Making & Using ‘Waves’ in Flood Management: A Data-Driven Approach

TL;DR: As a Chief Minister’s Fellow working in the District Collector Office Bhandara, under the visionary guidance of District Collector IAS Yogesh Kumbhejkar Sir, I developed a graphical model to predict and analyze the flood situation in real time that helped the District Disaster Management Authority (DDMA) in managing sever flood in the district.

Wainganga River in the Bhandara District is prone to heavy flooding during monsoon season. Earlier, floods in 2020 and 2022 have caused heavy damage to people and property. To avoid or minimize such destruction and to evacuate citizens to safer places immediately, the District Disaster Management Authority (DDMA) uses a variety of data to maintain the situation under control. Among such data sets, daily rainfall forecast, data of water entering and leaving the District in different ways, forecasts of the IMD & CWC, mobile applications like Damini, previous year’s analysis etc. are used.

Along with all such imminent efforts, this monsoon season, under the able leadership of District Collector IAS Yogesh Kumbhejkar Sir, we used a simple but innovative approach to build a model to analyze the real time data and to disseminate relevant instructions as per need.? Before understanding that approach and the model, let us understand the geography of the basin of the Wainganga River in the Bhandara District.

Wainganga River enters from North-East side of the district, flows diagonally towards Bhandara town in the western side of the district; it then marks the western border of the district for a few kilometers and then flows towards the South-East side of the district, as shown in the map below (blue dots). The Kardha Bridge on the river is situated in the Bhandara town. Gosi Khurd dam, which is a national project, is situated roughly 40 kms away from the Kardha Bridge in the Southern side. Almost the entire basin of the river in the district till Gosi Khurd dam lies under the flood prone area.

This monsoon season, Bhandara district experienced a severe flood situation during 15th to 17th September 2023. We were continuously getting half hourly data from the field officers. This data included the water levels of Wainganga River at Kardha Bridge and at Gosi Khurd dam. Collector sir came up with the idea to plot these water levels and the rate of their increase i.e. the first differential. In order to control the loss and give timely directions to relevant officers, Collector sir entrusted me the responsibility to develop a graphical model of this idea.? I started to develop a suitable model and after multiple iterations and discussions with Collector sir, I was able to develop a model that clearly helped us to analyze the flood situation in real time.

In the first graph, the red line shows the water level at Gose dam. The blue line shows the water level at Kardha Bridge. Yellow line shows the difference between both the water levels. On 15th September, as soon as we realized that the flood situation may worsen in the coming hours, Collector sir asked the Gose dam officials to open the gates of the dam and discharge the water. We can see the red line going downwards in the graph. Despite the gates of Gose being opened and water being continuously discharged, the water level at Kardha Bridge inevitably went increasing till the evening of 16th September which threatened the lives and property.

The second graph of our model, which contained the first differential of water levels, i.e. the rate of change of water levels, immensely helped us to evacuate people from their places. For example, between 4:30 pm to 5:30 pm on 15th September, the rate of change of water level at Kardha was about 0.5 m/hr. This was an alarming situation and immediate directions were given to the field officer to evacuate people on war footing. This increased water level inevitably destroyed the houses, farms, crops and other infrastructure but the team became successful in saving human lives and animals.

Once the flood water receded, we tried to analyze the graphs and other data sets like daily rainfall, volume of water entering and leaving the District in different ways. Following are some of our major observations from this analysis.

1)?Once the gates of Gose Khurd dam were opened, and free flow of entire river water was allowed, the water level at Kardha should have increased at a slow pace and shouldn’t have crossed the danger mark. But it increased with a very high rate and went almost 1.5 meters above the danger level. The difference between water levels at Gose and Kardha at one point of time reached to 5 meters which was highly alarming. This clearly illustrates the problem of heavy siltation, invasion of unwanted vegetation and unexpected structures (encroachments) in the river basin.

2) The rate of receding water level from the highest water level mark was certainly higher than the rate of increasing water level (clearly visible in the graph). That was actually a good sign.

Henceforth, whenever there will be possibility of the flood in the district, this model will be used to monitor the situation. As a CM Fellow working in the Bhandara District, I got this incredible chance to leverage real-time data and transform it into insightful graphical formats. This not only streamlined the analysis process but also empowered the administration to make informed decisions swiftly. I'm grateful for this opportunity to contribute to community resilience. Let's continue working together for a safer and more prepared future!

?#FloodManagement #DataDrivenDecisions #CommunityResilience

Chief Minister Fellowship Maharashtra, Government of Maharashtra