POLinSAR 2007, Frascati Studies of Ku- and X-Band Radar Interactions with the Snow Cover in Preparation of the CoreH 2 O Mission CoreH 2 O - COld REgion Hydrology High resolution Observatory Helmut Rott 1 , Don Cline 2 , Keith Morrison 3 , Thomas Nagler 4 , Jouni Pulliainen 5 , Helge Rebhan 6 , Jiancheng Shi 7 , Simon Yueh 8 1 University of Innsbruck, Austria 2 NOAA-NOHRSC, Chanhassan, USA 3 Cranfield University, UK 4 Environmental Earth Observation IT, Innsbruck, Austria 5 Finnish Meteorological Institute, Helsinki, Finland 6 ESA-ESTEC, Noordwijk, NL 7 University of Calfornia, Santa Barbara, USA 8 JPL-Caltech, Pasadena, USA
23
Embed
Studies of Ku- and X-Band Radar Interactions with the Snow ...earth.esa.int/workshops/polinsar2007/presentations/278_rott.pdfPOLinSAR 2007, Frascati Studies of Ku- and X-Band Radar
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
POLinSAR 2007, Frascati
Studies of Ku- and X-Band Radar Interactions with the Snow Cover in Preparation of the CoreH2O Mission
CoreH2O - COld REgion Hydrology High resolution Observatory
Helmut Rott1, Don Cline2, Keith Morrison3, Thomas Nagler4,Jouni Pulliainen5, Helge Rebhan6, Jiancheng Shi7, Simon Yueh8
1 University of Innsbruck, Austria2 NOAA-NOHRSC, Chanhassan, USA3 Cranfield University, UK4 Environmental Earth Observation IT, Innsbruck, Austria5 Finnish Meteorological Institute, Helsinki, Finland6 ESA-ESTEC, Noordwijk, NL7 University of Calfornia, Santa Barbara, USA8 JPL-Caltech, Pasadena, USA
Land cover, Water cycle, Albedo,Surface temperature, Wet lands,Eco-systems, Carbon cycle
OCEAN
Temperature, Salinity, Circulation,Sea level
Frozen GroundPermafrost
Heat exchangeGas exchangeCarbon cycle
Snow Cover
AccumulationMeltRunoffWater supplyExchange of heat, moisture momentumAvalanches
Ice SheetsIce Shelves
Mass balanceAccumulationExchange of heat, momentum and water
River andLake Ice
ExtentThicknessEnergy exchangeRunoff routing
Sea Ice
ThicknessGrowthSnow accumulat.MeltExchange of heat, momentum
GlaciersIce Caps
Mass balanceAtmospheric forcingRunoffWater supply
POLinSAR 2007, Frascati
Key Scientific Questions: The Role of Snow and Glaciers for Climate Research and Hydrology
The role of snow in the global climate system Improved understanding and modelling of snow cover - climate feedbacks are needed for adequate representation of snow in climate models.
The role of snow in hydrological processesIn high and mid latitudes snow cover is a key parameter of the water and energy cycle of land surfaces. The interactions between atmospheric circulation, land surfaces and hydrological processes are an open problem for climate modelling.
The role of snow for water management and flood control:
Improved models with distributed input data are needed to improve flood forecasting and water resources management.
Glacier mass balance and runoff modelling
Snow extent and mass (Snow Water Equivalent, SWE) are key input parameters for those models.Presently spatially distributed data of SWE are not measured directly, but inferred from few point measurements, resulting in large errors
POLinSAR 2007, Frascati
Variable Spatial Scale Repeat Interval Seasonal Snow Cover Snow extent 100 – 500 m 3-15 d Extent of melting snow 100 – 500 m 3 dSnow depth 200 – 500 m 3-15 dWater equivalent 200 – 500 m 15 dGlaciersFacies type 100 m 15 dWinter snow accumulation 200 – 500 m 15 dFrontal position, lakes 50 m 15 dFreshwater Ice Area 25 - 50 m 3 dSea IceSnow Depth 200 - 500 m 3 -15dIce Type 50 - 100 m 3 dIce Motion 0.5 to 1 km 3 dIce Melt 50 – 100 m 3 dSurface WaterExtent of open water areas 50 – 100 m 3-15 d
CoReH2O – Key Observation Requirements
POLinSAR 2007, Frascati
25 m x 8 m25 m x 10 mScanSAR resolution
<-25dB<-25dBNESZ
10 MHz10 MHzTransmit chirp BW
3 kW4 kWPeak transmitter power
100 km100 kmSwath width
46Nr. of ScanSAR beams
24 km16 kmAnt. beam footprint elev.
2.2 m x 4.4 m (parabolic)1.2 m x 3.3 m (3 reflectors)Antenna size
CoReH2O – Ku-band SAR Satellite Configuration - V01
Antenna: 3 parabolic reflectors. Overall length: 3.3 m (az.) x 1.2 m (el.)
Total satellite mass ≤ 850 kg
activefeeds
beam
dire
ctio
n
ϑ ϑ
x
f
Beam scanning by feed defocusing
ObservationDirection
Swath
Width
Satellite Ground Track
Multiple subswa
to achieve wide sw
POLinSAR 2007, Frascati
Antenna BeamBoresight
S/C Flight Direction
Radiator Panel
Solar Array
CoReH2O – X-band SAR Configuration - V01
X-band SAR Satellite Configuration Switch matrix for X-band SAR feed array
Technical Challenge: dual frequency SAR with single reflector antenna
POLinSAR 2007, Frascati
Combination of zeroth-order approximation of RT theory and field approach, with wave approach to define scattering in a small volume unit (DDA approximation). Theory by J. Pulliainen and M. Hallikainen, HUT. Software implemented by VEXCEL UK
Dense Medium Radiative Transfer, DMRT. Applied for numerical simulation of σ° for a wide range of snow conditions. Algorithm developed for SWE estimation with dual-frequency (9.6 & 17 GHz) & polarization (VV and HV) SAR by scattering decomposition (J.C. Shi, UCSB)
• Depolarization factor for dry snow proportional to the scattering contribution in co-polarization signals; used to decompose surface and volume scattering
• Frequency ratio of volume scattering components at X- and Ku-bands used for estimating snow optical thickness τ. Single scattering albedo estimated from τand σv.
• With the estimations of scattering albedo and optical thickness at two frequencies, SWE can be retrieved.
Models for Snow Scattering and SWE Retrieval
POLinSAR 2007, Frascati
Backscattering from a Multilayer Snowpack
Total backscattering0000
)cos()cos(
groundsnowsnowpackairsnowsnowairt
isnowpacksnowairtotal ATT −−−− ++= σ
θθ
σσσ
∑ ∏=
−
=⎟⎟⎠
⎞⎜⎜⎝
⎛=
n
i
i
jjitotsnowpack t
1
1
0
20,
0 σσ
∏=
=n
jjsnowpack tA
0
2
Multilayer volume Two-way
attenuation
h
d
θi
A single layer is split up into K volume units
DDA solution of scattered electric field by solving the field inside each cubical volume element (approximated as a dipole with volume of the cube) and by summing the fields originating from each scatterer
( )[ ]ttet
e
tVvolsnow h θκ
κ
θσσ sec2exp1
2
cos11,
1,
1,1, −−=° −
POLinSAR 2007, Frascati
Volume Scatter Modelling
Model Approach:• Applies a general modeling tool to describe volume backscatter and extinction• Combines the deterministic Discrete Dipole Approximation (DDA) with a common
vector Radiative Transfer (RT) approach
• An ensemble of infinitesimal snow samples representing certain ice volume filling factors and grain sizes are generated, i.e. random realizations of snow structure
• DDA solution on scattering amplitude is calculated for each 3-D grid of ice cubes (including multiple scattering, all polarisations)
• Volume scattering and extinction coefficients for different snow layers (with specific physical characteristics) are calculated from scattering amplitudes
• RT solution on backscattering coefficient at different polarizations is calculated applying the inherent properties determined by DDA calculation
POLinSAR 2007, Frascati
Monte Carlo Simulation of Snow Packs
• Collection of snow volume units with random localization and orientation of snow grains (ice crystals)
• Constant average properties in a single collection: snow density, mean snow grain size and snow grain size standard deviation
Snow grain r=0.4 mm
POLinSAR 2007, Frascati
Example of σ° at 17 GHz with DDA Approximation
30 40 50 60 70 80-50
-40
-30
-20
-10
0
incident angle (deg)
back
scat
terin
g (d
B)
VV-backscattering component for air-snow-ground-model with DDA in 17 GHz
totalvolumeair-snowsnow-ground
30 40 50 60 70 80-50
-40
-30
-20
-10
0
incident angle (deg)
back
scat
terin
g (d
B)
HH-backscattering component for air-snow-ground-model with DDA in 17 GHz
totalvolumeair-snowsnow-ground
30 40 50 60 70 80-25.6
-25.4
-25.2
-25
-24.8
-24.6
-24.4
-24.2
incident angle (deg)
back
scat
terin
g (d
B)
HV-backscattering component for air-snow-ground-model with DDA in 17 GHz
totalvolume
30 40 50 60 70 80-25.6
-25.4
-25.2
-25
-24.8
-24.6
-24.4
-24.2
incident angle (deg)
back
scat
terin
g (d
B)
VH-backscattering component for air-snow-ground-model with DDA in 17 GHz
totalvolume
Snow depth 2 m Grain r = 0.4 mm Cubic grain shape
0 0.2 0.4 0.6 0.8 1
-50
-40
-30
-20
-10
0
snow depth [m]
volu
me
back
scat
terin
g [d
B]
VVVH
Needles randomly alignedl = 0.8 mm r = 0.2 mmθ = 30°
POLinSAR 2007, Frascati
Minimum Maximum Interval
Exponential function
0.25 cm 3.0 cm 0.25 cm2.5 cm 25 cm 2.5 cm
Parameters
rms. height
Co-lengthCo-function
Ice FractionParticle Radius 0.2 mm 2.0 mm 0.2 mm
15 % 50 % 5 %
Frequency 9.5 GHz 17 GHz
Snow Temp. -15°C -3°C 3°C
Input parameters for Database Simulation
Simulation model Characteristics – second-order RT model1. Snow – Dense Medium Model (DMRT)2. Surface – Advanced Integral Equation Model (AIEM)3. Snow-surface interaction – Bistatic AIEM & DMRT
Surface scattering from air-snow interface and scattering term from the snow-ground interaction
Snow volume scattering term
Surface scattering term)f(spq0σ
Surface scattering from snow-ground interface
POLinSAR 2007, Frascati
Scattering Contribution Characteristics at 9.6 and 17 GHz 35° Incidence
Each Scattering Contribution in % for 3 Components
17-VV9.6-VV17-VH
Volume
Interaction
Surface9.6-HV
POLinSAR 2007, Frascati
Decomposing Scattering Components
)/BA(/)f('F tvvtvh
tvvsvvs σσ⋅+−≈σσ= 1
tvv
tvh
tvv
vvvs bafF σσσσ //)( +≈=
tvv
tvh
tvv
svvv
vvv BA σσσσσ /*/)( +⇔+
Depolarization factor is proportional to scattering contributions of either the direct volume scattering or the sum of volume and surface-volume interaction components
9.6 GHz 17 GHz
tvv/v
vv σσ
tvv/t
vh σσ tvv/t
vh σσ
tvv/)sv
vvvvv( σσ+σ
tvv/s
vv σσ
POLinSAR 2007, Frascati
Estimation of Albedo for Correction of Grain Size Effect
)(),.( 1769 ωω
)().(
1769
ττUsing the estimated
Optical thickness ratio is highly correlated to albedo
)().( 1769 ωω +
)().(
1769
ττ
)().(
1769
ωω
Grain Size Effect Correction
)f())f((sd)f(a τ⋅ω−=⋅κ 1
)(),.(),(),.( 17691769 ττωωUsing estimated
To derive the absorption part of the optical thickness
POLinSAR 2007, Frascati
Correction of Snow Density and Temperature Effects for SWE Retrieval
Correction on effects of snow density and temperature
1. Absorption coefficient is linearly related to snow density at a given temperature
2. Ratio of absorption coefficients from two frequency depends only on temperature
density
temperature
SWE estimation
)(/).(e))(/).(log(dd).(c)d).,((Logba)SWElog(
aaaa
sasa17691769
696917κκκκ
κκ⋅+⋅+
⋅⋅+⋅⋅+=
SWE estimation using τa
with simulated data base at θ = 35°
κa
POLinSAR 2007, Frascati
Signature and SWE Retrieval Studies with low Resolution Scatterometer Data
Snow accumulation at NASA-SW station in Greenland and QSCAT retrieval (S. Nghiem)
SWE retrieved from QuikSCAT Ku-band data using a radiative transfer model function comparedvto observations (Cline, 2004).
POLinSAR 2007, Frascati
In Situ (Snow Depth, Snow Pits)
CIR Aerial Photography/Orthoimagery
LiDAR DEM (Airborne Snow Depth Mapping)
AMSR-E (Spaceborne PM Radiometer)
QuikSCAT (Spaceborne Ku Scatterometer)
TerraSAR-X (Spaceborne X SAR)
POLSCAT (Airborne Ku Polarimetric Radar)
Alaska 2007-2008
Colorado 2006-2007Data Collection
Field Experiments:CLPX-II Cold Land Processes Experiment
Don Cline
Simon Yueh
POLinSAR 2007, Frascati
GB-SAR Snow Campaign10-18 GHz• Univ. Cranfield• ENVEO IT Innsbruck• ESA-ESTEC