Coastal megacities

[pzwsk]

List of coastal cities and their coastal flood hazard level on https://int.thinkhazard.org/

Dhaka, Bangladesh

Guangzhou, China

Bangkok, Thailand

Ho Chi Minh City, Vietnam

Jakarta, Indonesia

[matamadio]

Cross checking against the Fathom data (RP1000 extent).

Dhaka

Almost no flood pixls detected, hence the score reflects the data correctly.

Jakarta

Here the issue is the size of the urban extent, which "dilutes" the hazard and cause it to stay below the 3% area threshold. In fact, the smaller ADM2 units correctly capture the coastal flood risk:

HoChiMinhCity

No flood pixls detected, hence the score reflects the data correctly.

City	Current CF score	RP10/100/500/1000 area flooded (%)	RP10/100/500/1000 mean depth (cm)
Guangzhou	3 / HIGH	3.8 / 7.6 / 11.8 / 14.2	98 / 128 / 164 / 184
Bangkok	2 / MEDIUM	2.7 / 3.5 / 4.2 / 4.4	49 / 51 / 53 / 54
Jakarta	-1 / Not Affected	1.80 / 1.95 / 2.02 / 2.05	60 / 62 / 62 / 63
Dhaka	-1 / Not Affected	0.20 / 0.27 / 0.40 / 0.50	88 / 78 / 68 / 61
Ho Chi Minh City	-1 / Not Affected	0 / 0 / 0 / 0	0 / 0 / 0 / 0

VALIDATION

Checked all five against the underlying Fathom data and the historical record.

City	Result	Historical record	Verdict
Dhaka	-1 (URB)	~200km inland; documented storm-surge impact concentrates on coastal Bangladesh (Barisal, Khulna, Chattogram), not Dhaka	Correct
Guangzhou	3 / HIGH	Ranked #1 globally for present-day coastal flood losses (Hallegatte et al. 2013); documented typhoon storm-surge impact on the Pearl River Delta	Correct
Bangkok	2 / MEDIUM	Less direct typhoon storm-surge exposure than the Bay of Bengal/South China Sea belt, but chronic subsidence (~4cm/yr) and long flood-plain history	Plausible, reasonable middle call
Jakarta	-1 (URB) / 3 HIGH at ADM2 (North Jakarta)	Among the most severe, best-documented coastal/tidal flood cases globally (subsidence up to 25cm/yr in North Jakarta)	ADM2 correct; URB wrong
Ho Chi Minh City	-1 coastal / 3 HIGH fluvial	Documented flooding is tidal-river (Saigon River gauge levels exceeding warning thresholds), not open-coast storm surge	Correct — risk is captured, just under the fluvial category

Four of the five hold up cleanly against ground truth as-is. The fifth (Jakarta) confirms a structural limitation specific to large, multi-district URB extents rather than anything wrong with the underlying flood data or the area-threshold value itself.