Spatial analysis of multifractal spectra of the MERRA II meteorological time series

Abstract

The meteorological time series from the Modern Era Retrospective-Analysis for Research and Applications (MERRA) were analyzed using the Multifractal Detrended Fluctuation Analysis (MF-DFA) method for period 1979-2015 and in 248 grid points covering uniformly the Poland territory. MERRA combines observations distributed irregularly in space and time with an unchanging model and analysis system spanning the historical data record into a spatially complete gridded meteorological dataset (Rienecker et al., 2011). On the other hand the multifractal analysis is a powerful method to characterize long-range correlations within the time series through calculation of different scaling exponents for different parts of the series. The MF-DFA was widely used to analyze and compare the multifractality of recorded time series in various contexts, e.g. climate dynamics or aggregation effects (Baranowski et al. 2015, Hoffmann et al. 2017). Here, the MF-DFA is used for gridded daily air temperature, wind speed, wind direction and atmospheric pressure data and the objective of this study was to: a) verify whether, and to what extent multifractality occurs in the time series of meteorological variables from MERRA II data; b) compare the singularity spectra of the time series coming from different grid points spanned over Poland territory; c) analyze spatial patterns of multifractal properties and explain their similarities to the orographic features using geostatistical methods. The results show that MERRA II meteorological variables exhibit specific multifractal properties and spatial anisotropy.