Validação da reanálise do MERRA-2 com dados observados

Abstract

Climate data such as temperature are key to understand a place’s dynamic. Since the classic period to study climate is thirty years, some areas face difficulties due to the lack of broad data record. Thus, one solution to this situation is the employment of reanalysis, like the Modern-Era Retrospective analysis for Research and Applications, Version 2, known as MERRA-2. Considering that the city of Itajaí, located in Santa Catarina, south of Brazil, has one automatic weather station with data since 2010, this study aimed to, statistically, evaluate MERRA-2’s data to fill the twenty-year gap in climate records for this location, regarding temperature, by analysing daily temperature records from 2010 to 2020, through basic statistic tests and other tests like Anderson-Darling normality test, paired t-test, Mann-Whiney, Pearson correlation, Spearman correlation, bias and polynomial regression analysis with Fitted Line Plot, all on Minitab® 18. The comparison was between observed data from A868-Itajaí automatic weather station and MERRA-2’s data. As a result, it was reached that none of the data is normally distributed, all the means and medians are different from each other, however, the observed data and the MERRA-2’s data are highly correlated. That said, this research provides substantial information to reassure the use of reanalysis data to fill the gaps due to the lack of observed temperature data in order to have a broad study when trying to comprehend a local environment.

Keywords

• remote sensing
• reanalysis
• brazilian coastline
• sensoriamento remoto
• reanálise
• litoral brasileiro