Frost safety margins: a proposal based on daily minimum temperature

The concept of frost safety margins has been used in previous research as an indicator associated with the period between the occurrence of the meteorological phenomenon and the most sensitive phenological stage to freeze damage in deciduous perennial plants in mid and high latitudes. It is a relevant indicator in agriculture due to its use in agricultural planning and territorial management of productive areas, as it has been linked to damages from late frosts and as a fundamental criterion for false springs events in fruit trees. However, its implementation requires robust phenological records or experimentation under controlled conditions. The aim of this study is to propose a concept of frost safety margins based on daily minimum temperature data (TMIN). To this end, we used TMIN in raster format at a spatial resolution of 1 km² for the Aragón region (northeast Spain). Calculations were performed in the R programming language and the cartography was built using ArcMap 10.7.1® software. The indicator represents a state composed of two elements: a numerical one, without hierarchical order, which incorporates the moment of occurrence; and a character one associated with the difference in days between the penultimate and last frost per year. Once the states were classified, an annual sequence was formed for the historical series and analyzed as a Markov chain. The results show that the margins are persistent and their autocorrelation with lag-1 is statistically significant, especially with increasing altitude. It was also found that, with a greater number of types of margins, the forecast using Markov chains is lower, indicating a more random behavior. In much of the Aragonese territory, the stationary probability of safety margins corresponds to type 4a, i.e., a margin during the month of April with a difference between 1 to 7 days, while in the lowlands it corresponds to type 2a, a margin during the month of March with a difference between 1 to 7 days. The proposed safety margins indicator can be used in conjunction with empirical or modeled phenological data to obtain agroclimatic risk, making it a line of research that allows for expansion with other future research lines.