Stochastic process of extreme rainless periods in Serbia

Abstract

AbstractThis paper analyses extreme rainless periods for the territory of Serbia in southeast Europe. Treated as meteorological droughts from an agriculture perspective, the number of rainless periods longer than 20 days as well as their distributions is examined for the growing season (April–September) using data from a period of 50 years. The related stochastic process is modelled via the Zelenhasic–Todorovic (ZT) method for 20 evenly distributed locations across Serbia and important results are mapped in a geographic information system (GIS) environment. The principal components of the process are the number of rainless periods following a Poisson distribution, the theoretical exponential distribution of all droughts' duration and the theoretical double exponential distribution of the (yearly) longest drought's duration. These components are significant for agriculture under climate change because frequent and long‐lasting periods without or with very small amounts of rainfall (e.g., 3 mm·m−2·day−1) have a negative impact on crop production and require additional agricultural measures. This study shows that rainless periods long up to 45 days can be expected every 5 years. Extremely long droughts up to 75 days can be expected every 100 years. The saturation effect, visible if the distribution functions of the longest rainless periods are analysed for different return periods (100, 50, 20, 10, 5 and 2 years) at different locations in Serbia, illustrates that droughts of longer than approximately 55 days may occur every 50 years, while shorter droughts (30–33 days) may be expected every second year. Through quantifying the spatial and temporal distribution of rainless periods across Serbia, this work demonstrates that periods without rain are at least equally important as periods with rain. As both are aspects of a unique climate variability process, they influence agriculture and other socio‐economic sectors and have major impacts on the reliability, resilience and vulnerability of production systems, industries and services.