Rotterdam May 10 th, 2016 Ho Chi Minh City adaptation to increasing flood risk Image: VCAPS (2013) Paolo Scussolini, Ralph Lasage, Elco Koks, Andres Diaz Loaiza, Van-Thu Tran, Ho Long Phi Hazards in Ho Chi Minh City 2 VCAPS (2013) 1
Hazards in Ho Chi Minh City 3 VCAPS (2013) Hazards in Ho Chi Minh City 4 VCAPS (2013) 2
Hazards in Ho Chi Minh City 5 VCAPS (2013) Hazards in Ho Chi Minh City 6 VCAPS (2013) 3
Hazards in Ho Chi Minh City Policy objectives: Developing into a modern and attractive industrial city No more floods Protect strategic harbor Integrate climate adaptation into urban design Tackling also extreme temperatures Green-look 7 VCAPS (2013) Table of contents Our approach to flood impacts (Methods) Results Flood hazard present and future Flood risk Effect of subsidence Indirect economic damage Conclusions Next steps 8 4
Future scenarios Climate change Socio-econ. change 10/05/2016 Our approach Sea level rise 5 GCMs Indirect economic damage Storm surges Subsidence Hydro modeling TELEMAC 2D Flood risk: Area flooded People flooded Direct damage Flooding of infrastructure Socioeconomic growth 2100 2100 Flood maps present Flood with 5 return maps periods with Flood 5 return maps periods Vulnerability curves Popul. Growth; Land use change Damage scanner Discount rates Adaptation measures Flood damage 4 return periods 9 Our approach Future scenarios Present Year 2100 RCP 4.5 RCP 8.5 High-end Sea level rise (cm) 0 49 180 Land use (year) 2010 2025 2025 Population change (%) 0 +41 +41 Regional output of General Circulation Models: CanESM2 CNRM-CM5 HadGEM2-AO IPSL-CM5A-LR MPI-ESM-MR Jevrajeva et al. (2014) Projections World Bank (2014) DPA (2010) Demographia (2014) 10 5
Present 10/05/2016 Our approach Hydro modeling Hydrodynamic model TELEMAC 2D Boundary conditions: Extreme sea levels - 4 return periods Rivers discharge - 10-year event Precipitation extreme - 5-year event Calibration and validation by comparison with 6 gauge stations in the city 11 Flood hazard Present 13 6
RCP4.5 (2080-2100: 49 cm) RCP 8.5 High-end (2080-2100: 180 cm) 10/05/2016 Flood hazard Future (moderate Sea level rise) 14 Flood hazard Future (High-end Sea level rise) 15 7
Our approach 2010 2025 Land use maps 16 Using depth-damage curves specific to the land use Our approach Lasage et al. (2014) 17 8
Flood damage Without adaptation Annual damage Exposure may be much higher Underestimation Adaptation will occur Overestimation 18 Current adaptation: Sparse levees Private measures Wet proofing Adaptation Simulated adaptation: Business as Usual Ring dike Elevation Dry-proofing buildings 19 9
Adaptation strategies - simulated Business as Usual A number of sparse dykes and levees, but with poor effectiveness. Drainage system is outdated and scarcely maintained. Ring dike Two ring dikes are modelled that encircle the urban HCMC, with height of 2 and 3 m. A ring dike is part of a ministerial project that is likely to be undertaken soon (MARD, 2013). Elevation Dryproofing buildings We model the elevation of the areas at highest risk. We elevate land to 30 cm above the present 100-year flood height. With the on-going urbanization, it makes sense to use sand to elevate new building ground. We simulate retrofitting residential houses and small businesses, by sealing them to waters up to 1 and 2 m. This strategy is particularly interesting because of its scalability, low cost, and relatively fast implementation. 20 Adaptation - simulated Blue gradient : water depth of the 100- year flood (present) Structural adaptation: Ring dike Elevated areas 21 10
Results Indicators % Area flooded >1 m (100-year flood) Damage of 100-year flood (M$) Annual damage (M$/year) Annual # people flooded by > 10 cm (*10 3 ) Annual # people flooded by > 1 m (*10 3 ) Flooding of critical infrastructure Presen t Business as Usual 28 27 2,368 3,416 2,046 3,035 3,401 5,275 68 508 Year 2100 Scenario RCP4.5 (moderate sea level rise) Ring dike 2 m 25 (-9%) 2,382 (-30%) 2,145 (-29%) 1,559 (-70%) 397 (-22%) Additional adaptation strategies Ring Dryproofing dike Elevation 3 m 1 m 25 22 (-9%) (-20%) 2,382 2,322 1,709 (-30%) (-32%) (-50%) 2,145 (-29%) 1,559 (-70%) 397 (-22%) 1,990 (-34%) 4,443 (-53%) 215 (-98%) 1,376 (-55%) Dryproofing 2 m 1,184 (-65%) 1,049 (-65%) - -- ++ ++ -- -- -- 22 Results Indicators % Area flooded >1 m (100-year flood) Damage of 100-year flood (M$) Annual damage (M$/year) Annual # people flooded by > 10 cm (*10 3 ) Annual # people flooded by > 1 m (*10 3 ) Flooding of critical infrastructure Presen t Business as Usual 28 64 2,368 6,268 2,046 5,861 3,401 6,823 68 2,856 Year 2100 Scenario RCP8.5 (High-end sea level rise) Ring dike 2 m 64 (0%) 6,268 (0%) 5,861 (0%) 6,823 (0%) 2,856 (0%) Additional adaptation strategies Ring Dryproofing dike Elevation 3 m 1 m 64 63 (0%) (-2%) 6,268 5,814 5,194 (0%) (-7%) (-17%) 3,945 (-33%) 2,012 (-71%) 1,366 (-52%) 5,328 (-9%) 6,819 (-31%) 2,483 (-75%) 4,658 (-21%) Dryproofing 2 m 2,842 (-55%) 2,451 (-58%) - -- -- + -- -- -- 23 11
Adaptation - simulated Costs of adaptation measures Dike ring 0.9 M$ /km for 2 m high dike; 1.5 M$/km for 3 m high dike (Hillen et al., 2014) 8 M$/km (FIM report, 2013) 2.4K$/km Yearly maintenance costs (Lasage et al., 2014) 20-30 K$/km yearly maintenance (Hillen, 2008; Mai et al., 2008) Elevating land Price of sand plus transportation: 304520 Dong, or 14 $ per m3 (FIM, 2013) Take into account a compaction factor: need 1.5 times more sand than the calculated volume Dry proofing 645 $ per house 1 m (2.5 times more expensive than wet-proofing, Botzen et al. 2013 Ann. N.Y. Acad. Sci; Aerts al. 2013; Kreibrich et al. 2009) 24 Results Cost (M$) Benefit/cost ratio Net present value (M$) Benefit/cost ratio Net present value (M$) Year 2100 Scenario RCP4.5 (moderate sea level rise) Ring dike 2 m Additional adaptation strategies Ring dike Dry-proofing Elevation 3 m 1 m Dry-proofing 2 m 197 318 3,340 641 1,281 98 61 8 77 43 19,109 18,988 22,443 48,927 54,122 Year 2100 Scenario RCP8.5 (High-end sea level rise) 55 91 8 69 57 10,648 28,740 17,577 43,457 71,240 (with 2.5% discounting) 25 12
Results Summary Flooding is already very frequent and damaging: Billion(s) $/year By mid- or end of the century intensity and impacts will grow several-fold in the lack of adaptation, with most of the city and critical infrastructure flooded yearly (but airport may be ok) Indirect economic damage could be as much as double the direct damage in the present, and substantially increase in proportion in the future Adaptation can solve most of the impacts: Elevation the most effective on key indicators Cost-benefit analysis suggests all studied measures are economically very efficient 26 Effects of subsidence Subsidence rate in HCMC from Permanent Scatter Interferometric Synthetic Aperture Radar (PSInSAR)(Trung and Tong Minh Dinh, 2009) 27 13
Performance indicators 10/05/2016 Effects of subsidence mitigation measures Subsidence will affect flood risk, but is marginal compared to sea level rise (scenario RCP 4.5 by year 2050) No adaptation Low adaptation High adaptation PRESENT FUTURE Courtesy Timo Maas 29 Effects of subsidence mitigation measures Measures to mitigate subsidence by aquifer recharge 30 Courtesy Timo Maas 14
Indirect economic damage Courtesy Chris Kerklaan 31 Indirect economic damage Indirect/direct damage ratio Flood probability Indirect damages are larger than direct ones They increase with the magnitude of floods Indirect/direct ratio independent of sea level rise Courtesy Chris Kerklaan 32 15
Conclusions Ho Chi Minh city should act quickly to reduce floods, and consider ongoing sea level rise in its plans, to prevent several-fold increase in damage Investment in adaptation seems entirely justified and extremely efficient from the economic perspective Subsidence plays a minor role in flooding, compared to sea level rise Indirect economic damage could be double the direct damage, and increase in the future Institutional/policy structure of the city may be an obstacle to mainstreaming adaptation in the plans Adaptation should also take into consideration equity (who will be protected) 33 Next steps Simulations for year 2050 almost complete Flood Severity Depth h(m) Velocity v(m/s) Overturning parameter v * h(m²/s) Sliding parameter v².h(m) Include fatality impact calculations Negligible severity <0.45 <1.50 <0.50 <1.23 Low severity <0.8 <1.60 <1.00 <1.23 Include compounding effect of river and coastal flooding Coastal, pluvial and river flooding often coincide, with much larger impacts Examine other adaptation measures + combinations Develop adaptation pathways for the 21 st century Medium severity <1.00 <1.88 <1.00 <1.23 High severity >1.00 >1.88 >1.00 >1.23 Extreme severity >1.00 >1.88 >3.00 >1.23 Diaz Loaiza et al.. submitted Adaptation to be simulated: Change land use Manage dams and reservoirs Upgrade drainage system Create retention areas Wet-proof 34 16