Flood Detection

Develop a web tool, mobile device app or add-on for existing apps or websites that leverages NASA imagery and climate data to illustrate the impacts of our changing Earth in areas of interest to you. Some ideas to explore include:

• Use NASA Earth observations data, social media, smart phones, and Short Message Service (SMS) text phones to collect Earth observations and connect public in local, regional, and national networks to communicate about our changing planet.
• Examine current natural events curated by NASA’s EONET (Earth Observatory Natural Event Tracker) by browsing global historical and near real-time imagery from space.
• Upload images or other data points demonstrating visible observations and how they compare to satellite data. For example, generating early-warning alerts or validating precipitation rates reported from NASA’s Global Precipitation Measurement Mission.
• Integrate NASA imagery with a mobile assistant to allow for dynamic image generation based on a basic request structure.

We are attempting to link the information from the Global Precipitation Measurement database with Twitter trends so that we can identify flooded areas in real time and so predict where the floods will occur next.

REST

Data is retrieved from ... in HDF5 Format. This is parsed using a Java parser and exported into a JSON file and inserted into the database.

Databases

Riak KV Store

The details of the precipitation are inserted into a Riak NoSql database. The records are stored in JSON format with a primary key of timestamp | latitude | longitude. Secondary indexes are added on the latitude and longitude fields individually so that queries can be performed without supplying a date range.

Web Front End & Analysis

R Shiny