TIGGE * 1 2 1 2 JAMSTEC
2
3
Strategy Global Ens. Forecast Potential Occurrence - Timing ~ O(day) - Location ~ O(100km) time Regional Hi- Res. Forecast - Estimate Quantitatively - Timing ~ O(hr) - Location ~ O(10 km) High Impact Weather Event Occurs! 6
1. Global Hi-Res. Deterministic Forecast Only My Choice! Global Lo-Res. Ensemble Forecast 2. + Regional Hi-Res. Forecast 3. Global Hi-Res. Deterministic Forecast + Regional Hi-Res. Forecast 7
2008 11 2007 2009 10Lupit
2008 11
2008 11
Vietnam : Flood : November 2008 Torrential rainfall in early November in northern and central Vietnam caused severe flooding that left at least 85 people dead, damaged 180,000 houses and devastated over 265,000 hectares of rice and vegetable fields. An estimated 600,000 people were affected. United Nations Office of Coordination for Humanitarian Affairs (OCHA) http://ochaonline.un.org/roap/disasterhistory/ tabid/4837/language/ja-jp/default.aspx IRIN http://www.irinnews.org/report.aspx?reportid=81253 EM-DAT http://www.emdat.be/database
Genesis Potential using TIGGE EPSs for Vietnam Flood in Oct. 2008 192hr 216hr TD TD TD TD TD Color Shading: Percentage of Members, which exceed 90 Percentile of Us
2007 3
Genesis Potential using TIGGE EPSs for Gobi Sand Storm in Mar. 2007 96hr Russia Mongolia China Color Shading: Percentage of Members, which exceed 95 Percentile of Us
WMO/WWRP
Strike Probability in 120 km, 4 days CMA MSC ECMWF JMA-W JMA-T KMA NCEP STI UKMO http://tparc.mri-jma.go.jp/cyclone
An Example of the Strike Probability Map
JMA-WEPS Taipei Sinlaku (km)
Lupit 2009 10
Global Deterministic High Resolution Oct. 20-22 2009 ECMWF WWRP JMA
Global Deterministic High Resolution ECMWF Oct. 17-19 2009 WWRP No Recurvature! JMA
WWRP Oct. 20 Oct. 21 Oct. 22 JMA-W JMA-T ECMWF
Oct. 17 Oct. 18 Oct. 19 JMA-W JMA-T WWRP ECMWF Member Spread at Day 5
Model Representations for Extreme Us
Model Representations for Extreme Us
Model Representations for Extreme Us
Model Representations for Extreme Prcp
Model Representations for Extreme Prcp
Parameter Candidates for Target Disasters Disaster Typhoon Flood Wind, Dust Storm Draught Cold Surge Heat Wave Parameters Us, Ps, Vor850 Us, Precip, Vor850 Us Ts, qs, Precip Ts, T700, Us Ts
Tropical Cyclone ITC=max(P(Us), P(Ps), P(Vor850)) P(Us): Percentage of members, which are above 95 percentile P(Vor850): Percentage of members, which are above 95 percentile P(Ps): Percentage of members, which are below 5 percentile Flood IFL=max(P(Precip), P(Us), P(Vor850)) P(Precip): Percentage of members, which are above 95 percentile P(Us): Percentage of members, which are above 95 percentile P(Vor850): Percentage of members, which are above 95 percentile Wind/Dust Storm IWD=P(Us) P(Us): Percentage of members, which are above 95 percentile
Draught IDR=max(P(Ts), P(qs), P(Precip)) P(Ts): Percentage of members, which are above 95 percentile P(qs): Percentage of members, which are below 5 percentile P(Precip): Percentage of members, which are below 5 percentile Cold Surge in 10-day average ICS=max(P(Ts), P(T700), P(Us)) P(Ts): Percentage of members, which are below 5 percentile P(T700): Percentage of members, which are below 5 percentile P(Us): Percentage of members, which are above 95 percentile Heat Wave IHW=P(Ts) P(Ts): Percentage of members, which are above 95 percentile
TIGGE 38
PDF JRA-25/JCDAS PDF GSMaP JCDAS 39