Rain-related Tweets in flooding days: a brief analysis on Tamanduateí river basin

Abstract

Flooding has become an increasingly frequent phenomenon and reason for several economic and human losses worldwide, such as the São Paulo city, Brazil. The literature has shown that Voluntary Geographic Information (VGI) data is a valuable complementary source for monitoring and analyzing hydrological and meteorological events. Face with this motivation, we extracted VGI from the Twitter social network and analyzed the frequency of meteorological-related posts before flooding/non-flooding events and its correlation with physical-meteorological measures. The obtained results indicate that Twitter is a valuable VGI source for monitoring flooding events.