Time Decomposition and Short-Term Forecasting of Hydrometeorological Conditions in the South Baltic Coastal Zone of Poland

Abstract

This article presents an analysis of time-series for hydrometeorological conditions determining the behavior of the natural environment in the South Baltic coastal zone of Poland. The analysis is based on monthly data for average air temperature, total atmospheric precipitation, and average sea level during the 50-year period from 1966–2015 for three coastal stations in Hel, Ustka, and Świnoujście. Time decomposition of these hydrometeorological conditions and formulation of short-term forecasts were carried out using ARIMA modelling. This study identifies the seasonal and non-seasonal parameters that determine both current and future hydrometeorological conditions. Moreover, it indicates the spatial differences among features of the analyzed time-series, estimated parameters of the selected models, and forecasts. The ARIMA models used for the Polish Baltic coastal zone are somewhat spatially homogenous. This is especially true of the models for average monthly air temperature, which are identical across the entire coastal zone (2,0,1)(2,1,0)12. Very similar are the models for average monthly sea level across the central and west coast (1,0,0)(1,1,0)12. The model for the east coast, however, was determined to be slightly different (2,0,2)(2,1,0)12. In contrast to those for air temperature and sea level, the models used for atmospheric precipitation were different for each site. Among the parameters modelled, the effect of AR(p) processes was greater than that of MA(q) processes. The monthly models for Ustka are an example of this: average air temperature (2,0,1)(2,1,0)12, atmospheric precipitation (0,0,3)(2,1,0)12, and average sea level (1,0,0)(1,1,0)12. Time decomposition of extreme hydrometeorological conditions has an important utilitarian significance. The climate of the Polish Baltic coastal zone is getting warmer, the sea level is rising, and the frequency of extreme hydrometeorological events is increasing. Time decomposition of hydrometeorological conditions based on monthly data did not reveal long-term trends. In the last half-century, hydrometeorological conditions have been conducive to erosion of coastal dunes and cliffs. These factors determine changes in the natural environment and limit the development potential of the coastal zone. The time decomposition, modelling, and forecasting of hydrometeorological conditions are thus very important for many areas of human activity, especially those related to management, protection, and development of the coast.