Catastrophic flood in Pakistan once again raises a question: why people build stuff in river's flood basins? (we had the same in 1997 and 2010 in Poland)
Mostly because such floods happen rarely (decades away), and the interest to capitalise on cheap ground is reverse proportional to the declining memories of damage.
So at any given moment you have people *always* questioning government policies on one of these contradictory grounds:
* "why is government not allowing us to use this cheap land because of some flood that happened decades away?!"
* "why hasn't government told us not to use this cheap land when there was a likelihood of flooding?!"
Few people realise that significant parts of many towns are built in flood basins — go check yours, and you'll be surprised. Just like I was when I worked on business continuity plan for my employee in Krakow back in 2010, and found out that most of the historic centre of my home town technically is a flood basin of either Vistula or Rudawa river.
A precautionary approach would be to just leave these without any construction, but then what you do if people built stuff there for hundreds of years already? In reality we're doing a kind of intergenerational gamble here — massive floods usually happen not more frequently than every few decades, so around one or more generations.
You can run a successful business and capitalise on cheaper flood basin land for decades, and never experience a flood. Or you may just have no luck and have your business wiped by a flood.
An "once in a century" flood is merely a layman's term for 1% statistical chance of a large flood in every given year, but it doesn't mean it will happen at century intervals. It only means that after 100 years the likelihood of such flood comes very close to 100% (still, no guarantees it will happen, but very likely). In Krakow, we had these "floods of century" first in 1997 and then in 2010...