The reconstruction of past temperatures with ground ice, especially with ice wedges is possible, but reduced by the missing correlation of ice veins to the exact year of their formation. Therefore, we performed recent cryogenesis tracer experiments on Samoylov Island, Northern Siberia, to calibrate a stable isotope thermometer for ice wedges. In 2002, a low centre polygon (20 m in diameter) was selected showing clearly visible (several mm wide) frost cracks and a well-developed relief between polygon wall and centre. Here, coloured Lycopodium spores, (each year a different colour) have been applied in autumn to a polygon with recent ice-wedge growth (between 2002 and 2009) to trace ground ice formed in the considered years. The spores should to be transported with snow/snowmelt into the open frost crack in spring. In 2010, 13 permafrost cores were taken from the ice-wedge polygon. In the cold laboratory, these ground-ice cores have been processed and individual ice veins have been sampled for spore identification and stable isotope analyses. Single ice veins were melted and examined under the light microscope for spores. All spores colours could be recovered, with red spores (2002) being most abundant followed by green (2003) and violet (2004). An overall decrease in spore abundance is probably related to changing hydrological conditions and polygon degradation. Additionally, frost cracking experiments (breaking cables) were installed to the polygon, intended to break when a sudden rupture (i.e. a frost cracking event) takes place. Data loggers detected the precise moments of cracking on Samoylov Island generally occurring between November, 8 and February, 10 (N=12). The stable isotope composition of every single melted ice vein has been analysed with a Picarro L2120i water isotope analyser and, if spores were present, attributed to the respective year of formation. However, attribution to the year of formation is complicated by eventual occurrence of more than one colour in a sample, which has been assessed by statistical methods. The isotope composition of single ice veins could be measured for the first time in high resolution with a Picarro laser-optical spectrometer and showed no significant isotope fractionation during freezing. The combination with meteorological data allows correlating the local temperature with the δ18O of ice-wedge ice formed in a discrete year. The δ18O and δD values of recent ice veins correspond roughly to the LMWL in Tiksi (δD=7.57 δ18O–6.8), indicative for meteoric precipitation stored in ice wedges. However, the broad scatter of the data shows large inter-annual and seasonal variability of the frost cracking process. A first comparison with meteorological data indicates that December temperatures may best explain the variability in recent ice wedge isotope composition.