07 big skyearth_dlr_7_april_2016

High resolution maps of PWV and 3D
reconstruction of atmosphere refractivity
Giovanni Nico
g.nico@ba.iac.cnr.it
Consiglio Nazionale delle Ricerche (CNR)
Istituto per le Applicazioni del Calcolo (IAC)
Bari, Italy

Outline
Interesting features meterological datasets
Active vs. passive remote sensing
Interaction of e.m. with atmosphere
Meteorological databases
SAR interferometry (new high-resolution meteorological data?)
 GPS tomography

What am I interested in?
Linear features are
related to transport of
moisture in atmosphere?
Laminar or turbolent flow ?
Are there frontal zones?
Are isolated anomalies
related to some specific
atmospheric phenomenon?
High resolution image of atmosphere

What am I interested in?
3D images of atmospheric refractivity
Can I identify/retrieve
specific patterns ?
Relevance
I could better study/identify
different types of clouds:
•Stratiform cloud
•Small cumulus clouds
•Cumulonimbus
•Ice clouds
Deeper knowledge about
atmospheric dynamics

Active vs. passive remote sensing
The Sun energy is
 reflected, for visible
wavelengths, or
 absorbed and re-emited,
as it is for thermal infrared.
detect energy when the
naturally occurring
energy is available
Passive Sensors
can only
take place
during the
day (sun)
time
Thermal-IR energy
can be detected by
night or day as long
as the amount of
energy is large
enough to be
recorded.

Active vs. passive remote sensing
They provide their own energy source
for illumination.
Active Sensors
The sensor emits radiation which is directed
toward the target to be investigated. The
radiation reflected from that target is
detected and measured by the sensor.
The advantages of active sensors is that they can
operate at any time of day
Active sensors can be used for examining wavelengths that
are not sufficiently provided by the sun, such as microwaves.

Passive remote sensing
The microwave energy recorded by
a passive sensor can be:
1. Emitted by the atmosphere
2. Reflected from the surface
3. Emitted from the surface
4. Transmitted from the
subsurface
Because the wavelengths are so long, the energy available is
quite small compared to optical wavelengths.
Thus, the fields of view must be large to detect enough
energy to record a signal. Most passive microwave
sensors are therefore characterized by low spatial
resolution.

Active remote sensing
As with passive microwave sensing, a major
advantage of radar is the capability of the radiation
to penetrate through cloud cover and most weather
conditions.
Because radar is an active sensor, it can also be used to
image the surface at any time, day or night
The two main
advantages are:
“All-weather”
“Day and Night”
Because of the fundamentally different way in which an active
radar operates compared to the passive sensors a radar
image is quite different from images acquired in the visible.

Interaction with atmosphere
Fevereiro 2012
DEGGE, João Catalão Fernandes
[jcfernandes@fc.ul.pt]
11
Particles and gases in the atmosphere can affect the
incoming light and radiation.
Two mechanisms:
Scattering
Absorption

Scattering occurs when particles
or large gas molecules present in
the atmosphere interact
with and cause the
electromagnetic radiation to be
redirected from its original path.
How much scattering takes place depends on the wavelength of the
radiation, the abundance of particles or gases, and the distance travelled
by radiation.
Three types of scattering:
Rayleigh Mie
NonSelective

Rayleigh scattering causes shorter wavelengths of energy
to be scattered more than longer wavelengths
Rayleigh scattering is the dominant scattering mechanism
in the upper atmosphere.
Mie scattering occurs when the particles in the atmosphere
have 0about the same size as the radiation wavelength.
Examples are: Dust, pollen, smoke and water vapor
Mie scattering occurs mostly in the lower portions of the
atmosphere where larger particles are more abundant

Nonselective scattering
occurs when the particle size
is greater than the radiation
wavelength.
All wavelengths are scattered
about equally.
Water droplets and large dust particles can cause
this type of scattering.
This type of scattering causes fog and clouds to appear
white to our eyes because blue, green, and red light are
all scattered in approximately equal quantities
(blue+green+red light = white light)

Absorption: this phenomenon
causes molecules in the
atmosphere to absorb energy at
various wavelengths.
Ozone, carbon dioxide, and
water vapor are the three main
atmospheric constituents which
absorb radiation.
Ozone serves to absorb the harmful (to most living things) ultraviolet
radiation from the sun. Without this protective layer in the atmosphere our
skin would burn when exposed to sunlight.

Effect of the atmospheric refraction on microwave signal propagation in
a horizontally stratified atmosphere in which the refractive index
decreases with height: delay in the wave propagation
Royal Observatory of
Belgium. GNSS Research
Group

Because these gases absorb
electromagnetic energy in
very specific regions of the
spectrum, they influence
where (in the spectrum) we
can "look" for remote
sensing purposes.
Those areas of the spectrum which are not severely
influenced by atmospheric absorption and thus, are useful to
remote sensors, are called atmospheric windows.

Meteorological databases
MODIS (Moderate-Resolution Imaging Spectroradiometer)
spatial resolution of 1x1
1 day-time acquisition
36 spectral bands
0.4 – 15.0 m
The MODIS PWV product represents the total atmospheric column water vapor
Satellite Terra (1999)
Satellite Aqua (2002)
MOD05 Terra product
spatial resolution of 5x5
2 day and night acquisitions
MOD07 Terra product

AVHRR (Advaced Very High Resolution Radiometer)
spatial resolution of 1.1x1.1 km at nadir
 ground swath of about 2000 km
 6-8 acquisitions per day (by combining two operational satellites)
Five channels
• C1  0.58 – 0.68 m
• C2  0.73 – 1.1 m
• C3  3.6 – 3.9 m
• C4  10.3 – 11.3 m
• C5  11.5 – 12.5 m
AVHRR images can be used to get an overview of
the general atmospheric situation, the position of
frontal zones and the type of cloud cover
Cold cirrus clouds
Warmer medium and lower level clouds
Combination of channels 1, 2 and 4

Meteosat
spatial resolution of 5x5 km at nadir
 1 acquisition per half an hour
Three channels
• C1  0.5 – 0.9 m
• C2  5.7 – 7.1 m
• C3  10.5 – 12.5 m
Water Vapor

It will carry the Flexible Combined Imager (FCI) with a spatial resolution of 1–2 km
at the sub-satellite point and 16 channels (8 in the thermal band), and an infrared
sounder (IRS) that will be able to provide unprecedented information on
horizontally, vertically, and temporally (four-dimensional; 4-D) resolved water
vapor and temperature structures of the atmosphere.
Humidity and temperature profiles will be generated on the vertical hybrid-sigma
coordinates of the ECMWF forecast system (91 levels)
Meteosat Third Generation (MTG)

Global Atmospheric Models:
ERA-Interim (European Center for Medium-Range Weather Forecasts – ECMWF)
North American Regional Reanalysis (NARR)
Modern Era-Retrospective Analysis for Research and Application (MERRA)
Global and regional reanalysis of atmospheric data provide estimates of
atmospheric variables several time a day at different pressure levels.

ERA-Interim is a atmospheric reanalysis of the ECMWF, following ERA-40. It
provides estimates of temperature, water vapor partial pressure, and
geopotential height along 37 pressure levels, on a global 0.7° grid, at 0:00, 6:00,
12:00 and 18:00 UTC daily, from 1989 to present.
NARR is a regional model that provides estimates of temperature, water vapor
partial pressure, and geopotential height along 29 pressure levels, on a Northern
Hemisphere Lambert Conformal Conic grid centered on the USA, at 0:00, 3:00,
6:00, 9:00, 12:00, 15:00, 18:00 and 21:00 UTC daily, from 1979 to the present.
MERRA is a global reanalysis, providing temperature, water vapor partial
pressure and geopotential height along 42 pressure levels, on a global grid (0.5°
along longidute and 0.75° along latitude), at 0:00, 6:00, 12:00, and 18:00 UTC
daily, from 1979 to present.

The geopotential height is defined to compensate for the decrease of
gravitational attraction with the geometric height z, as
zR
zR
H
e
e



where Re = 6536.766 is the mean Earth radius
 Te
RH
e s
100
The partial pressure e of water vapor is computed from the relative humidity RH
and temperature
Interested people can search for the Clausius-Clapeyron equation giving the
saturation partial water vapor pressure es 

Numerical Weather Models (NWMs)
The Weather Reseach & Forecasting (WRF) model can be used to generate 3D field of
atmosphere temperature, pressure, geopotential, water vapor fraction and liquid water.
Spatial resolution 1kmx1km

NWMs  atmospheric phase delay
  
 h
hydwethydwetatm dhNNRRR
0
6
cos
10


GPS stations overlaid to delay map

Stratified atmosphere? Turbolent atmosphere?

SAR interferometry
WRF

Atmospheric signal in radar interferometric phase images?

Synthetic Aperture Radar (SAR)
h
S
Range (R)
Azimuth(Az)

Azimuth aperture
Azimuth aperture
beamwidth

Radar frequencies
RADAR acronim for RAdio Detection And Ranging
SAR = Synthetic Aperture Radar (Radar ad apertura sintetica)
Band name Frequency (GHz) Wavelength (cm)
P 0.3-1 30 – 100
S 1-2 15 – 30
L 2-4 7.5 – 15
C 4-8 3.8 – 7.5
X 8 - 12.5 2.4 – 3.8
Ku 12.5 – 18 1.7 – 2.4
K 18 – 26.5 1.1 – 1.7
Ka 26.5 - 40 0.8 – 1.1
W > 50 < 0.6
SAR sensor
ALOS-2
RADARSAT-2, SENTINEL
COSMO-SKY-MED, TERRASAR-X
GROUND-BASED SAR

GPS (20180 km)
L-band
Sentinel-1 (693 km)
C-band
ALOS-2 (628 km)
L band
CSK (620 km)
X-band
11000 km
Ionosphere = dispersive medium
Propagation delay in atmosphere

Synthetic Aperture Radar (SAR) Interferometry (InSAR)
Spaceborne radar satellites
• Simultaneously
• Spaced in time
• Hi-res topography
• Motions
• Crustal deformation
• Atmosphere
Multiple observations of surface
Applications

InSAR data acquisition
z
baseline
slant range
The baseline is the distance between
“time coregistered” orbits

SAR interferometry
  DD 





4
2
2
(MASTER) S1(t)
Pixel = A1ei1
D1
D2
Pixel = A2ei2
 12
4
DD 



(SLAVE) S2(t)

InSAR Geometry - height
For =2: Height ambiguity
B=200m
h 2= 43.7 m

Differential SAR interferometry (DInSAR)
After interferogram flattening, the interferometric phase contains
both altitude and motion contributions:
If there
is a
DTM
Phase DTM
contribution
Differential
Interferogram

DInSAR Applications
Terrain displacements (earthquakes, landslides, subsidences, glaciers, etc…

InSAR phase contributions
topo noiseatmoDisplacementk

The atmospheric contribution
Longer wavelength microwave radiation
can penetrate through cloud cover, haze,
dust, as the longer wavelengths are not
susceptible to atmospheric scattering.
Radiation travel path can be affected by
atmospheric humidity, temperature and pressure
Two SAR images not simultaneous, can be affected
differently by the atmosphere with consequences on the
interferometric phase.
BUT

W
f
N
T
P
k
T
P
k
T
P
ksn ed






 4.131.401)( 22
v
3
v
21
 
vacatm
atm dsdsn(s)
    ZTDFdzn
H
  
 0
0atm 1
sin
1
Propagation delay in atmosphere

Precipitable Water Vapour and InSAR
Let us suppose to have an interferogram corrected for topography
     )()()()(, M
atm
S
atm
M
def
S
def
SM tttttt  
where
 tdtdef



4
)( 
 
 tPWV
M
tatm





14
)(
 


cos
1
M is the mapping function








'
2
3
6
2
10
k
T
k
R
m
vOH
The precipitable water vapor is
the total amount of water vapor
In a vertical column of the
atmosphere if it would all
condense

If terrain deformation can be neglected, InSAR can provide maps of PWV temporal
changes
A set of independent measurements of PWV by a network of permanent GPS
stations can be used to calibrate InSAR measurements. Each station measures the
mean PWV in a circular area with a radius of about 3.8 km depending on the cut-off
angle set in the GPS processing. The idea is to use GPS estimates of PWV at the
acquisition times of master and slave SAR images to compute an independent set
of PWV.
  rSM MttPWV  


4
,

Latitude[deg]
Longitude [deg] GPS stations
PWV[mm]
30/08/2009  04/10/2009

Longitude [deg]
Latitude[deg]
GPS stations
PWV[mm]
04/10/2009  08/11/2009

Longitude [deg]
Latitude[deg]
GPS stations
PWV[mm]
12/04/2009  17/05/2009

Longitude [deg]
Latitude[deg]
GPS stations
PWV[mm]
21/06/2009  26/07/2009

Longitude [deg]
Latitude[deg]
GPS stations
PWV[mm]
26/07/2009  30/08/2009

A refinement of the PWV can be obtained by accurately estimating the mean vertical
temperature used to compute the constant .
Usually Tm is obtained by a linear regression with the surface temperature Ts







  '
2
36
10 k
T
k
R
m
v

How can we estimate the absolute PWV?
Relative PWV
InSAR
17/05/09 – 12/04/09
Absolute PWV
WRF model
12/04/09
Absolute PWV
17/05/09

What can we do with InSAR estimates of PWV?
InSAR interpolation to 10km x 10km grid

Assimilation of InSAR PWV maps in NWMs
Spatial distribution of the cumulative
difference in the water vapor mixing ratio
(QVAPOR) in g/kg (positive mean increase
with respect to the analysis field)
QVAPOR vertical profile before
and after the data assimilation

Assimilation of InSAR PWV maps in NWMs
Cumulative difference of
hydrometeors in mm Hydrometeors vertical profile
before and after data
assimilation

Sentinel-1 
High-resolution mapping of PWV on a regional scale
Footprint of all swaths for each
segment (S1 and S2) and the
network of GPS permanent
stations

Sentinel-1 
Phase contributions due to the temporal change of the dry and ionospheric
components of refractivity have been removed!!! Differences with respect to
100kmx100km SAR interferograms (e.g. Envisat)

Sentinel-1 

wetNASWD 
GNSS (Global Navigation Satellite System) tomography








 
 PWV
dr
T
P
k
T
P
kZSWD w 2
v
3
v
2
16
10
K2 = 71.6 k mb-1
K3 = 3.747 105 k2 mb-1
Z-1 = empirical inverse wet
compressibility factor
SWD
observations
unknown
refractivity
3D tomographic grid model































wet
M
wet
MNN
voxM
N voxvoxobsobs
aa
aa
N
N
.
SWD
SWD
1
1111





1

The A matrix is filled using a ray tracing algorithm to measure the sub-path
distance travelled by the total SWD in each voxel
The main drawback of GNSS tomography formulation is its ill-posedeness
resulting from the sub-optimal coverage of the grid model due to he GPS
geometry properties  the A matrix is not invertible due the large amount of
zeros values ini correspondence of the empty voxels mainly in the lowerr part of
the tomographic part of the tomographic model.
To overcome this problem a set of constraints or additional information
concerning the grid model are added
wetN
B
A
0
SWD













Each row of matrix B contains a constraint imposed to the tomographic model
The most common ones are weighted averages using horizontal and vertical
smoothing functions or the inverse distance from the neighbors in each
horizontal layer
Another useful contraint sets refractivity values to zero above a given height of
the troposphere or impose that they follow an atmospheric standard profile.
Can we use meteorological databases to properly fill the A matrix?

     0
TT
NASWDPAPAPANN 
1
00
wet
If a first guess solution N0 of refractivity is available, a solution can be provided
by the damped least square method
P = vector with the weight of each SWD observation
P0 = error covariance matrix of the a priori solution N0

GPS tomography + SAR interferometry
wetINSAR
GPS
INSAR
GPS
.N
B
A
A
0
SWD
SWD





















Units: g/m3
Doy 230 (17/8) 11:30

wetMODIS
GPS
MODIS
GPS
.N
B
A
A
0
SWD
SWD





















GPS tomography + MODIS
Radiosonde: Strong vertical
variability
Tomography: Smoother and
smaller vertical variability

A few references + contacts for data
P. Benevides, G. Nico, J. Catalão, and P. Miranda,
“Bridging InSAR and PS Tomography: A New Differential Geometrical Constraint,”
IEEE Transactions on Geoscience and Remote Sensing, 54(2), 697–702, 2016.
P. Mateus, G. Nico, and J. Catalão,
“Maps of PWV Temporal Changes by SAR Interferometry: A Study on the Properties of Atmosphere’s Temperature Profiles,”
IEEE Geoscience and Remote Sensing Letters, 11(12), 2065–2069, 2014.
P. Mateus, G. Nico, R. Tomé, J. Catalão, and P. Miranda,
“Experimental Study on the Atmospheric Delay Based on GPS, SAR Interferometry, and Numerical Weather Model Data,”
IEEE Transactions on Geoscience and Remote Sensing, 51(1), 6–11, 2013.
P. Mateus, G. Nico, and J. Catalão,
“Can spaceborne SAR interJ. Catalão, ferometry be used to study the temporal evolution of PWV?”
Atmospheric Research, vol. 119, no. 0, pp. 70–80, 2013.
João Catalão, University of Lisbon, jcfernandes@fc.ul.pt
Giovanni Nico, CNR-IAC, g.nico@ba.iac.cnr.it

07 big skyearth_dlr_7_april_2016

More Related Content

What's hot

Viewers also liked

Similar to 07 big skyearth_dlr_7_april_2016

Recently uploaded

07 big skyearth_dlr_7_april_2016