Flash flood in the mountainous region of Rio de Janeiro state (Brazil) in 2011: part I—calibration watershed through hydrological SMAP model

Abstract In January 2011, the Mountainous Region of Rio de Janeiro state underwent one of the largest natural hazards in Brazil, affecting more than 300,000 people, causing more than 900 dead and irreparable human losses and enormous economic losses. From a hydrometeorological point of view, for the prevention of this type of disaster, alert systems are used worldwide, seeking to anticipate these phenomena by issuing warnings to the population, using tools such as hydrometeorological monitoring and modeling. This study evaluated the applicability of the hydrological soil moisture accounting procedure (SMAP) model with daily discretization as a flood prevention tool in Brazil. It was done by adjusting sub-watershed parameters in the semiautomatic calibration process, inserting observed data of precipitation and stream flow, as well as evapotranspiration in the calibration of the model, between 2008 and 2010. The model was validated twice, between the years of 2011 and 2013, and during the disaster period. The results show that the efficiency coefficient reached in the calibration stage was 1.713, with a relative error of 12% between the calculated and the observed stream flow, and in the validation, the results were 1.401 and 28%, respectively. In the validation of the disaster, the calculated stream flow was overestimated, with an average relative error of 53%. Flash flood prevention through the use of the daily SMAP model has shown promise for use in operational centers.