Abstract: Recent flooding of river Jhelum in Kashmir caused large scale devastation in the Kashmir valley. Also, cloud burst over Leh in Jammu &Kashmir two years ago, and at Kedarnath temple vicinity in Uttarkhand a year ago caused significant loss of life and property. In addition, a number of tourists were swept away due to opening of a damís flood gates in Himachal Pradesh a few months ago. In this paper, we present an innovative solution to the frequent flooding problem faced in our country year after year. Named 'Integrated Weather and Flood Alerting System', the system is automated flood detection and early warning system designed using Internet of Things technologies. The system uses a novel approach and low-cost sensing devices to measure and analyze multiple weather and river parameters, namely, rainfall in the river basin / catchment area, water flow in the river, and water level at various locations along the river length to accurately predict flooding of the river in real-time. Timely flood alerts are then disseminated simultaneously and in real-time to nearby population, and the general public through multiple delivery mechanisms such as: 1) Audible alerts (using Hooters near human settlements on the river bank), Visual alerts (using flash lights deployed near the river banks), 3) text-based SMS alerts to all public in the vicinity of the river, 4) graphical, map based alerts on a Web-based dashboard for general public to view flood situation in the river, and 5) live alerts on online social media sites such as Facebook and Twitter for the public. The system consists of self-contained sensing and alerting units that can work autonomously when deployed at a river site, in addition to working in a networked environment. The system offers a low-cost alternative to traditional flood alerting mechanisms in use today.

Keywords: Raspberry Pi, IWFAS Field Unit, Twitter, Facebook, Short Message System, Database.