I read an interesting article on the BBC website today about the extensive flooding in Chennai, India. Tamil Nadu has seen huge amounts of rain recently, in the wettest December for 100 years and large parts of the city, including the airport are underwater. I’ve been paying particular attention to this event as I worked for a number of years with companies in Chennai and have many friends there. The article I mentioned fits in with a strange phenomenon I’ve written about before on the blog; the rush to apportion blame for flooding, something I don’t believe happens to the same degree for other natural disasters.
Essentially the article is blaming poorly planned development and a lack of awareness of nature for the flooding, which seems to be a common refrain when discussing adverse flood impacts. What struck me while reading the article was the disconnect between the natural event and the effects of that natural event, which got me thinking about what we even mean (popularly) when we talk about “a flood”.
I think it can be useful to break a flood down into the two components: the event and the effects. Thinking about the event itself, the EU flood directive defines a flood as “a covering by water of land not normally covered by water”. In almost all definitions of a flood there is no mention of impacts, so philosophically a flood could occur even if no-one was there to see it or feel its effects! A flood is not contingent on human impacts, but is a natural event.
By and large, water covering an area of normally dry land is going to have its source in (abnormal) rainfall. Although there is some evidence human activities can have an impact on rainfall patterns, we can normally take the source rainfall as a fixed variable, i.e. there is next to nothing that can be done by humans to manage rainfall rates. When thinking about the raw volume of water within a catchment basin the principle losses from the system are going to be evapotranspiration, and typically these losses will be low during a rainfall event. So what we are left with is an essentially invariable quantity of water (in terms of human impact) delivered to, and within, a catchment for a given storm.
In terms of human impact on whether a flood occurs we are therefore left with our ability to affect timings of delivery of the rainfall into rivers and down the channel network to areas which may or may not be flooded. Our ability to influence the speed of water transfer is largely driven by land use. As most people will be aware, impermeable concreate areas transfer water quickly and efficiently into channels, forested areas tend to do so more slowly (with a spectrum of land use in between). It is possible; in short duration events (as shown in some of my work), to manipulate the speed of water transfer in different parts of a catchment and so affect flood magnitude and duration. However, there is a really important concept in hydrology related to this issue, which is “time to equilibrium”; this is the time taken for a drop of rain falling at the most distant part of the catchment to reach a downstream area we are interested in. IF a rainfall event persists for a length of time approaching the time to equilibrium (e.g. a long duration rainfall event), and if the rainfall rate is sufficient to lead to flooding, then any land use effects of slowing or speeding transfer of water will be irrelevant in terms of whether a flood will occur. Flooding may be quicker, deeper, longer or shorter, depending on human management, but it will occur in that rainfall event. This is important because from a hydrological point of view floods occurring from long duration weather events such as the ones in Chennai or in Somerset a few years ago are inevitable. It is never correct (hydrologically) to say that human impacts caused the floods.
Where human intervention is vital is in the effects of that inevitable flooding upon infrastructure, but more importantly, people. As the article points out, the human interventions can include:
- Developments on high risk areas, where infrastructure and people are exposed to the flooding
- A disconnect between people and the environment such that at-risk groups are ill prepared for flooding and thus the impacts of flooding upon them are more severe.
- A lack of planning on the part of emergency services and governments such that evacuation and mitigation plans are poorly thought out, ineffective or completely absent.
These are the areas which are ripe for critical analysis by government bodies; principally how can they make sure they minimise the number of people and quantity of vital infrastructure that is exposed to flood risk. How can they make sure that individuals and organisations are aware of the risk they are living with and know what to do when a disaster occurs? In the case of Chennai tackling planning laws and the application of them to prevent building in high risk areas and increasing awareness of risk would be really important steps to minimising future flood impacts. I’ve written about this before in Indian flooding, where (in a different hydrological context) the narrative was very similar, focused unhelpfully on stopping floods, rather than mitigating their impacts.
I can only disagree with Nityanand Jayaraman’s concluding statement:
“Clearly, indiscriminate development and shoddy urban planning have led to the floods in India’s fourth most populous city.”
However I think a compelling case could be made for a modification:
Clearly, indiscriminate development and shoddy urban planning have directly led to, or at least exacerbated the impacts and damage from floods in India’s fourth most populous city.
It can actually be damaging to the aim of minimising flood impacts to suggest that floods such as the one in Chennai would be avoidable with better planning, as this can actually hamper raising awareness of living with risk. It’s important we focus on what we can change and do better in order to properly prepare cities around the world for a likely future increasing in flooding through climate change.