Abstract: Agro-meteorological quantities are often in the form of time series, and knowledge about their temporal scaling properties is fundamental for transferring locally measured fluctuations to larger scales and vice versa. However, the scaling analysis of these quantities is complicated due to the presence of localized trends and nonstationarities. The objective of this study was to characterise scaling properties (i.e. statistical self-similarity) of the chosen agro-meteorological quantities through multifractal detrended fluctuation analysis (MFDFA). For this purpose, MFDFA was performedwith 11 322 measured time series (31 yr) of daily air temperature, wind velocity, relative air humidity, global radiation and precipitation from stations located in Finland, Germany, Poland and Spain. The empirical singularity spectra indicated their multifractal structure. The richness of the studied multifractals was evaluated by the width of their spectrum, indicating considerable differences in dynamics and development. In log-log plots of the cumulative distributions of all meteorological parameters the linear functions prevailed for high values of the response, indicating that these distributions were consistent with power-law asymptotic behaviour. Additionally, we investigated the type of multifractality that underlies the q-dependence of the generalized Hurst exponent by analysing the corresponding shuffled and surrogate time series. For most of the studied meteorological parameters, the multifractality is due to different long-range correlations for small and large fluctuations. Only for precipitation does the multifractality result mainly from broad probability function. This feature may be especially valuable for assessing the effect of change in climate dynamics.

Abstract: The most usual records of observable agro-meteorological quantities are in the form of time series and the knowledge about their scaling properties is fundamental for  transferring locally measured fluctuations to larger scales and vice-versa. However, the scaling analysis of these quantities is complicated because of  the presence of localized trends and nonstationarities. The objective of this study was to characterize scaling properties (i.e. statistical self-similarity) of the  chosen agro-meteorological quantities through multifractal detrended fluctuation analysis (MFDFA). The MDFA analysis was performed  for time series of the air temperature, wind velocity and relative air humidity (at the height of 2 m above the  active surface) as well as the soil temperature (at 10 cm depth in the soil). The studied data were hourly interval, 12 years’ time series from the agro-meteorological station in Felin, near Lublin, Poland. The empirical singularity spectra  indicated their multifractal structure. The richness of the studied multifractals was evaluated by the width of their spectrum, indicating their considerable differences in dynamics and development. The log-log plots of the cumulative distributions of all the studied absolute and normalized meteorological parameters tended to linear functions for high values of the response  indicating that these distributions were consistent with the power law asymptotic behaviour. Additionally, we investigated the type of multifractality, that underlies the q-dependence of the generalized Hurst exponent, by analyzing the corresponding shuffled and surrogate time series. For majority of studied quantities, the multifractality was due to different long-range correlation for small and large fluctuations.

Abstract: Meteorological time series are used in modelling agrophysical processes of the soil-plant-atmosphere system which determine plant growth and yield. Additionally, longterm meteorological series are used in climate change scenarios. Such studies often require forecasting or projection of meteorological variables, eg the projection of occurrence of the extreme events. The aim of the article was to determine the most suitable exponential smoothing models to generate forecast using data on air temperature, wind speed, and precipitation time series in Jokioinen (Finland), Dikopshof (Germany), Lleida (Spain), and Lublin (Poland). These series exhibit regular additive seasonality or non-seasonality without any trend, which is confirmed by their autocorrelation functions and partial autocorrelation functions. The most suitable models were indicated by the smallest mean absolute error and the smallest root mean squared error.