Development of a heavy snowfall alarm model using a Markov chain for disaster prevention to greenhouses

Global climate change has, in recent years, increased the frequency and intensity of meteorological disasters. In particular, damage to greenhouses caused by heavy snowfall continues to occur on farms in South Korea. Because heavy snowfall occurs over a relatively long time compared to other sudden meteorological disasters, if an appropriate warning system for heavy snowfall events is in place then damage to greenhouses can be prevented. However, the existing snowfall warning system in South Korea consists only of a simple alert that is unable to anticipate future risks to an area caused by heavy snowfall. In this study, the aim was to develop a stochastic alarm model to minimise the damage to greenhouses caused by heavy snowfall events. A Markov chain was designed to construct the alarm model using snowfall data from nationwide weather stations as well as from greenhouse standards with various safety criteria. The snow depth data were divided into several sections using a cluster analysis, and the failure probabilities of the greenhouses were derived for specific time interval according to current snow depth. This method considers the weather characteristics of each region as well as various greenhouse standards because the risks of heavy snowfall vary depending on location and type of the installation. Using the alarm model developed in this study, it is possible to predict and therefore manage the negative impacts of heavy snowfall events on greenhouses.