GeoAI Framework Unveiled to Assess School Flood Vulnerability in Bihar
In a significant development for disaster management, researchers from the Central University of South Bihar (CUSB), in collaboration with Arizona State University in the USA, have pioneered an innovative GeoAI-based framework. This advanced system is specifically designed to assess school exposure to varying flood-depth scenarios in Bihar's highly vulnerable regions. The research has been published in the latest edition of the internationally reputed natural hazards journal, Springer Nature.
Addressing a Critical Need in Flood-Prone Bihar
Bihar faces a severe and persistent flood threat, with a staggering 73% of its land area prone to inundation. Alarmingly, 76% of the population in the northern part of the state lives under the constant threat of annual flooding. The newly developed framework aims to provide a scientific basis for disaster-resilient educational infrastructure planning, offering a crucial tool for policymakers.
According to team leader Somnath Bera, "The proposed framework offers a scientific basis for disaster-resilient educational infrastructure planning. It can assist policymakers in prioritising flood-resilient school construction, retrofitting, and emergency preparedness. The framework is scalable and can be applied to other flood-prone regions across the country."
Historical Context and Vulnerability of Schools
The study underscores north Bihar's long history of devastating floods, with major events recorded in 1973, 1978, 1988, 2007, 2014, 2016, 2019, and 2020. These recurring disasters have repeatedly damaged educational infrastructure. For instance, during the 2007 floods, nearly 1,960 schools in Bihar suffered extensive damage, while 159 schools were completely destroyed in 2008. These events highlight the acute vulnerability of schools across the region.
Floods frequently disrupt educational infrastructure in Bihar, adversely affecting student safety, attendance, and learning continuity. Compounding the issue, schools often serve as emergency shelters during disasters, making it imperative to assess their vulnerability under different flood conditions.
Methodology and Key Findings from Darbhanga Case Study
The research paper, titled "Developing a framework for assessment of the school’s exposure to flood-depth scenarios: a case study of Darbhanga, Bihar, India," was authored by Chandrakanta Bishui, Somnath Bera, Shivam Priyadarshi, Pritha Ghosh, and Kshitij Dahal. Utilizing Google Earth Engine (GEE) and a sophisticated machine learning model, the team generated detailed flood depth susceptibility maps for three categories:
- Flood depths below 1.5 metres
- Flood depths between 1.5 and 3 metres
- Flood depths exceeding 3 metres
A comprehensive spatial analysis of 2,010 schools in Darbhanga district revealed critical insights. The findings indicate that nearly 20% of the district has a high probability of experiencing floods exceeding 3 metres in depth. Under extreme flood scenarios, 43 schools were identified within high-susceptibility zones. The study also highlights that rural schools are comparatively more vulnerable than their urban counterparts.
Implications and Recommendations for Strengthening Resilience
With Darbhanga's literacy rate at 56.56%, repeated flood disruptions place additional strain on the education system. The study proposes several actionable recommendations to mitigate risks:
- Elevating school buildings above predicted flood levels
- Retrofitting high-risk schools with flood-resistant materials
- Improving drainage systems around educational institutions
- Integrating flood exposure maps into district-level planning processes
The GeoAI-based framework is not limited to Darbhanga. It can be scaled across other North Bihar districts such as Madhubani, Sitamarhi, Saharsa, and Muzaffarpur. This scalability offers a pathway to strengthen disaster-resilient education systems throughout the region, potentially safeguarding thousands of students and preserving educational continuity in the face of natural disasters.
