Assam Floods: Could the Govt Have Predicted the Disaster Using Tech?

Technology can help authorities gain actionable information much before the disaster actually takes place.

An elderly woman rows a makeshift raft to cross a flood-affected area in Assam.

(Photo: PTI)

Assam and a few northeastern states have been witnessing frequent disasters this year, with heavy rains causing widespread disruption and destruction. Hills are facing landslides and slush movements, while rampant flooding across urban and rural areas is impacting work and movement. In a span of two weeks, floods have wreaked havoc in southern Assam, particularly in the Barak valley. Silchar, the second largest city of Assam and a major transportation and transit hub to connect Mizoram, is currently seeing flood waters reach neck-deep levels in almost the whole town for a week now. Users on almost all social media platforms have been highlighting the plight of the city.

While excessive rains have been the cause of floods almost every monsoon in Assam, this time, the breach of an embankment in Bethukandi is responsible for the Barak river water gushing out into the city of Silchar. Out of the many embankments of Barak Valley, the Bethukandi dyke is important as it is responsible for restricting the flow of the water of the Barak River towards the city. River Sonai, which comes from Mizoram, gets merged into River Barak near Bethukandi. During monsoon, Sonai brings in a good flow from the Mizo Hills, which results in increasing the level of Barak.

• Technical advancements can give us a source where we can predict the level of water, get guidance on weather reports and can estimate the quality and measures to prevent soil erosion, a prime cause of flooding.

• In many countries, 3D Mapping, LiDAR, data centres and Drones/Unmanned Aerial Vehicles are used to detect increasing water levels in rivers and predict floods.

• Technologies like high-performance computing (HPC), artificial intelligence (AI) and machine learning (ML) can be applied to optimise our flood response.

• HPC and AI, which mimic the brain's processes, can carry out parallel computations with massive data that can map the approximate model of flooding with almost 36 hours of lead time.

• The government has to look beyond just flood management and focus on a flood prediction approach.

The Impact of Cyclone Asani

Almost all embankments of Barak need proper maintenance, but this has been neglected by successive governments. Credible estimates show that around 3 lakh people are directly affected in the Barak valley in this round of flooding. Similar situations were also seen in 1985 and 2004, but the intensity was lower and the water didn’t stay for as long in the area.

Monsoon in Northeastern states generally comes early as compared to the rest of India. However, this year, it came even sooner. Back in the month of May, the region experienced heavy rainfalls due to the severe cyclonic storm “Asani”, which was part of the 2022 North Indian Ocean cyclone season and resulted in floods and landslides. It cut off connectivity in almost all the eight states in the region. At that time, Silchar and the other two districts of Barak Valley faced a mini flood situation, which left many homeless. The situation improved slightly towards the end of May, and in the early weeks of June, people had just started to return to normalcy.

But then came the real monsoon from the second week of June, and the northeast region again started receiving heavy rainfall. The situation worsened quickly for the Barak Valley.

These rains and the resulting disasters have brought to the fore the need to find the means to address such situations sincerely and urgently. Much of the solutions will require actual infrastructure development on the ground, such as the strengthening of embankments and building rehabilitation accommodation on higher grounds. Along with that, the construction of artificial water gates and water barrier socks has to be undertaken.

Artificial Intelligence, Machine Learning: Can Tech Help?

However, from an early warning and disaster management perspective, technology can be used in an effective way. Technical advancements can give us a source where we can predict the level of water, get proper guidance on weather reports and can fully estimate the quality and measures to prevent soil erosion, which is a prime source of flooding. In many countries abroad, technical measures to manage natural calamities using technical advancements like 3D Mapping, LiDAR, data centres and Drones/Unmanned Aerial Vehicles are used to detect increasing water levels in rivers and help predict floods. They can even keep proper track of the embankments and would give better security control over the various embankments in and around the rivers.

In this context, it will be pertinent to touch upon two areas that the state needs to focus on where today’s emerging technologies can make a difference – flood management and agriculture productivity optimisation, including the tea industry. Both these areas need a focused, long-term approach, and the effort has to start now. Technologies like high-performance computing (HPC), artificial intelligence (AI) and machine learning (ML) can be applied to optimise these sectors. The availability of space technology and its application in the northeast region, such as the mapping of forest gap areas, the expansion of the land area for horticulture development, identification and rejuvenation of wetlands and diversion of floodwater, have already been started by the North Eastern Space Applications Centre (NESAC) at Shillong.

Acting Before the Disaster Takes Place

The government has to look beyond just flood management and focus on a flood prediction approach, based on data from NESAC. This can help the government acquire actionable intelligence in advance, which will help in crafting appropriate response strategies that could range from evacuating the predicted area of impact to sandbagging to arranging for rations for the displaced population.

A flood forecasting system commonly consists of a numerical weather prediction model to provide rainfall forecasts and a hydrological/hydraulic model to predict the hydrological response.

Though widely used for flood forecasting, hydrological models provide only a simplified representation of the physical processes of flooding and cannot reliably predict the highly transient flooding process from intense rainfall, in which case, a fully 2D hydrodynamic model is required. HPC and AI, which mimic the brain's processes, can carry out parallel computations with massive amounts of data that can effectively map the approximate model of the flooding with almost 36 hours of lead time, enough to mount a credible response. Some of the recent cyclone warnings have worked on HPC.

It is important to address this perennial problem with a more pragmatic approach and have solutions effected in a time-bound manner. The heavy flooding at Silchar should push the government to take up flood management in a more coordinated fashion and much before the disaster actually takes place.

(Subimal Bhattacharjee is a commentator on cyber and security issues around Northeast India. He can be reached @subimal on Twitter. This is an opinion piece and the views expressed are the author’s own. The Quint neither endorses nor is responsible for them.)