In conversation with nature

A study from Bihar shows that official flood prediction systems are inadequate and can be improved by integrating local knowledge into it.
The Gandaki, as it enters India from Nepal. (John Pavelka - Flickr: Looking Down the Valley from Kagbeni)
The Gandaki, as it enters India from Nepal. (John Pavelka - Flickr: Looking Down the Valley from Kagbeni)

“There is a special type of black ant that is visible just before and during heavy rains. They start coming out in large numbers with their eggs in their mouths and only travel in a straight line, like a railway track,” says Chandrika Mahato when asked how he predicts rains and floods. Like a number of people living in the floodplains of the Gandaki river in the West Champaran district of Bihar, Mahato is a keen observer of nature and can predict weather events. The Gandaki, a highly flood-prone river, originates in Nepal and flows through the state of Bihar.

This working paper, The river itself warns us: Local knowledge of flood forecasting in the Gandaki river basin, West Champaran, India published by HI-AWARE informs that people living in the floodplains of the river are constantly exposed to the wrath of the river and have acquired precise knowledge to gauge the signs of impending floods and how to cope with them. While the existing, official flood prediction systems are inadequate, the conventional knowledge of the local people is not utilised fully. This knowledge still remains undocumented and often ignored.

The paper discusses the findings of a study that documents the methods used by local communities of the Jogapatti and Nautan blocks of West Champaran district in Bihar to forecast floods and rainfall and its relevance in designing a system to accurately predict the weather.

Watching out for signs of floods

Local men and women identify and pick up very subtle signs of impending floods from the behaviour of the river and its surrounding environment.

Phenomenological--related to human sensations

Sounds that feet make in the water, sensations of excess humidity and heat, etc are used to judge the arrival of rains and floods. Halla or noise is used to warn about the rising water levels in the village. When flood approaches a house, the residents shout out warnings to the rest of the village. While shouting, they also send out details of the houses, the grain storage structures and the cattle affected by the flood to give an idea of the nature and the intensity of the flood. The sound of the mud and bamboo grain storage structures collapsing also help locals to judge the intensity of the flood and the damage done.

Ecological--based on the behavioural changes of animals and insects

Keen observation helps locals identify changes in nature. The sudden emergence of insects like kutki and bhuchain in large numbers, intensified singing of an insect known as jhingur, the sudden proliferation of mosquitoes and dragonflies or damselflies, croaking of peela mendak (yellow-bellied frog), are all indicative of imminent heavy rains and floods.

Riverine--based on the behaviour of the river

Water levels are watched and measured rigorously during monsoons. For this, a khuta (a long piece of wood) or a laga (a long bamboo pole) is planted on the riverbed near the bank. People who have farms near the bank keep an eye on the nishan (mark of the water level). Any rise in the level of the water is communicated to everyone in the market. This, combined with information from the radio on water release from the Gandaki barrage upstream in Nepal, gives them an idea of the flood situation. The appearance of the river also gives out signs. For example, the loss of transparency or a change in colour of the water to ochre or yellow indicates imminent floods. The behaviour of fish such as baikhi is another indicator--if it jumps to enter the smaller streams when they get close to clearer or fresh water, it indicates imminent floods.

Meteorological--based on wind and cloud patterns

Rainfall patterns are also gauged by looking at the colour, shape and direction of clouds. Light coloured clouds are thought to bring rain.

Celestial--based on rain-related constellations

Constellations are also observed for predicting rains and floods.

Government indicators to predict floods

The government method of weather prediction is top-down, bureaucratic and lacks coordination between different departments. Although radio and newspapers track water releases from the barrage, it is unclear how information travels to a radio station, newspapers, and government officials. At the district level, information reaches the district magistrate who passes it onto the block development officer (BDO) and the district superintendent of police (DSP).

The BDO shares this information with the village headman and others through cell phones. The person in charge at the block police station receives information from the district police headquarters who alerts the village chowkidar on his mobile phone.

The Bihar Water Resources Department also provides information to the village, which they get directly from the barrage. The junior engineer (JE) informs the local embankment chowkidar to alert people. Announcements are also made from the local police station and a police constable goes around the villages alerting people. The embankment chowkidar also shouts out an alert when he receives information of a possible flood situation from the JE.

This information has many limitations in terms of quality, the level of detail and specificity of data.

Way forward

Locals have developed a system of weather forecast by combining official indicators with their own knowledge of the ecological, riverine, and other indicators. They also use information from television, mobile phones and newspaper to their advantage. This makes the system of prediction better in terms of quality and level of detail. Centralised official systems of flood forecasting, on the contrary, have a number of limitations and often isolate themselves from local knowledge.

The paper ends by saying that it is time local knowledge is given due credit. Better communication and coordination between flood forecasting scientists and local knowledge holders can go a long way in generating information and designing coping strategies for people living in flood-prone areas.

A copy of the paper can be downloaded from below: