IMPACT OF GPS RADIO OCCULTATION DATA ON ANALYSIS AND PREDICTION OF AN INTENSE SYNOPTIC-SCALE STORM OVER THE SOUTHERN OCEAN NEAR THE ANTARCTIC PENINSULA Shu-Ya Chen 1, Tae-Kwon Wee 1, Ying-Hwa Kuo 1,2, and David H. Bromwich 3 1 University Corporation for Atmospheric Research, Boulder, Colorado 2 National Center for Atmospheric Research, Boulder, Colorado 3 The Ohio State University, Columbus, Ohio
Outline Introduction Model configurations and experimental designs Simulated results and Discussion Summary
Introduction u Precipitation accompanied by synoptic-scale cyclones is one of the major sources of water mass over Antarctica. However, the structure and developing mechanism of the synoptic-scale cyclones near Antarctica are little known compared to other areas on the globe due to scarcity of observation. Therefore, GPS Radio Occultation (RO) data could be a vital source of information over the Antarctic area. u In this study, we investigate a synoptic-scale storms, with a rapidly enhancement to 923 hpa during three days, by assimilating GPS RO data. u Before investigating the data impact, many factors from model configuration, initial condition, assimilating strategy, and the way of data used have been tested.
Model domain & GPS RO distribution u Track from ECMWF analysis (black-dot line) L 923hPa u Sea-Level Pressure decreases ~70 hpa during 3 days u SLP of 923 hpa at 2007/12/12_18UTC u GPS RO (gray dots) : total of 2403 during 08 15 December 2007 Contour interval : 10 hpa
Many factors were tested. NG: assimilates only GTS data WG: assimilates both GTS and GPS RO data Ø Model resolution (against GPS) 44 soundings 0000 UTC 0600 UTC 1200 UTC 1800 UTC N N Height (km) 30 25 20 15 10 GPS RO data around the storm within 1000-km radius. Ø Assimilated strategy (against EC) Ø Initial time (against GPS) 5 0 N OBS - FST NG WG 0 0.5 1 1.5 Standard Deviation (N-unit) Init. Time: 0800,0812,0900,0912,1000,1012,1100 Z NG: GTS data only; WG: GTS plus GPS RO data; WG_A: discarding GPS RO above 50 hpa; WG_B: with vertical data thinning for GPS RO; WG_C: combine the effect of WG_A and WG_B. T
Model configurations and experimental designs. Model Domain WRF v3.3.1 and WRFDA v3.3.1 (3DVAR) 401x401x55 grid-points with 30-km horizontal resolution Model top: 10 hpa Observations Conventional data (GTS) GPS RO (COSMIC, CHAMP, SAC-C, GRACE) Physics Experiments Microphysics: WSM-5 class scheme Longwave radiation: rrtmg scheme Shortwave radiation: Goddard short wave Surface layer: Monin-Obukhov (Janjic Eta) Land surface: Unified Noah land-surface model Boundary layer: Mellor-Yamada-Janjic(Eta) TKE Cumulus: Kain-Fritsch (new Eta) scheme Data source for initial condition: NCEP FNL analysis (2007/12/09_00 UTC) First guess for WRFDA: 6h WRF forecast Cycling: three updated cycling (06, 12, and 18 UTC on 09 th Dec.) NG: assimilates only GTS data WG: assimilates both GTS and GPS RO data Forecast for 5 days Verification: ECMWF with 25km horizontal resolution and NCEP with 1 o x 1 o resolution
Track and Intensity 12_18 UTC 1218 The simulated intensity for WG is closer to the global analyses.
Verifications on grid points (whole domain) Black line: NG Gray line: WG Solid line: against EC Dash line: against NCEP Z T Q
Verified against ECMWF (SD) A square box of 900 km x 900 km, which is centered at the storm, is extracted for comparison. Temperature (unit:k) water vapor mixing ratio (unit: g kg -1 ). Pressure (hpa) Date_Hour Date_Hour
500-hPa RV & SLP ECMWF NCEP 2007/12/12_18 UTC weaker in RV NG WG Ø Weaker relative vorticity in NCEP may be caused by coarse resolution.
WG-NG UV + T @ 500 hpa UV + T @ 850 hpa UV + q @ 850 hpa 24h 48h 72h
Potential Temperature & Geopotential height (500 hpa) NG f48h WG f48h WG-NG f48h larger temperature gradient à baroclinicity à deepening trough
Difference of Eady growth rate between WG and NG at 500 hpa. Baroclinicity effect : Eady growth rate (Bromwich et al. 2011) WG usually has bigger Eady growth rate (baroclinicity) than NG above 500 hpa during the cyclonic development WG-NG at f54h WG-NG at f60h M2 g dθ σ Eady = 0.31, M2 = N θ dy N : Brunt - V a is a l a frequency WG-NG at f66h
Meridional cross-section x
EC NG WG Potential Vorticity 2007/12/11_18 UTC NG-EC WG-EC
Water vapor mixing ratio & Vertical velocity (WG-NG) 48h More moisture and bigger vertical velocity for exp. WG during the cyclonic development. 54h 60h 66h
Summary and future work Ø Ø Ø Ø Ø Verifications against either ECMWF or NCEP show that assimilated with GPS RO data has positive impact. Both experiments (NG and WG) present similar track predictions as the global analyses, but assimilated with GPS RO data (WG) can further improve the prediction of cyclonic intensity. Assimilating GPS RO provides a larger temperature gradient, then deepening the trough and further facilitating the cyclonic development. WG improve the structure of potential vorticity, and then it propagates to lower levels. WG also provide more moisture and high vertical velocity in lower troposphere. The role of moisture field from GPS RO that contributes on cyclonic development will be further investigated.
THE END ~ Thank you