Saving lives from riverine floods— New possibilities with technology

New Update
Kerala Floods

By: Vasanth kumar, Head of SAP Co-Innovation Labs, South

Asia, ANZ and Middle East


According to a World Bank study, there are about 4,400 natural flood events recorded from 1900-2015 across the 5 continents, consisting of coastal, flash, and riverine floods. Of these 57% constitute riverine floods. About 1,800 events have occurred in Asia which is the highest amongst the 5 continents, affecting 3.5 bn people and causing death of 6.8 mn people, which is 98% of deaths globally. Hence, the seriousness to tackle the flood-based disaster rests more with Asian countries.


In India, floods and droughts affect vast areas of the country, transcending state boundaries. As per XII Plan Working Group Report, total flood affected area in the country is about 50 mn Ha, out of which 24 mn Ha lies in Ganga Basin States, ie, nearly 50% of flood affected area in the country lies in Ganga Basin States. What could then be the steps to mitigate a natural disaster caused by riverine floods? What approach is required to be in tune with the technological times in which we live? Can we do something to live up to the old adage? Is prevention better than cure? The root cause for this preventable disaster is absence of early warning system and absence of responsible and active disaster management of monitoring system.



In India, central government departments or organizations like CWC, IMD, etc, dealing with floods are requested to furnish flood data (information on flood and flood damages during the monsoon season).

While the CWC takes into account the real time water level rise in rivers, the run-off factors, storage capacity in the barrages, and history to give flood alerts three days in advance, this is far from real-time and still not effective to save lives.


The existing ways to tackle flood is more archaic, which poses innumerable challenges, ranging from information repositories stored in silos with various central and state agencies, extreme dependence on manual process for collection, complexity in dealing with static reports stored in Google drive (the reservoir basin weekly bulletin from CWC), which lacks an integrated approach to gain insights and lack of ability to take action in real-time to mitigate risks during harsh flood environments. Although it is a fact that end-to-end Early Warning Systems (EWS) reduced death tolls as reported by UN Economic and Social Commission for Asia and the Pacific (UN ESCAP), this type of EWS is widely used to predict the movement and intensity of cyclones.

However, early warning flood detection and prediction is considered expensive and computationally complex. Some case studies, research publications based on Embedded Networked Sensor Systems (SenSys) exist. However, to my understanding, the ability to collect, monitor, and provide real-time analysis of the river basins and reservoirs usage of robust adaptive distributed algorithms for flood warning for covering a large geographic area such as a river basin is still pursued as a research topic.



The system to be designed has to take into consideration

the unique characteristics of the state and local departments responsible for flood management operations in India. What makes creating such an early warning flood monitoring and forecast system challenging, is to leverage the sensor networks and the power of real-time computational capability with the latest in-memory technology.

In a first such attempt, a system has been built recently with inputs from state irrigation department with unique abilities to tackle the typical needs of a large river basin in India. With the support of state irrigation authorities, field trials, and implementation of the system will be taken up in one of the major river basins in India. This real-time flood EWS, made in India solution, uses latest world-class technology and has the potential to avoid loss of property and lives by accurate monitoring of flood situations and provide clear and timely distribution of warnings to all those at risk and build the resilience and response capability both at the national and local levels.



Although more work is required to implement the river monitoring sensor network system and invest in developing multiple linear regression models that predicts river flooding with inputs from several types of local data collected from space as well as ground surveys, the flood EWS is the first step towards putting a system in place to monitor the most flood prone river basins.

Such a flood monitoring and forecasting solution would soon become a reality to enable state irrigation departments to centrally monitor, analyze, and forecast flood situations and help save lives of many thousands who perish during riverine floods.

However, the implementation depends on the budget allocation for disaster management solutions and would typically involve coordination of multiple institutions under the central and state governments like MoWR, CWC, IMD, GFCC, NDMA, SDMA.

sap world-bank riverine-floods flood-management vasanth-kumar