eralized scheme to retrieve wet path d ater vapor radiometer measurements app European geodetic VLBI network Jung-ho Cho 1,2 , Axel Nothnagel 2 , Alan Roy 3 , and Ruediger Haas 4 1 Korea Astronomy and Space Science Institute 2 Geodetic Institute of the University of Bonn 3 Max Plank Institute for Radio Astronomy 4 Onsala Space Observatory of Chalmers Technical University 4 th IVS General Meeting, Concepcion, Chile, Jan. 9~11, 2006 To check the possibility of improvement in VLBI positioning introducing WVR WPD instead of estimation WVR: Water Vapor Radiometers WPD: Wet Path Delay
27
Embed
A generalized scheme to retrieve wet path delays from water vapor radiometer measurements applied to European geodetic VLBI network Jung-ho Cho 1,2, Axel.
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
A generalized scheme to retrieve wet path delays from water vapor radiometer measurements applied to
European geodetic VLBI network
Jung-ho Cho1,2, Axel Nothnagel2, Alan Roy3, and Ruediger Haas4
1Korea Astronomy and Space Science Institute
2Geodetic Institute of the University of Bonn
3Max Plank Institute for Radio Astronomy4Onsala Space Observatory of Chalmers Technical University
4th IVS General Meeting, Concepcion, Chile, Jan. 9~11, 2006
Purpose: To check the possibility of improvement in VLBI positioning results introducing WVR WPD instead of estimation
WVR: Water Vapor Radiometers WPD: Wet Path Delay
Contents
2/15 4th IVS General Meeting, Concepcion, Chile, Jan. 9~11, 2006
Tropospheric delay in VLBI
Water vapor monitoring instruments
WVR network & WVR inter-comparison campaign
WPD retrieval scheme of four European VLBI sites
Results
Concluding remarks
L = S n ds – G L = S (n – 1) ds + S - G
Elgered (1993)
Tropospheric delay in VLBI
John W. Birks
Water vapor contents in troposphere are highly variable even in a short period as well as long period. It causes an unpredictable tropospheric path delay of radio signal propagation. Although its size of 10~30cm is relatively small, water vapor is one of the biggest pending problem in the space geodesy techniques.
Especially in VLBI, global scale network is normally used. That means the tropospheric condition of each site is different enough. But it is not enough to get stable 1mm-precision with conventional estimation. We need to find a proper instrument that can be used as monitoring the water vapor in troposphere directly.
Daily variance of water vapor contents in troposphere
3/15 4th IVS General Meeting, Concepcion, Chile, Jan. 9~11, 2006
+ Temporal resolution+ The most direct way+ Continuous observation
+ Global observation+ Good resolution for ocean
+ Temporal resolution+ Continuous observation + Free from raining
+ Possible to profiling
+ Vertical distribution
+ Temporal resolution+ The most direct way+ Continuous observation
+ Global observation+ Good resolution for ocean
+ Temporal resolution+ Continuous observation + Free from raining
+ Possible to profiling
Weak points Weak points
- Expensive & sporadic observation- Drift while ascending
- Spatial resolution- Instrumental calibration- Saturation by dew and rain
- IR: Invisible in cloudy condition- Microwave: Land area, Temporal resolution
- Vertical distribution- Calibration for absolute IWV
- Beginning stage
- Expensive & sporadic observation- Drift while ascending
- Spatial resolution- Instrumental calibration- Saturation by dew and rain
- IR: Invisible in cloudy condition- Microwave: Land area, Temporal resolution
- Vertical distribution- Calibration for absolute IWV
- Beginning stage
VLBI
N.A.
N.A.
(Future)
4/15 4th IVS General Meeting, Concepcion, Chile, Jan. 9~11, 2006
Elgered (1993)
Water vapor absorption model and its observations by WVRs
MICAM (WVR Inter-comparison Campaign)
• Dutch weather service facility in Cabauw• Eight WVR, Radar, Ceilometers, Radiosonde• Separation btw. WVR: 30m• Total freq.: 47 different freq.
Westwater et al. (2004)
5/15 4th IVS General Meeting, Concepcion, Chile, Jan. 9~11, 2006
JPL D2, 21.0/31.4 GHz9 sessions
Radiometrics, 23.8/31.4 GHz,1 session
Astrid, 20.7/31.4 GHz, 37 sessions
25 freq., 18.8~25.7 GHz, 1 session
IEEC, Barcelona (europa.ieec.fcr.es/.../ recerca/gnss/euro_net.gif)
European geodetic VLBI & WVR network
6/15 4th IVS General Meeting, Concepcion, Chile, Jan. 9~11, 2006
Water vapor sensing instruments collocated at Wettzell
Campaign period: April 11~19, 2005
Wettzell fundamental station, Germany
Instruments• 3 ETH series WVR instruments 2 from BKG & 1 from ETH, Zurich• 2 Radiometrics 1 from Univ. BW & 1 from TU Dresden• Sun spectrometer from ETH Zurich• Radiosondes launched with balloons• GPS & VLBI
VLBI session• R1 and R4 analysed by TU-Vienna• GPS observations analysed by IGS
7/15 4th IVS General Meeting, Concepcion, Chile, Jan. 9~11, 2006
13/15 4th IVS General Meeting, Concepcion, Chile, Jan. 9~11, 2006
- 60
- 50
- 40
- 30
- 20
- 10
0
- 60
- 50
- 40
- 30
- 20
- 10
0
Comparison of vertical components btw standard solution (left) and WVR solution (right)
Status/Instrument Retrieval method
Impacts
DSS65 First operation: WAVEFRONT(‘96) JPL D2 type, 21.0/31.4 GHz Advanced WVR was developed for more precise atmospheric calibration Euro-session No. : 9 (‘99~ )
Resch(‘83) PD = Cr1 + Cr2 Tb1 + Cr3 Tb2
Johansson(‘93) PD = Cj1 [ 1 + Cj2 COS
(t – Cj3) – Cj4 (Tb – Cj5) ]
20~30 mm reduction in vertical Concentration was changed Resch: -2 mm Johansson: 3 mm Small changes in baseline-rate & WRMS
Effelsberg
First operation: Dec. ‘04 25 channel WVR: 18 GHz ~ 26 GHz Mounted on top of the Antenna Euro-session No. : 1 (‘05~ )
Tahmoush & Rogers(‘00 ) PD = Ctr Tb-peak
~15 mm increment in vertical Reference: Effelsberg
Onsala60
First operation: ‘90 Astrid, 20.7/31.4 GHz Konrad WVR was developed for meteorological project Euro-session No. : 37 (‘90~ )
Johansson(’93) PD = Cj1 [ 1 + Cj2 COS
(t – Cj3) – Cj4 (Tb – Cj5) ]
~7 mm reduction in vertical concentration was degraded ~5 mm Small changes in baseline-rate & WRMS
14/15 4th IVS General Meeting, Concepcion, Chile, Jan. 9~11, 2006
Concluding remarks
Future Task Verification of the GPS aided WVR WPD calibration
Concluding remarks Impacts of adopting WVR WPD as a tropospheric calibration are shown
• Four WVR data of European geodetic VLBI network are collected
• Three different kinds of WPD retrieval methods are applied and results are compared
Alternative WVR WPD retrieval method is planed• New approach with mixture of GPS and WVR for WPD calibration
15/15 4th IVS General Meeting, Concepcion, Chile, Jan. 9~11, 2006
Thank you for your attention.
Supplementary slides
4th IVS General Meeting Concepcion, Chile, Jan. 9~13, 2006
● European geodetic VLBI network Operation: 1990~present Application: Monitoring of local tectonic motion & glacial rebound etc.
● Motive To check the possibility of improvement in VLBI positioning results introducing WVR WPD instead of model calibration
● Primary obstacle Unpredictable water vapor contents in troposphere
● Solution Theoretical model, Radiosonde, WVR, GPS etc.
● Aim Check the impact of WVR calibration on the quality of the results of the European VLBI network and plan generalized WVR WPD retrieval scheme as a proposal
Study summary
4th IVS General Meeting Concepcion, Chile, Jan. 9~13, 2006
Primary error sources of the WVR WPD
4th IVS General Meeting Concepcion, Chile, Jan. 9~13, 2006
Instrumental calibration
Brightness temp. modeling
WPD retrieval algorithm
Elevation mask
Error sources Error item
Gain error & drift Offset error
Theoretical brightness temp.Theoretical opacity
Coefficient error
Different elevation mask btw. stations
Characteristics
Unstable behavior of raw dataDrift while observing
Laboratory values; 5~10% error for 20~32 GHz frequencies
5% of opacity model uncertainty Non-unique mapping problem
Inconsistent tropospheric delay under 5deg. of elevation mask
Primary error sources of the GPS WPD
Observation circumstances
Model uncertainty
Error sources Error item
Physical obstacle Radio interference
Inaccurate hydrostatic partmodeling
Characteristics
Causing site-dependent error
Depending on the precision of surface met. measurements
Contemporary WVR instruments
4th IVS General Meeting Concepcion, Chile, Jan. 9~13, 2006
Frequencies Characteristics Developer
NOAA/ETLDual-channel Radiometer
Dual-channel20.6 (23.87) and 31.65 GHz
• Less affected by rain drops• Internal calibration using three switches named Hach• Tip-cal calibration once per week
NOAA ETL, USA
MWR(Microwave Radiometer)
Dual-channel23.8/31.4 GHz
• Portable WVR • Calibration: Noise diode or Tip-cal method• Dew blower and moisture detector
Radiometrics co., USA
TROWARA(Tropospheric Water Vapor Radiometer)
Dual-channel21/31 GHz
• Continuous observation for IWV and LWP• Internal calibration every an hour using Tipping curve
IAP, Bern Univ., Swiss
MTP5(Meteorological Temp. Profiler)
Single-channel61 GHz
• Measure temperature from the surface to 600m altitude• Solid-state Dicke type super heterodyne receiver (1KHz)• Bandwidth: 2GHz
Attex co., Russia
MWP(Microwave
Profiler)
12-channel5ch.- 22~30GHz 7ch.- 51~59GHz
• Portable WVR, 32kg.• Self correction for frequency drift error• Measure infrared temp., Tsurface, H, and P
Radiometrics cp., USA
MICCY(Microwave
Radiometer for Cloud
Cartography)
22-channel10ch.- over 22.235GHz 10ch.- under 60GHz2ch.- around 90GHz
• Single-sideband total power radiometer• Heterodyne receiver filter bank design• Internal calibration using highly stable noise diode
MIUB, Germany
HATPRO(Humidity and Temp. Profiler)
14-channel20~60 GHz
• Total power radiometer that can detect directly to receiver• Each receiver & frequency are designed as filter bank• Flexible channel bandwidth• Reducing IF interference: high stability and accuracy
Radiometer Physics GmbH
ASMUWARA(All-Sky Multi-Wavelengt
h adiometer)
9-channel18~151 GHz
• Wideband Thermal infrared Radiometer• Equipped Camera and rain-drop sensor as well
IAP, Bern Univ.,Swiss
GSR(Ground-based
Scanning Radiometer)
23-channel50~380 GHz
• Modification version of WVRs in North pole region• New set of thermally stable calibration targets
NOAA ETL, USA
Westwater et al. (2004)
Path Delay from Various Inversion Method I
● Classical inversion coefficients: Resch (1983) & Keihm (1995) Assuming that 31.4 GHz frequency has only continuum emission 20.7 GHz frequency has water vapor line and continuumWe can get the water vapor component by subtracting scaled 31.4 GHz from 20.7 GHzThen convert from brightness temp. to PD using scale factorPD = Cr1 + Cr2 Tb1 + Cr3 Tb2 Madrid and Effelsberg Tb1: Brightness Temp. for 20.7 GHz, Tb2 : Brightness Temp. for 31.4 GHz
● Include Locality & Seasonal variation: Johansson (1993)
Measure spectrum from 18 GHz to 26 GHz in 30 channels with sweeping radiometerSeparate continuum from line emission by fitting a frequency-squared baseline and a van Vleck-Weisskopf water vapor line profilePD = Ctr Tb-peak Effelsberg Tb-peak: Water vapor spectral line intensity at 22.235 GHz
Models include many atmospheric chemical constituentsMany hundreds of transitions and their Einstein rate coefficientsMultiple layers in atmosphere, each with T, P, partial pressure water vapor,Cloud liquid water, Aerosols
+ Relationship btw. PWV and PD: Delgado et al.(ALMA MEMO No. 451) An idea using PWV from a lot of method using GPS and WVR together It may can be a generalized WVR WPD retrieval method because almost every WVR has identical PWV retrieval method. So we can spare time to get the site-and-instrument dependent WVR WPD retrieval method and just use simple value of relationship btw. PWV and PD. For example Wettzell Radiometrics uses the value of 6.50 i.e. PD = 6.5*PWV. Then we can use GPS PD as a reference PD value. There are so many studies on proof of GPS PD accuracy and precision compared with WVR PD. So we can adjust the value compared with WVR PD and GPS PD for each site. This is my idea but it will be shown as a plan in 2006 IVS meeting.
4th IVS General Meeting Concepcion, Chile, Jan. 9~13, 2006
Water vapor sensing instruments collocated at Wettzell
Results; Wettzell
4th IVS General Meeting Concepcion, Chile, Jan. 9~13, 2006
- 30.0
- 20.0
- 10.0
0.0
10.0
20.0
30.0
40.0
Up East North
Medicinawith WVR
- 100.0
- 80.0
- 60.0
- 40.0
- 20.0
0.0
20.0
Up East North
Nyales20with WVR
- 10.0
0.0
10.0
20.0
30.0
40.0
50.0
60.0
Up East North
Onsala60with WVR
- 20.0
- 10.0
0.0
10.0
20.0
30.0
40.0
Up East North
Svetloewith WVR
Standard solution WVR/Resch model WVR/Johanssen model
Euro-63
The Onsala60-DSS65 baseline result shows relatively big degradation of WRMS after introducingWVR data. But we have to note that there are only four sessions included. This means that theOnsala60-DSS65 result is easily changed by a single value.
Results; Onsala60-DSS65 baseline
4th IVS General Meeting Concepcion, Chile, Jan. 9~13, 2006
The UD (Up-Down) components have been computed with respect to the standard solution. Thereforethe reference UD component is set to zero and the other results are reported relative to this. The averageVertical components are all smaller when WVR data has been used.
Summary of the multi session results
4th IVS General Meeting Concepcion, Chile, Jan. 9~13, 2006
Design of low-cost radiometer
4th IVS General Meeting Concepcion, Chile, Jan. 9~13, 2006
Results Flexible radiometer design Several improvements from MICAM Low maintenance every 3 months
WP 2600 Description of work Design a low cost microwave radiometer for automatic, high accuracy LWP measurement Estimation of cost for different levels of LWP accuracy Development of a calibration concept to Guarantee low maintenance