B1678
Title: Trend methods in time series: Comparison and application within the 14th sustainable development goal
Authors: M Rosario Ramos - FCiencias.ID Associacao para a Investigacao e Desenvolvimento de Ciencias (Portugal) [presenting]
Clara Cordeiro - DM-FCT, Universidade do Algarve and CEAUL (Portugal)
Abstract: Research into the analysis of trends in time series remains of considerable interest and relevance, taking advantage of computational tools and the diversity of data sources, such as time series of environmental variables. Therefore, following previous research, a comparative study of methods for detecting monotonic trends in a time series is carried out through a simulation study. Parametric and non-parametric tests are considered, such as tests on the slope and Mann-Kendall test, under several scenarios of autocorrelation, among other characteristics. In the first step, the seasonal effect is removed, using more than one method. To improve the power of the tests, a modification is applied, and resampling is used, such as Bootstrap, when relevant. Data is generated with a behaviour similar to real-time series from sea variables and marine resources with the aim of contributing to the 14th Sustainable Development Goal (SDG) - conserve and sustainably use the oceans, seas and marine resources for sustainable development.
