constructing high- resolution, absolute maps of...

KIT – University of the State of Baden-Wuerttemberg and National Research Center of the Helmholtz Association

INSTITUTE OF PHOTOGRAMMETRY AND REMOTE SENSING

www.kit.edu

Fadwa Alshawaf, Stefan Hinz, Michael Mayer, Franz J. Meyer [email protected]

Constructing high-resolution, absolute maps of atmospheric water vapor by combining InSAR and GNSS observations

2 Water vapor mapping by combining InSAR & GNSS March 24, 2015 F. Alshawaf, Institute of Photogrammetry and Remote Sensing

Atmospheric water vapor

(In)SAR: (Interferometric) Synthetic Aperture Radar

GNSS: Global Navigation Satellite Systems

Weather and Climate:

Most active greenhouse gas Key element in the hydrological

cycle





Highly variable in time/space Available data are limited in

temporal/spatial resolutions



cycle





Highly variable in time/space Available data are limited in

temporal/spatial resolutions

Noise

Geodesy and Remote Sensing:

Source of error Methods for error mitigation Empirical models Calibration using external data Time series analysis

Sig

nal



cycle


GNSS

Data Combination

2D fields of partial wet delay

Pointweise total wet delay

PS InSAR

Objectives

2D fields of total precipitable

water vapor


(In)SAR (2003-2008) GNSS (since 2002) Meteorology MERIS (Reference)

Study area and data sets

Upper Rhine


Atmospheric wet delay from PS InSAR

+++= atmospherentdisplacemetopographyji φφφφ ,

noisereforbit φφφ ++

2004 2005 2006 2007 2008 2009-800

-600

-400

-200

0

200

400

600

800

Temporal Baseline [years]

Perp

endi

cula

r Bas

elin

e [m

]

p

Track 294

master

Poster „Constructing high-resolution maps of atmospheric water vapor using InSAR“



+++= atmospherentdisplacemetopographyji φφφφ ,

noisereforbit φφφ ++

2004 2005 2006 2007 2008 2009-800

-600

-400

-200

0

200

400

600

800

Temporal Baseline [years]

Perp

endi

cula

r Bas

elin

e [m

]

p

Track 294

master

Persistent Scatterer InSAR (PSI)

Interferograms

Series of temporal and spatial filtering

Wet day difference maps


Least squares inversion Maps of partial wet delay → water vapor Very good agreement with MERIS observations

PSI: Partial water vapor MERIS: Partial water vapor

MEAN [mm] -0.01

STD [mm] 0.86

RMS [mm] 0.87

Corr. Coeff. 0.88


April 23, 2007




MEAN [mm] -0.01

STD [mm] 0.86

RMS [mm] 0.87

Corr. Coeff. 0.88


+ Maps of high spatial resolution − Partial measurements → no absolute values

April 23, 2007


How to reconstruct the total water vapor content? It is not only one offset → A value has to be determined for each point



MEAN [mm] -0.01

STD [mm] 0.86

RMS [mm] 0.87

Corr. Coeff. 0.88


April 23, 2007


Data combination

Total water vapor

Elevation-dependent water vapor

Long wavelength water vapor

Partial water vapor


Data combination: Approach

GNSS zenith total path delay

GNSS absolute water vapor content

Meteorological data P, T, RH

P: Pressure T: Temperature RH: Relative Humidity

Model elevation-dependent signal

min)exp()exp()( LzCzzCzL ∆+α−α+α−=∆

Elevation-dependent water vapor

Model parameters

GNSS site altitude

)(zL∆

min,, LC ∆αz



GNSS zenith total path delay



Model a 2D linear trend

+ -


Model elevation-dependent signal


Data combination: Application to the data

0 200 400 600 80024

26

28

30

32

34

36

Altitude [m]

Wat

er v

apor

[mm

]

observedfitting

[mm] [km-1] [mm]

8.0375 4.1342 25.0434

C α minL∆


Data combination: Application to the data

Compute a value at each persistent scatterer → digital elevation model is required

Iterative solution

0 200 400 600 80024

26

28

30

32

34

36

Altitude [m]

Wat

er v

apor

[mm

]

observedfitting

[mm] [km-1] [mm]

8.0375 4.1342 25.0434

C α minL∆


GNSS total path delay



PSI maps of partial water vapor content

Maps of absolute water vapor content

+ + Non-turbulent water

vapor maps

+ +

Estimate 2D linear trend parameters

+ -


Estimate elevation-dependent parameters



Water vapor: PSI+GNSS Water vapor: MERIS

Difference

Data combination: Results

MEAN [mm] -0.43

STD [mm] 0.84

RMS [mm] 0.91

Corr. Coeff. 0.92

April 23, 2007 April 23, 2007


Water vapor: PSI+GNSS Water vapor: MERIS

Data combination: Results

April 23, 2007

Day CC [%] RMS [mm] MEAN [mm] STD [mm]

June 27, 2005 75 1.00 0.07 1.00

September 5, 2005 87 0.88 0.15 0.86

July 17, 2006 80 0.76 -0.03 0.75

April 23, 2007 92 0.91 -0.43 0.84

Alshawaf, F., S. Hinz, M. Mayer, F.J. Meyer (2015), Constructing accurate maps of atmospheric water vapor by combining interferometric synthetic aperture radar and GNSS observations, Journal of Geophysical Research: Atmospheres, 120 (4), pp. 1391–1403.


GNSS

Data Combination 2D fields of total precipitable

water vapor

2D fields of partial wet delay

Pointweise

total wet delay

PS InSAR

Correcting for atmospheric errors

Correcting for atmospheric errors

WRF 3D fields of total

precipitable water vapor

Data Fusion

Conclusions and Outlook

WRF: Weather Research and Forecasting


Conclusions and Outlook

Thank you very much for your attention

constructing high- resolution, absolute maps of...

Documents