Evaluation of the TanDEM-X Digital Elevation Model by PPP GPS -Analysis and Intermediate Results-

Transcription

1 Evaluation of the TanDEM-X Digital Elevation Model by PPP GPS -Analysis and Intermediate Results- XXIV FIG International Congress 2010 TS 8C New GNSS Applications and Developments Jürgen Schweitzer¹, Bimin Zheng¹, Volker Schwieger¹, Detlev Kosmann² ¹, ² German Aerospace Centre, German Remote Sensing Data Centre, Wessling Germany Sydney, Australia, April 15, 2010 Structure Background and Motivation Precise Point Positioning Conclusions and Outlook XXIV FIG International Congress, Sydney, Australia Saturday, 08 May Sydney, Australia, April

2 Motivation First Mission using Synthetic Aperture Radar (SAR) First homogeneous global Digital Elevation Model (DEM) Accuracy: 6 to 10m (resolution 30m) New DEM Mission: TandDEM-X Expected accuracy: 2m relative (resolution 12m) Kinematic evaluation method : Precise Point Positioning (PPP) 2 Precise Point Positioning Characteristics One Receiver No reference station Continuous dual-frequency code and carrier phase observations Precise correction models required 3 Sydney, Australia, April

3 Precise Point Positioning Characteristics Extract from IGS Products (IGS, 2009) Broadcast Ultra-Rapid (predicted half) Rapid Final Orbits Sat. clocks Orbits Sat. clocks Accuracy Latency Sampling Interval ~100cm ~2.5ns SDev ~5cm 1.5ns SDev Real time Real Time Daily 15 min Orbits 2.5cm 15 minutes 1 41 hours Sat. clocks 25ps SDev 5 minutes Orbits 2.5cm 15 minutes days Sat. clocks ~20ps SDev 30 seconds 4 Precise Point Positioning Processing Procedure Raw Data 10Hz RINEX files Software GIPSY OASIS (GOA II) from JPL, USA GPS-Inferred Positioning System and Orbit Analysis Simulation Software - Freeware for research and teaching purposes - Command line based software runs on Fedora 10 - No user manual CSRS-PPP from National Resources Canada Canadian Spatial Reference System - Web - Interface 5 Sydney, Australia, April

4 Precise Point Positioning Processing Procedure Parallel Processing Averaging of positions with dh < 1m Results compared with PDGPS 6 Sydney, Australia, April

5 Current Status of Data Acquisition Position Direction Start End Length [km] Receiver Forward 05/13/08 05/15/08 99 Leica GX Munich - Ukraine Backward 05/15/08 05/1/ Leica GX1230 Forward 06/09/08 06/13/ Leica GX Munich - Sao Martinho Backward 06/15/08 06/28/ Leica GX Beijing - Gauquan Forward 10/04/08 10/11/ Leica GX1230 Laguna Verde - Punta Forward 11/21/08 11/23/ Leica GX De Choros Backward 11/23/08 11/24/ Leica GX1230 Vina Del Mar - Mar Forward 11/24/08 11/28/ Leica GX Del Plata Backward 12/01/08 12/09/ Leica GX Nairobi - Outjo Forward 06/14/09 06/24/ TPS HIPER_GGD Krasnojarsk - Belgorod Forward 0/20/09 0/31/ Leica GX Recife - Porto Veiho Forward 0/31/09 08/0/ TPS HIPER_LITE Total length = 2949 km 8 Results of Post-Processing Duration RMS_dh Point density Position Direction Epochs [Hour] Avail. [m] [1/km] Forward % Munich Ukraine Backward % Munich Sao Forward % Martinho Backward % Beijing Gauquan Forward % Laguna Verde Punta Forward % De Choros Backward % Vina Del Mar Mar Forward % Del Plata Backward % Nairobi Outjo Forward % Krasnojarsk Belgorod Forward % Recife Porto Veiho Forward % Weighted mean 59.2% Sydney, Australia, April

6 Availability Results of Post-Processing 80,00% 0,00% 60,00% 50,00% 40,00% 30,00% 20,00% 10,00% 0,00% Availability Forward Backward RMS_dh [m] 0,0 0,60 0,50 0,40 0,30 0,20 0,10 0, RMS of height difference Forward Backward 10 Results of Post-Processing Results of PDGPS Num. of MEAN_dh RMS_dh Ref. Station comparisons [m] [m] LINZ OBE OBE SALA URUM No ref. Station 5 Sant No ref. Station NVSK BRAZ Sydney, Australia, April

7 Analysis of Accuracy-Height Difference Correlation Height difference of each epoch: Position accuracy for each epoch: Average Standard deviation of Correlation 12 Analysis of Accuracy-Height Difference Correlation in cm GIPSY filtered GIPSY rejected CSRS filtered CSRS rejected MEAN Stdv MEAN Stdv MEAN Stdv MEAN Stdv track subtrack s3d s3d s3d s3d s3d s3d s3d s3d avail % % % % % % % % % % % % 13 Sydney, Australia, April 2010

8 Analysis of Accuracy-Height Difference Correlation Correlation not existent Correlation existent 14 Conclusions and Outlook Evaluation of TanDEM-X by PPP processing of the first 8 tracks Average height RMS: 0.52m Average availability: 59% The correlation between height difference and internal accuracies is obvious, but not stringent for all data. Further tracks in Ukraine, Canada, West Africa, Australia and India are available and will be analyzed soon 15 Sydney, Australia, April

9 Thanks for your attention! Dipl. Ing. Jürgen Schweitzer / Dipl. Ing. Bimin Zheng/ Prof. Dr. Ing. habil. Volker Schwieger Institute for Applications of Geodesy to Engineering, University Stuttgart Geschwister Scholl Str. 24 D 014 Stuttgart Germany CONTACTS Phone: / 84061/ Fax: E mail: juergen.schweitzer@iagb.uni stuttgart.de bimin.zheng@iagb.uni stuttgart.de volker.schwieger@iagb.uni stuttgart.de Dipl. Ing. Detlev Kosmann German Remote Sensing Data Centre, German Aerospace Centre Münchnerstr Wessling Germany Phone: Fax: E mail: detlev.kosmann@dlr.de Sydney, Australia, April