Introduction
================

**Background & Motivation**

The annual average number of extreme weather events has more than doubled globally since 1980. These events can have devastating impacts on communities worldwide, but the diversity of people's experiences is difficult to capture and communicate. Thus, it is imperative to develop and refine approaches for responding to extreme weather events that draw upon all available tools. Social media platforms serve as massive repositories of real-time actionable and situational information during extreme weather events. Consequently, the DSSG Solve team, in collaboration with the International Water Management Institute (IWMI), aims to develop research methodologies that leverage online social media data to identify and better understand the experiences and needs of vulnerable communities affected by natural disasters. Through our goal, we hope to derive actionable insights to supplement on-the-ground efforts and help inform critical decision- and policy-making processes.

For the initial phase, we focus on a recent extreme weather event - Cyclone Amphan, which made landfall in South Asia on May 20, 2020, and affected 18 million people along the way. It was the most damaging storm in the history of the Indian Ocean, rendering hundreds of thousands of people homeless, ravaging agricultural lands and causing billions of dollars in damage. We develop a research tool that enables users -- from governments to donors to development organizations -- to tap into the online discourse around Cyclone Amphan and investigate the cyclone’s social, political, and economic effects.

**Research Focus**

We first aimed to characterize how collective knowledge about Cyclone Amphan was produced on Twitter. Twitter is a decentralized microblogging platform, meaning that anyone from anywhere in the world can add their commentary to an issue — thus influencing its narrative and adding layers of interpretation to on-the-ground realities. After exploring who and what is shaping the narratives around Cyclone Amphan, we aimed to answer: Can Twitter content help identify unmet needs of people affected by Cyclone Amphan? If so, how?