With the increasing availability and advancements of mobile technologies with internet, it has become possible to quickly catch up with latest news around the world and share information through social media such as Facebook and Twitter. Especially during and after a disaster, the information obtained from citizens can be helpful and valuable for local authorities in identifying the extent, intensity and impacts of the hazardous event. Application-driven data collection techniques using mobile devices and social media allowed for the collection of citizens-based information in real time.

Are you aware that you could even contribute for better management of natural hazards (e.g. in the case of floods) through your tweets? In the recent years, researchers have developed approaches for real-time flood mapping using the social media network such as Twitter, enabling citizens to report the information related to flood conditions.

One of such examples is the work of Deltares (https://www.deltares.nl) and Floodtags (https://www.floodtags.com/); which an independent research institute and a social enterprise based in The Netherlands for water management.  They have studied how Twitter data could be used for disaster response with a pilot case in Jakarta (Indonesia), where floods pose a continuous thread to citizens every year. Taking advantage of citizens’ activeness on social media, in particular on Twitter, a concept was developed which combined observed information (i.e. depth and location of floods) with a Digital Elevation Model (DEM) to produce flood extent maps in almost real-time [1].

Another example is an open source platform called PetaJakarta.org, which allows citizens to collect, disseminate and share real-time flood information. It is a research project led by the University of Wollongong in collaboration with the Jakarta Emergency Management Agency and Twitter Inc. This project proved the use of social media and free and open source software (FOSS) for crowd-sourcing of relevant information for decision-making and response coordination. During the pilot study period, citizens reported locations of floods events using Twitter and this contributed to achieve a publicly accessible real-time flood map at PetaJakarta.org. The reported data were cross-validated for flood assessment and management in real-time [2]. A map showing geo-located tweets (from the Twitter #DataGrant) related to flooding over the monsoon season (2013/2014) is illustrated in Figure 1.Twitter box

Figure 1. Geo-located tweets related to flooding in Jakarta for the monsoon season (2013/2014). (Source: PetaJakarta.org)

Another promising example is the research work of German researchers. In this study, they proposed a method which leverages contents of social media for rapid flood mapping. For this purpose, a tool called “PostDistiller” was developed for filtering geo-located posts with photo links (Figure 2). An inundation map was manually created based on information obtained from photos shared through Twitter and Flickr. This developed approach was applied for rapid mapping during the June 2013 flood in the city of Dresden, Germany. It was proven that flood depth estimates could be derived within 3 to 4 hours (in the case of Dresden). Therefore, compared to traditional data sources such as remote sensing satellite data, achieved results are encouraging as social media can provide data more rapidly for assessment [3].  

image map

Figure 2 Location of useful photos retrieved with PostDistiller and inundation depths estimates. Photos by Denny Tumlirsch (@Flitzpatrick), @ubahnverleih, Sven Wernicke (@SvenWernicke) and Leo Käßner (@leokaesner). (Source: [3])

These studies have shown that social media play an important role and contain useful information for flood management, while the availability of reliable data could be a limitation, particularly in places where social media is not frequently used. Besides, it also provides an alternative way for dealing with lack of traditional data sources in certain places. On a positive note, these social media tools provide an alternative way for dealing in places where there is lack of baseline data with increasing accessibility and availability of mobile devices with internet nowadays, we could foresee that there is quite a high potential of applying social media such as Twitter for real-time flood mapping in near future.

The author @AyeZarChi is a PhD candidate working in the field of risk management applying geo-information tools and writes on disasters and information technology related topics in this blog.


[1] http://media.egu.eu/media/documents/2015/25/eilander_scientific_abstract.pdf

[2] https://petajakarta.org/banjir/en/

[3] http://www.nat-hazards-earth-syst-sci.net/15/2725/2015/nhess-15-2725-2015.pdf