA project
26 July 2010
10
SAMOSA
O3: SAR altimeter theoretical
By:
model
•  An semi-analytical waveform model has been derived
for:
• 
Elliptical antenna pattern
• 
Mispointing error in elevation and azimuth
• 
Errors in range cell migration correction
• 
Surface Scattering pattern
• 
Non-linear ocean wave statistics
• 
Spherical Earth surface effects
The SAMOSA project
26 July 2010
11
SAMOSA
• 
• 
!
• 
Given the new shape of SARM echoes
there is a need to define a new theoretical
model to retrack waveforms
The SARM single-look waveform has been
defined (per Burst) as the convolution:
#c & #c &
W (" i , f a ) = PFS (" i , f a ) **SPTR (" i , f a ) * % (Pz % (
$2' $2'
Where τ refers to delay time, or range
window, with respect MSL; fa corresponds
to a Doppler bin; PFS is the average flat
surface response; SPTR the radar system
point target response; and Pz the surface
elevation pdf. The major difference with
respect conventional altimetry is that SPTR
is defined as:
SPTR (" i , f a ) = sinc 2 (Tf a ) sinc 2 (" u s" )
Where T is the long-track boxcar time, τu
the useful pulse length and s the chirp
slope.
Along-track time
HHI1E-F0GA6>1GA61D3GG060.41;EJ1G6A>1#14A1*$"1>
1
!
#$+
#$*
#$)
@A6>-B3C0D1E-F0GA6>
• 
Derivation of the waveform
model
;EJK#
;EJK!
;EJK%
;EJK&
;EJK'
;EJK"
;EJK"$"
;EJK($"
;EJK)$"
;EJK*$"
#$(
#$"
#$'
#$&
#$%
#$!
#1
!!"
!!#
!"
#
"
,-./012345160780941:;<1=>?
!#
!"
!
The SAMOSA project
26 July 2010
12
SAMOSA
O4: SAR Altimeter re-tracker
By:
•  A SAR altimeter re-tracker has been developed based on
the analytical expression derived in the previous
objective
•  Good agreement was found between theoretical &
numerical SAR waveforms
•  A DPM for the Sentinel-3 SRAL retracker has been
delivered.
The SAMOSA project
26 July 2010
13
SAMOSA
LRM & SAR L1B Scenario C1
LRM
Along-track time
SAR
LRM: slow decaying trailing edge &
increased tilt of leading edge for high
SWH
SAR: peaky waveforms, fast decaying
trailing edge, peak broadening at high
SWH
The SAMOSA project
26 July 2010
14
SAMOSA
Multi-looking
…
…
…
…
#$
#!
!"
"
antenna beamwidth
!
Doppler-beam selection
& incoherent integration
over multiple bursts
surface
Multi-looked SAR L1B
Waveform at 18Hz
The SAMOSA project
26 July 2010
15
SAMOSA
How good are O4 and O5 for the
coastal altimetry community?
See related movie StarSARAltCO.mp4
The SAMOSA project
26 July 2010
16
SAMOSA
O5: Coastal & Inland Waters
By:
•  What improvements does SARM altimetry offer for
coastal and inland waters? Two types of scenario have
been modelled:
−  Lakes
−  Coastal Zones
The SAMOSA project
26 July 2010
17
SAMOSA
Coastal zone results
DEM picture courtesy of D.Brockley MSSL
The SAMOSA project
26 July 2010
18
SAMOSA
O6: ASIRAS and SAMOSA
By:
•  Evaluate the SAMOSA re-tracker against real SAR
altimetry data
•  Gain information on the differences between spaceand airborne data
•  Processing scheme for airborne data to allow
comparison with space-borne data
The SAMOSA project
26 July 2010
19
SAMOSA
SIRAL vs. ASIRAS
SIRAL
ASIRAS
Along-track beam width
1.0766°
10°
Cross-track beam width
1.2016°
2.5°
25 W
5W
13.575 GHz
13.5 GHz
350 MHz
1,000 MHz (100-1,000MHz)
49 ms
4 µs (4-80 µs)
Pulse repetition frequency
18.182 kHz
2.5 kHz (2-15 kHz)
Approximate footprint size
15 km x 250 m
120 m x 10 m
Measurement range gate
0.46875 m
0.08783 m (0.109787 m)
Average platform velocity
7,389 m/s
65 m/s
717 km
2.7 km (0.2-10 km)
Transmitted power
Center frequency
Transmitted bandwidth
Transmitted pulse length
Average range to surface
The SAMOSA project
26 July 2010
20
Along-track geometry
(Doppler selection)
0.8
0.6
0.4
0.2
0
15
5
12
B
50
75 in
b
0
10 am
e
20
25
0
60
70
80
90
100
1
110
R
e
ang
bin
40
01
150 1.0
13
Normalized amplitude
Normalized power
SAMOSA
0.8
0.6
0.4
0.2
0.0
60
70
80
90
100
110
120
130
140
Range bin
The SAMOSA project
26 July 2010
21
SAMOSA
• 
• 
• 
• 
• 
SAMOSA main progress
Investigations to reduce SAR mode to LRM
Definition of SAR Altimetry waveform model
Development of SAR retracker
Investigations with simulated data
Analysis of performance of SAR retracker with ASIRAS
data
The SAMOSA project
26 July 2010
22
Contact details:
David Cotton
SATOC
SAMOSA PROJECT COORDINATOR
[email protected]
SAMOSA