Statistical modelling of agrometeorological time series by exponential smoothing

Abstract

A b s t r a c t. 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.

Primena Tehnika Uglačavanja Kod Referentne Evapotranspiracije U Srbiji

“…ET is a process which depends on factors such as air temperature, relative humidity, solar radiation, wind speed and geographical position of place [2]. Plenty of authors have used different smoothing techniques for forecasting time series [1,[3][4][5][6]. In order to determine the most suitable models to generate forecast Murat et al (2016) compared several exponential smoothing models on the data of air temperature, precipitation and wind speed from Jokioinen (Finland), Dikopshof (Germany), Lleida (Spain) and Lublin (Poland) [3].…”

Model prediction and climatology of aerosol optical depth (τ550) and angstrom exponent (α470-660) over three aerosol robotic network stations in Sub-Saharan Africa using moderate resolution imaging spectroradiometer data

“…The RMSE is the square root of the average squared values of the differences between the forecast and the corresponding observation. Those errors have the same units of measurement and depend on the units in which the data are measured [34]. Figures 1 -3 show the validation plots of τMODIS against τAERONET for the study area using both retrievals, including the seasonal validation.…”

“…They all have good coefficient of determination (R 2 is between 0.690 -0.935) while their respective values of standard deviation (SD) is small and fairly constant (SD between 0.130 and 0.145). The deviation from unity of the slope of correlation plot represents systematic biases and are mainly due to aerosol model assumptions, instrument calibration or the choice of the lowest 20-50 percentile of the measurements [15,34]. Table 2 shows the comparison of slope and R 2 values for different sites in the study area for the two satellite sensors and the combined data set.…”

Dynamics of meteorological time series on the base of ground measurements and retrospective data from MERRA‐2 for Poland

“…Long‐term meteorological data (usually longer than 30 years) are often analysed for studies concerning climate change impacts on various aspects of human living conditions (Pirttioja et al ., 2015; Fronzek et al ., 2018; Ruiz‐Ramos et al ., 2018; Rodríguez et al ., 2019). Existing historical observation data show spatio‐temporal irregularities (Vose et al ., 2012; Murat et al ., 2016, 2018; Stott et al ., 2016; Gos et al ., 2020) and were obtained with the use of a variety of instruments and measurement techniques which have been modified over time (Jones and Wigley, 2010). Additionally, inconsistencies in ground‐level meteorological data are present in long time series due to issues with instrument relocations (Brohan et al ., 2006), urbanization (Peterson et al ., 1998), and land‐use changes (Montandon et al ., 2011).…”