Prediction of Riverbank Erosion 2022 (Jamuna, Ganges and Padma Rivers)


Riverbank erosion foreshadows the untold miseries undermining the socioeconomic development of Bangladesh. While the international community is lauding Bangladesh as a rising star due to its rapid economic growth, riverbank erosion is threatening the economy and security of the country. Bank erosion alone has rendered millions homeless and has become a severe social hazard. Due to the magnitude of scale and braided nature, structural interventions to protect against riverbank erosion are very costly. Bangladesh Water Development Board (BWDB) initiated structural measures to reduce loss by riverbank erosion by constructing significant river training structures in the 1990s. These, although have proven somewhat useful, still fall short in providing a solid measure due to the vastness of the problem and associated costs of construction. Thus, along with structural measures, less costly nonstructural measures, like erosion prediction, can be used to reduce the loss due to riverbank erosion and lessen the suffering of people.

Center for Environment and Geographic Information Services (CEGIS), a public trust under the Ministry of Water Resources is continuing its scientific support to the government for progressive development in Disaster Risk Reduction (DRR) regarding riverbank erosion. In the past, CEGIS has developed a unique method/tool for predicting riverbank erosion separately for three different rivers by analyzing time series satellite images. This tool is based on a stochastic formula, using which CEGIS can predict erosion vulnerability for three different probabilities (70%, 50%, and 30%). The 70% probability line represents the high-risk areas while the 50% and 30% probability line provides information about risk and less risk areas.

Since 2004, CEGIS has been predicting the riverbank erosion along the Jamuna, the Ganges and the Padma Rivers based on its developed tool. Prediction is not only limited to the identification of the vulnerable locations along the riverbanks but also provides detailed information on the vulnerable features (land, settlement, road, educational institutions, government & non-government offices, health centers, mosques/temples, etc.) located within the erosion vulnerable locations. The current version of the tool tabulates the vulnerability of different structures in the map with magnitude.

CEGIS predicted the bank erosion along both banks of the Jamuna River in 2004 and the morphological changes in 2005 and 2006 of the same river under the framework of the Environmental Monitoring Information Network (EMIN) project. In 2005, CEGIS applied the tools for predicting bank erosion and morphological changes along the Padma River. In 2007, it predicted erosion and morphological changes of the Jamuna, the Ganges and the Padma under the framework of EMIN and Jamuna-Meghna River Erosion Mitigation Projects (JMREMP). In the later year up till 2018, erosion prediction efforts were funded by different projects of the Bangladesh Water Development Board (BWDB). Since then riverbank erosion prediction exercises have been exclusively supported by the Research and Development fund of CEGIS.

Like the previous years, CEGIS has predicted seventeen (17) probable vulnerable locations for 2022 along both banks of the Jamuna, the Ganges, and the Padma rivers. Among these nine (9) locations are in the Jamuna River; six (6) are in the Ganges and two (2) are in the Padma River. The prediction is not only limited to the identification of the vulnerable location but also provides information on the vulnerability of the land, settlement, and other physical infrastructures of the predicted locations.

The riverbank erosion is predicted to occur in twelve (12) districts situated along the banks of the major rivers. These are Kurigram, Jamalpur, Gaibandha, Sirajganj, Tangail, Manikganj, Pabna, Kushtia, Rajbari, Rajshahi, Faridpur, and Madaripur.

CEGIS regularly monitors erosion prediction to assess the robustness of the prediction tool. Satellite images of early 2022 were used to evaluate the predictions made in 2021. It was observed that erosion prediction made in the previous year showed a good match with the occurrences although there were few locations of slight riverbank erosion where erosion was not predicted. However, the overall accuracy of this prediction tool is approximately 70-80%.

The results of the applied methods are in good agreement with the actual erosion in the field. In 2007, CEGIS disseminated erosion prediction to the community on a pilot basis at four locations along the Jamuna and the Padma Rivers. The community took the prediction positively and used it to save their properties. The predictions have proved to be useful means of reducing the suffering of the erosion vulnerable people, which instigated UNDP to take initiatives to include the erosion vulnerable people in the Safety Net Programme utilizing the experiences of CEGIS. It is to be mentioned here that BRAC disseminated the erosion prediction results to the local community to reduce the loss of properties and suffering of people during the period from 2016 to 2018. CARE Bangladesh also disseminated erosion prediction results to the local community living along the banks of the Jamuna River in 2020.