Banca de QUALIFICAÇÃO: ALAN VICTOR DOS SANTOS SILVA

Uma banca de QUALIFICAÇÃO de MESTRADO foi cadastrada pelo programa.
STUDENT : ALAN VICTOR DOS SANTOS SILVA
DATE: 30/08/2022
TIME: 09:00
LOCAL: WEB CONFERÊNCIA
TITLE:

ANALYSIS OF AREAS SUSCEPTIBLE TO FLOODING IN THE SÃO MIGUEL RIVER BASIN THROUGH PREDICTIVE MODELS AND COMPUTATIONAL LANGUAGE R

KEY WORDS:

Predictive models; floods; Machine Learning.

PAGES: 40
BIG AREA: Ciências Humanas
AREA: Geografia
SUMMARY:

Studies focused on watersheds are important because they represent open systems with a quick response to changes in the dynamics of the elements, whether internally or in the surrounding systems, and given their socioeconomic relevance to the population, they are excellent territorial management units. Therefore, the need for policies and projects for the use and occupation of the basin must be adapted to its dynamics. Predictive models are widely used in various areas of science, also given the difficulty in obtaining and processing large volumes of data that has been overcome recently with advances in generation, storage and processing capacity, it is also in this scenario that their most contact occurs. closely with the natural sciences, which have to deal with data of a different nature and large volumes. That said, there are already predictive models with the objective of analyzing natural systems and, with their behavior patterns, generate probabilities for future scenarios, which allows us to foresee these scenarios and provide subsidies to avoid or mitigate the damages that may arise. In this work, two models will be used together, the flow model, which from the physical characteristics of the São Miguel River Basin (BHSM) and precipitation data generates probable flow levels, and the data model to predict the possibilities of floods that will be calculated by ANN (Artificial Neural Networks) trained with data from previous flood events of the last 30 years. Finally, the trained ANN will have the data generated by the flow model as input for flow data, thus allowing to predict and quantify the hydrological potential, as well as the probability of flooding events in the watershed.

BANKING MEMBERS:
Externa à Instituição - DANIELE TORRES RODRIGUES - UFPI
Presidente - 1119902 - JOSE VICENTE FERREIRA NETO
Interno - 2501620 - MELCHIOR CARLOS DO NASCIMENTO