Introducing ‘Intelligent (Open) Data Ecosystem for Assam — Flood Response and Management (IDEA-FRM)’

Authors: Kabeer Arora and Sai Krishna

This article is part of CivicDataLab’s initiative Open Contracting India, and the project Intelligent Data Ecosystem Assam — Disaster Risk Reduction.The team members who contributed are Kabeer Arora Gaurav Godhwani Jeeno Soa George Phani Datta Surampudi Sai Krishna Dammalapati shreya agrawal

The potential of data in providing solutions to the challenges presented by climate change is well established now. However, in India, the necessary good quality, machine-readable and interoperable data that can inform strategy and help plan for climate change and disaster response is currently unavailable. Such data is scattered or siloed across different agencies, at different scales and in different formats, making it difficult for the decision-makers and relevant stakeholders to make data-informed decisions. This often results in inefficient processes and policies or ad-hoc responses that fail to adequately cater to urgent, often life-saving needs in times of emergency.

We at CivicDataLab undertook a journey to identify and collate all such relevant datasets which can help in creating data-driven solutions to address the issue of floods in the state of Assam. This was done as a part of a project called ‘Intelligent Data Ecosystem for Assam — Flood Response and Management (IDEA-FRM)’ in collaboration with Open Contracting Partnership and supported by the Patrick J McGovern Foundation.

One big challenge we realised after conducting the scoping study, primary field research and talking to different stakeholders, was that there is no agreed-upon definition of what constitutes ‘climate data’ or in this case ‘flood data’. Depending on the case and stakeholder we engaged with, the definition kept changing from weather and satellite data to flood-related losses and damages data, from data on infrastructure to data on relief money spent. Moreover, these datasets also did not follow set standards and were scattered across different scales, administrative boundaries and jurisdictions.

Through this project, we first mapped the governance framework and related datasets which get produced and used for taking actions for responding to floods. Following this, we collated, standardised and organised the said data in broad categories to create a holistic data ecosystem that can enable various decision-makers, researchers, data scientists etc to experiment and propose data solutions to address the issue of floods and climate change in Assam.

We classified the available datasets into five categories:

1. Satellite and weather-related data: This data informs the factors and extent of floods and can help in developing early warning systems using predictive modelling. It consists of details like precipitation, flood inundation, elevation, slope, river discharge, distance from rivers, land-use characteristics and lithology of the study area. We have shared all the data sources in the repository here: Satellite and weather-related data
2. Demographic data: Datasets in this category help us understand how floods interact with settlements and human lives, we measure demographic vulnerability by studying the population, sex ratio, deprived population, household access to drinking water and sanitation. These datasets were sourced from the Socio-Economic Caste Census, National Family Health Survey-5, Assam Statistical Handbook and National Census. The variable-wise data source is shared in the repository here: Demographic data
3. Access to infrastructure data: This layer evaluates the proximity to physical infrastructure like roads, rail, hospitals and community shelters which becomes critical to understand the coping capacity of settlement in case of a disaster. These datasets are sources from NESAC, GeoSadak and Pradhan Mantri Gram Sadak Yojana. The variable-wise data source is shared in the repository here: Access to infrastructure
4. Past damages data: The Assam State Disaster Management Authority (ASDMA) maintains a robust data system that collects information on damages due to floods in every monsoon season. This is high-frequency (day-level) data at the revenue circle level disseminated using the Flood Reporting and Information Management System (FRIMS). This data includes the population affected by floods, houses damaged, crop area affected, infrastructure affected etc. to understand the severity of flood impact in various settlements in the past. We cleaned all these past damages datasets and shared them in the repository here: Past Damages
5. Government response data: Government response before, during and after a disaster is important while proposing any solution. We looked through the lens of public expenditure, as seen in government procurement data, budget data and relief distributed for flood management to understand the government response. We identified the tenders issued by the state government that are related to flood management and geo-coded them to the revenue circle level to get spatial insights on government response. All these datasets are shared in the repository here: Government Response

We used the above datasets to develop data models to assess flood preparedness levels of different districts in Assam that can assist the authorities in streamlining the funds where they are needed the most. We are now developing an easy-to-use front end for the decision-makers to run the model and consume the insights on a near real-time basis. We believe that the wider community of researchers, entrepreneurs, decision-makers, students, etc., can utilise these datasets for many more use cases, in a much more holistic manner, in service of the people of Assam.

We have published these datasets under an open license. We hope to see their use for innovative solutions by different members of the community. You can access the Integrated Data Ecosystem for Assam- Flood Response and Management (IDEA-FRM)’ datasets here: IDEA-FRM.

Please feel free to contact us at open-contracting@civicdatalab.in for any queries or suggestions.