21st January - 09:30 (Link: https://videoconf-colibri.zoom.us/j/87292833142)
Presentation of the Statistics field as a part of the Profile in Actuarial Sciences, Statistics and Operation Research of the MSc in Mathematics and Applications - Isabel Gomes (DM / FCT NOVA, MSc coordinator)This talk will give an example on the use of Statistics applied to Ecology. Understanding the real impact posed by wind farms, regarding avian and bat populations, implies mortality estimation. Mortality assessment is based on counting detected carcasses in the wind farm. There are several sources of uncertainty in estimating real mortality, which include carcass removal and the observers’ detection ability. In addition, the spatial distribution of the dead animals is known to be highly dependent on the wind turbines locations. In this presentation we discuss some options to account for non-uniform density of carcasses around the turbines, while estimating mortality.
Portugal wildfires from the perspective of extreme value theory - Nathalia RibeiroEvery year, thousands of forest areas are lost in southern European countries due to wild land fires, causing social, environmental and economic damages. Given this situation, it is very important that governments are always looking for preventive solutions to minimize the impacts caused by these natural phenomena. Through the extreme values theory, the burned areas that occurred between January 1, 1980 and December 31, 2019 in mainland Portugal were modeled based on the block maxima approach, the r largests annual observations per block and peaks over the threshold. The daily burned areas of mainland Portugal were also modeled using the peaks over the threshold approach, in their entirety and regionalized in 8 regions considered homogeneous with regard to their incidence of fires, area burned and geographical characterization. Subsequently, the runs and intervals declustering methods were applied in each region for the daily burned areas and summer months daily burned areas, to filter the excesses considered dependents. The data analysis showed that the occurrences of extreme fires are more likely to happen in the summer months, it was also evident that a small part of the fires that occurred in the period was responsible for the largest part of the burned area. Through the application of statistics of extremes, inferences about fires were obtained, such as exceedance probabilities and return levels for annual return periods. Based on the geographic stratification, it was possible to demonstrate the differences between the daily burned areas distributions for each region, although they all fitted to heavy-tailed distributions. The declustering method proved to be a necessary improvement, due to the fact that the excesses have a tendency to cluster in stationary series and, in case of fires, these generate extremely high and unrealistically burned areas.
__________________________________________________
Regina BispoRegina Bispo has MSc in Probability and Statistics (Faculty of Sciences, Lisbon University, Portugal) and a PhD in Probability and Statistics (Faculty of Sciences, University of Lisbon, Portugal), with a specialization on Experimental Statistics and Statistical Data Analysis. She is currently an Assistant Professor at Departamento de Matemática of FCT NOVA and a researcher at CMA- Centro de Matemática e Aplicações, FCT NOVA, Portugal. Her main research areas include applied statistics in ecology and marine biology.
Nathalia RibeiroNathalia Ribeiro is graduated in actuarial sciences in 2015 and a master student in Mathematics and Applications in the field of Actuarial Science, Statistics and Operations Research at Universidade Nova de Lisboa. During bachelor, she worked for 3 brazilian pension funds and after her graduation she worked as an actuarial consultant for several cooperatives at Unimed Group, the biggest health cooperative group in the world. In 2018 she moved to Portugal to do her master degree. Currently, she still lives in Portugal and works at Future Healthcare Group as an actuary and team leader of the Technical, Risk and Actuarial Department.