3WCDRR: UN GAR Global Risk Assessment

Transcription

1 3WCDRR: UN GAR Global Risk Assssmnt

2 Scoping Mting Agricultural Risk Assssmnt Incorporating drought and th agricultural sctor in th GAR Global Risk Modl 7 9 Fbruary 2017 Bouldr, Colorado GAR Global Risk Assssmnt : At Glanc MULTI-HAZARD FULLY PROBABILISTIC RISK ASSESSMENT RESULTS AT GLOBAL LEVEL Dr. Prof. Omar-Dario CARDONA A. National Univrsity of Colombia at Manizals Rprsntativ INGENIAR Intrnational Cntr for Numrical Mthods in Enginring & INGENIAR LTDA

3 Why a Global Risk Assssmnt? ü Masuring is ssntial to dcid; what is not dimnsiond cannot b administratd ü An oprational pictur of risk improvs risk knowldg and provids an ovrall risk landscap ü Risk assssmnt is ky to awar but also to concrn dcision-makrs of thir rsponsibility ü Disastr risk is a contingnt liability and thrfor a sovrign risk for th socity ü Risk rduction and prvntion ar dutis for risk govrnanc and for th nations accountability ü Track DRM progrss ovr tim mans considring th dvlopmnt transformation trad-offs

4 COMPREHENSIVE APPROACH TO PROBABILISTIC RISK ASSESSMENT

5 Cyclonic Wind and Sismic Hazard St of stochastic scnarios ü Mutually xclusiv ü Collctivly xhaustiv ü Admit probabilistic rprsntation.ame FORMAT

6 Sismic Hazard Assssmnt Intnsity xcdanc curvs in ach nod of th global grid T = 100 yars ν = 0.01 Excdanc rat (1/yar) T =500 yars ν = T = 1000 yars ν = Intnsity masur Intnsity masur Hazard maps for diffrnt rturn priods

7 Sismic Hazard Assssmnt Earthquak Hazard Maps Cardona t.al. (2015) T= 0.5 sc, Tr = 475 yars

8 Cyclonic Wind Hazard Tropical Cyclon Wind Hazard Maps Cardona t.al. (2015) Tr=100 yars

9 Othr Probabilistic Hazards Tsunami, Floods, Volcanic Ash, Storm Surg, Climat Chang Scnarios Tsunami: Run-up hight for Tr 500 yars Løvholt t.al. (2014) Floods for Tr 1000 yars CIMA (2014) Volcanic ash load for Tr 100 yars Bar-Crozir t.al, 2014

10 Global Exposur Databas 5x5 km grids & 1x1 km in th coasts GRID, UNEP (2014)

11 Vulnrability Vulnrability functions for diffrnt hazards

12 Vulnrability Vulnrability functions for arthquaks SEISMIC DESIGN LEVEL: M MDR (%) 100% 90% MDR Var. 80% 70% 60% 50% 40% 30% 20% 10% 0% Spctral Acclration, Sa (g)

13 Vulnrability Vulnrability functions for wind QUALITY LEVEL: M MDR (%) 100% 90% MDR Var. 80% 70% 60% 50% 40% WOOD REINFORCED MASONRY REINFORCED CONCRETE FRAMES 30% 20% 10% 0% Wind Spd (kph)

14 Risk Modling: Loss Assssmnt Hazard Exposd Assts Vulnrability Risk PDF of Loss Losss Economic Human Loss LOSS EXCEEDANCE CURVE PROBABLE MAXIMUM LOSS AVERAGE ANNUAL LOSS Loss Excdanc Rat v(p) AAL PML

15 Risk Modlling: Loss Assssmnt Considrd vnt from th stochastic st Considrd asst (or grid) at analysis ara Considrd building(s) class(s) Vulnrability function slction Hazard intnsity at analysis sit Loss and varianc calculation Ys No Any othr building class at th sit? No Has all ara bn analyzd? Ys Loss probability distribution function calculation for th vnt Ys Hav all vnts bn considrd? No Loss xcdanc rat calculation using th loss PDF associatd to ach vnt and its occurrnc frquncy

16 f 0 ( l Evnt i) = f ( l Sa) f ( Sa Evnt i) dsa Vulnrability Hazard f(x) f(x) v Evnts ( p) = i= 1 p P Pr ( l L Evnt i) F A p ( Evnt i) Loss xcdanc rat (#/yar) Losss P

17 Loss Excdanc Curv Govrnmnts nd to dfin a risk rduction/financing stratgy 10 Rtntion (budgt / contingnt crdit) Transfr Rtntion 0.1 Loss xcdanc (#/yar) Planning / Prvntion / Mitigation Rgulation (Cods / Standards / Land us) (Rsidual Risk) Warning / Emrgncy rspons Layr 1 Layr 2 Layr 3 Layr Rturn priod (yars) Loss ($ 000) 1 = High probability & low/modrat losss 2 = Mdium probability & modrat/high losss 3 = Low probability & high losss 4 = Vry low probability & vry high losss

18 Extnsiv and Intnsiv Risk Assssmnt Analytical (propctiv) Loss Excdanc Curv 1, a r] / y [1 t ra n c a d c x s L o ] rs a [y d rio p rn tu R , , , ,000 10, ,000 Economic loss [Million $USD] Analytical curv

19 Extnsiv and Intnsiv Risk Assssmnt Empiric (rtrospctiv) Loss Excdanc Curv 1, a r] / y [1 t ra n c a d c x s L o ] rs a [y d rio p rn tu R , , , ,000 10, ,000 Economic loss [Million $USD] Historic curv Analytical curv

20 Proposal of a Hybrid Loss Excdanc Curv 1, a r] / y [1 t ra n c a d c x s L o AAL ] rs a [y d rio p rn tu R , , PML ,000 10, ,000 Economic loss [Million $USD] Historic curv Analytical curv Hybrid curv 100,000

21 Sismic Risk Assssmnt Risk Maps Distribution of AAL by country, US$ millions

22 Sismic Risk Assssmnt Risk Maps Distribution of AAL rlativ to th xposd valu by country

23 Sismic Risk Assssmnt Risk Maps Distribution of PML 1500 by country, US$ millions

24 Sismic Risk Assssmnt Risk Maps Distribution of PML 1500 rlativ to th xposd valu by country

25 Risk Rankings of Countris Using Avrag Annual Loss (AAL) rsults

26 Multi-hazard Risk Assssmnt Risk Maps Distribution of AAL by country, US$ millions

27 Multi-hazard Risk Assssmnt Risk Maps Distribution of AAL rlativ to th xposd valu by country

28 Som Global Figurs by Rgion and Economy Lvl Exposur Valu (PC)

29 Som Global Figurs by Rgion (Multi-hazard) AAL Valu AAL/PC

30 Som Global Figurs by Economy Lvl (Multi-hazard) AAL Valu AAL/PC

31 Disastr Risk Dvlopmnt Implications Indx (DRDI)

32

33 Sismic Risk Assssmnt Risk Maps CAPRA-Viwr

34 Th Challng: From Global to Local

35 Rgarding th GAR s Global Risk Modl: ü A fully probabilistic multi-hazard risk assssmnt has bn conductd at global lvl. Rsults ar usful for comparisons and rankings among countris ü In th past, risk had bn stimatd mainly basd on historical rcords. Th GAR s Global Risk Modl taks into account vnts that hav not yt occurrd. This is a ky issu for SFA follow-up. ü Countris must carry-out risk assssmnts with highr rsolution at sub-national and local lvl whn th rquird information is availabl. ü From global to local: th sam arithmtic can b usd for any rsolution lvl.

36 Global Risk Assssmnt (Modl): Earthquaks, Tsunami, Tropical Cyclons (w and w/o cc), Rivrin floods, Volcanic ash-fall, Droughts?

37 Undrstanding probabl losss du to hazard vnts (not only disastrs) crats powrful incntivs for countris to dvlop planning options and tools to cop with risk, including allocating th sustaind budgtary rsourcs ncssary to rduc th potntial damag and safguard th transformation of dvlopmnt.

38 It is ncssary to xplor how probabilistic risk modls and holistic modls can b intgratd in an ffort to dvlop a mthodology that can provid th tru bnfits of aggrgation and disaggrgation of hard and soft risk drivrs to idntify th corrctiv and prospctiv intrvntions to b don.

39 Far bttr an approximat answr to th right qustion, which is oftn vagu, than an xact answr to th wrong qustion, which can always b mad prcis John W. Tuky