Impact forecasting for Red Cross in Tanzania

The Tanzanian Red Cross Society, Deltares and FloodTags analysed publicly available online media (including Twitter, Jamiiforums and various online news websites) for early action and flood response in Dar Es Salaam. With this the Red Cross can 1) map and monitor flood related events in real-time (sign-up and request access here) and 2) use historic event data for future impact forecasting. The results are integrated in a mobile website here. FloodTags is currently fully operational at the Tanzanian Red Cross.

“This tool provides us with an online way of collecting information on flooding. Just by looking at it from our desk, we can see the level of damage, the water levels, people that call out for help harvested from the different media considered in the tool”

 - Renatus Mkaruka, Head of Disaster Management at the Tanzania Red Cross Society -

Hereunder we explain the tools one by one:

FloodTags Dashboard

The first tool enables the Tanzania Red Cross Society (TRCS) to monitor and analyze flood information from a variety of online media sources. It uses flood related data from international sources such as Twitter, but also draws from local sources of data, being the flood information shared by Red Cross members and volunteers through WhatsApp and the Tanzanian social medium JamiiForums. The system monitors and analyzes the data, and distributes the analysis results through an online dashboard (see below figure).

In case a flood event is detected (i.e. thresholds are exceeded indicating a flood), users who have subscribed to alert emails, will receive a notification in their inbox. They can then go to the online dashboard, and review all incoming flood information in real-time. In addition, this dashboard can be used to analyze historic online flood information as well. Users can select data based on publishing date and source of the data. Also, they can filter information by refining it to information about locations, searching by specific keywords, or only showing data from areas where heavy rainfall was observed. In summary, this dashboard allows users to quickly get an overview of all available online flood information, and analyze this information in detail. As a result, with the FloodTags Dashboard, the Tanzanian Red Cross is able to

  • Detect the onset of floods sooner (seconds to minutes)
  • Get a good overview of situiational awareness during an event on the basis of social media posts and shared messages from TRCS staff and volunteers (and act accordingly)
  • Use that data as impact information for fund requests to donors, like the disaster relief emergency fund (DREF).

Figure: Screenshot of the FloodTags Dashboard clearly showing the floods in Dar Es Salaam end of October

Impact Forecasting App

While the first tool provides the TRCS information about what is going on in real-time, the second tool is aimed at providing forecasts on the basis of historical infomation. Therefore we analysed 140.000 online news articles and detected 175 individual flood events in Tanzania over the past 10 years. By analysing those historic flood and water scarcity data together with data about hydrologic variables such as precipitation and discharge, relations between the two were  established and hydrologic variables can be directly translated to expected impact. This now enables the TRCS to forecast the expected impact of a flood event, based on the severe weather alerts they receive from the Tanzania Meteorological Agency (TMA): In the online tool, the user fills in the province of interest, the current condition of the area, and the forecast amount of rainfall. The tool will then search for similar situations in the past, and calculate the expected impact from news articles about this past situation. This provides impact estimates to the user, such as the number of casualties of an event or people evacuated. Additionally, users of the tool can follow the links to the news articles, to get an in-depth overview of the past event.

Figure: Screenshots of the Impact Forecasting App