Climate applications of ground-based GPS measurements

1. Climate applications of ground-based GPS measurements Junhong (June) Wang Earth Observing Laboratory National Center for Atmospheric Research Boulder, CO, USA Collaborators: Liangying (Liz) Zhang (EOL), Aiguo Dai (CGD), Teresa Van Hove (UCAR/COSMIC) and Joel Van Baelen (CNRS) Thank the support from NOAA Climate Change Data and Detection program

2. Earth Observing Laboratory C-130 ELDORA Dropsonde HIAPER Mobile GAUS S-pol ISS MAPR ISFF

3. EOL Sounding Systems GAUS TAOS Mobile GAUS Dropsonde Drift sonde Reference sonde

4. Outline • Motivation • GPS measurement and analysis techniques, and GPS PW dataset • Application #1: Quantifying systematic errors in global radiosonde humidity data • Application #2: Diurnal variations and reanalysis validations • Application #3: State of climate and climate trend • Summary

5. Available Water Vapor Observing Systems Satellite (Geostationary, polar-orbiting, GPS) Rawinsonde network Aircraft (research, commercial, remotely-piloted) (in-situ, remote sensing, dropsondes) Remote sensing (radar, lidar, radiometer, GPS ..)

6. Global radiosonde network (WMO report, July 2007) • Problems: • Errors and biases • Spatial and temporal inhomogeneity • Spatial sampling errors • Diurnal sampling errors Results: The role of radiosonde observations in climate studies is limited. Solutions: To quantify radiosonde errors and correct them.

7. How does GPS estimate precipitable water? Total delay = Ionosphere + dry + wet Noise (geodesy) Signal (meteorology) ZTD = ZHD + ZWD ZWD = ZTD - ZHD PW =  * ZWD  = f (Tm)

8. Why use GPS data for PW?

9. Global ZTD data from International GNSS Service (IGS)~420 stations, 1997-present, 2-hourly 1997: ~100 2010: ~421 See Wang et al. (2005, 2006)

10. Analysis Procedure

11. Feb. 1997 to Feb. 2009 • 2 hourly (0100, 0300, …, 2300 UTC) • 380 IGS, 169 SuomiNet, 1223 GEONET • Accuracy: < 3 mm • Ps, Tm, ZHD and ZWD also available • Request data: junhong@ucar.edu A global, 12-year, 2-hourly PW dataset from ground-based GPS measurements(Wang et al. 2007, JGR)

12. Highlight of GPS-PW data

13. Hurricane Ernesto (24 Aug – 1 Sep. 2006)

14. Matched GPS and radiosonde data (< 50 km in distance, < 100 m in elevation, < 2 hours; 14 types and 136 stations) • Humidity sensors: • Capacitive • Carbon hygristor • Goldbeater’s skin Wang and Zhang (2008a)

15. Capacitive Carbon hygristor Goldbeater’s skin median -1.67 1.97 0.81 S.D. 1.72 4.15 1.93 Systematic errors – mean biases Wang and Zhang (2008a)

16. Vaisala RS80-A with cover PW difference (mm radiosonde-GPS) without cover Sensor boom cover with cover Vaisala RS80-H PW difference (mm radiosonde-GPS) without cover Wang et al. (2002) Impacts of the sensor boom cover on Vaisala RS80 dry bias Wang and Zhang (2008a)

17. Day/Night difference for Vaisala sondes RS80A Day Night RS90 RS80 RS90/92 Wang and Zhang (2008a)

18. Temporal inhomogeneity of radiosonde PW data carbon hygristor capacitive with cover capacitive carbon hygristor Goldbeater’s skin Carbon hygristor Miami, U.S.A Suwon-Shi, Korea Relative PW differences (% Radiosonde-GPS) Beijing, China Wang and Zhang (2008)

19. PW comparison (Radiosonde-GPS) Homogenization global radiosonde humidity data RH comparison (TiMREX) Improved RS2-91 RS2-91 RSII-80 RH wet bias corr. at T<0C RSII-56 Courtesy of Aiguo Dai and Paul Ciesielski RH (%)

20. Summary on systematic errors of global radiosonde PW data • Systematic biases: systematic biases in three types of widely-used humidity sensors, including dry biases in capacitive sensor and wet biases in carbon hygristor and Goldbeater’s skin. • Characteristics: The dry bias in Vaisala sondes has larger magnitudes during the day than at night, especially for RS90 and RS92. The biases vary with PW and radiosonde types. • Temporal inhomogeneity: Several categories of radiosonde type changes, such as from VIZ to Vaisala and from RS80 to RS92, have been detected by the time series of PW differences between radiosonde and GPS. • Diurnal sampling errors: Diurnal sampling errors of twice daily radiosonde data are generally within 2%, but can be as much as 10-15% for the once daily soundings.

21. Global PW diurnal anomaly Globe S. H. N. H. • The diurnal cycle is less than 5% of annual mean PW • Larger magnitude in summer than in winter • Peak around late afternoon to early evening • An order of magnitude smaller than seasonal variation

22. PW diurnal variations in four regions Europe 30-70S Month Month LST LST mm N.H. Mountains Darwin region Month Month LST LST Wang and Zhang (2008b)

23. Comparisons of PW diurnal cycle between reanalysis and GPS Data

24. Diurnal cycle in U.S.A. in JJA (Amplitude)

25. Diurnal cycle in U.S.A. in JJA (Phase)

26. Diurnal Phase transition over Great Plains in U.S.A. Carbone et al. 2002

27. Seasonal variations of diurnal and sub-monthly variability over Europe GPS NCEP/NCAR JRA ERA-40 mm Wang and Zhang (2008c)

28. Summary on PW diurnal cycle • The PW diurnal cycle is small, but significant. Global, N.H., S.H. annual mean peak-to-peak amplitudes are 0.66, 0.53 and 1.11 mm, respectively. On global and hemispheric average, PW peaks from late afternoon to mid-night. • Seasonal variations of diurnal cycle in different regions are shown. The sub-monthly variability of PW has much larger magnitude than the diurnal cycle. • The PW diurnal cycle in reanalyses is compared against the GPS data over U.S.A. and Europe. Over U.S.A., NARR performs best. It is not clear whether it is due to its high resolution or assimilation of surface moisture data. Over Europe, ERA performs better than others.

29. Contributions to State of Climate for 2008 and 2009 Mears et al. (2010)

30. TCWV difference between 2008 and 2009 Dramatic moistening of the Tropical Pacific in 2009; This moistening extending across the Amazon basin in COSMIC; Good agreements among three. SSM/I & GPS COSMIC Mears et al. (2010)

31. Global PW anomaly from SSM/I (ocean), G.B. GPS (land) and COSMIC Mears et al. (2010)

33. 2-hrly combined to 5-min PPP-derived ZTD products Relative to absolute antenna phase model How good is the GPS?

34. Summary • Dataset: A global, 12-year, 2-hourly GPS-PW dataset is created for various scientific applications. You are welcome to use it (junhong@ucar.edu). • 2. Climate applications: The dataset is used to quantify systematic errors in global radiosonde PW data, study diurnal variations, validate global reanalysis products and study water vapor trend. • 3. More information: • Wang, J., and L. Zhang, 2008: Validation of Atmospheric Precipitable Water in Three Reanalysis Products using Ground-based GPS Measurements, extended abstract for Third WCRP International Conference on Reanalysis, Jan. 28 – Feb. 1, 2008, Tokyo, Japan. • Wang, J., and L. Zhang, 2009: Climate applications of a global, 2-hourly atmospheric precipitable water dataset from IGS ground-based GPS measurements, J. of Geodesy, 83, 209-217. • Wang, J., and L. Zhang, 2008: Systematic errors in global radiosonde precipitable water data from comparisons with ground-based GPS measurements. J. Climate, , 2218-2238. • Wang, J., L. Zhang, A. Dai, T. Van Hove and J. Van Baelen, 2007: A near-global, 8-year, 2-hourly atmospheric precipitable water dataset from ground-based GPS measurements, J. Geophys. Res., 112, D11107. doi;10.1029/2006JD007529. . • Wang, J., L. Zhang, and A. Dai, Global estimates of water-vapor-weighted mean temperature of the atmosphere for GPS applications. J. Geophys. Res., 110, D21101, doi:10.1029/2005JD006215, 2005.