Risk of late frost events in viticulture – temperature trends, phenological observations and spatial analysis based on digital terrain models in Franconia/Germany

Abstract

In the course of global warming a general decrease of frost days per year and in spring can be stated. However, in viticulture this matter of fact doesn´t lead to a reduction of the risk of damages to grapevines by late spring frost events. The observed warming of the spring months, especially of April, causes an ever earlier budbreak.

This study aims at analyzing phenological data and determining the frost risk for different study areas in the Franconian wine-growing region by means of geographical information systems (GIS).

The increase of the monthly average temperature of April in Veitshöchheim (Franconia, Germany) is appr. +2.6 K (1991-2012), as obtained from a linear regression analysis. In the same period, the phenological data collected for Müller-Thurgau show a significant earlier budbreak of appr. 12 days. Since the year 2006, quite large annual fluctuations in budbreak can be observed in Veitshöchheim. The potential risk of damages to buds and young shoots of the grapevine by late spring frosts is particularly increased in years with very early budbreak, as occurred for example 2007, 2009 and 2011 in Veitshöchheim. The latest frost event in May 2011 resulted in severe damages to grapevines not only in the Franconian region.

The delineation of areas that are identified to be exposed to an increased frost risk would provide necessary basic information for planning mitigating measures in viticulture. In this study three different methods of defining risk areas based on digital terrain models (DTMs) were compared and first results are presented for the region of Volkach in Franconia. The results have been verified with information of local wine-growers with respect to a survey of damages caused by winter and spring frosts in 2001/2002, and to a new one that is being carried out since January 2013. The first method chosen was the “relative height above channel network” (SAGA GIS), a hydrological tool that can also be used to calculate cold air flow. First results match quite well with endangered areas along the depth lines. However, neither the land use which influences the production of cold air, nor obstacles to cold air movements in terrain like trees or villages are taken into consideration. Secondly, the focal mean statistic tools provided by the Spatial Analyst of ESRI ArcGIS™ have been evaluated. But it seems that this method, extracting valleys and concave parts of the terrain, tends to overestimate endangered areas. A more complex classification based on a score system which includes factors like slope inclination, relative height, length of a valley, land use or obstacles to cold air flow led to improved results. However, it is very work-intensive and time-consuming. The calculation of zones with increased frost risk implies further studies to be applied to the whole Franconian wine-growing region.