A software package for exploratory time series analysis of environmental data

In Brunetti et al. (2006), Brunetti et al. (2009) the authors examined the variability and changes in precipitation in the Greater Alpine Region (GAR) over the period from 1800 to 2003 and introduced an innovative method for analyzing and visualizing trends and moving averages within the data series. This approach involves representing values from analyzed subseries on a color-coded chart, where the x-axis indicates the temporal position of the subseries' midpoint and the y-axis represents the width of the subseries being analyzed. The color scale highlights the entire possible spectrum of significant trends and averages in the dataset, offering a highly detailed and quantitative description of observed phenomena. This visualization method effectively identifies key features and trends across various timescales, enabling a comprehensive understanding of the data. In this contribution, we present an R package designed to perform such analyses for observational time series and to visualise climatological data in a graph, particularly those gathered within the eLTER network. Beyond replicating the aforementioned analysis, we introduce a complementary function for analyzing the full spectrum of correlations between two temporal data series. These visualization techniques enable preliminary data exploration to ensure the quality and suitability of datasets for further statistical analysis. They also allow for an in-depth examination of time series, including the correlation spectrum between two historical datasets. Preliminar results We performed the analysis on precipitation time series (Figs 1, 2) representing two distant eLTER sites, Svartberget in Sweden (Anonymous 2019) and Aigüestortes in Catalonia, Spain (Gacia 2018) over the period from 01/01/2005 to 13/11/2017. This analysis showed a negative correlation between the two precipitation patterns observed at the sites, without highlighting any particular statistical significance for long periods of analysis. In both datasets, seasonality in the precipitation patterns can be noted, along with a small anomaly after 2010 (the year ends at the time mark 2198 as annotated in Fig. 2) when the Catalonian site observed a drier period and the Swedish site a wetter period than usual (just following the period observed in Spain) Bartholy and Pongrácz 2013. What is relevant is a recent (after time mark 4000, i.e., 2016) shift towards a positive correlation in precipitation patterns between the two sites which appears to hold a certain significance. The presented approach not only enhances the reliability of climate and environmental studies but also offers a robust framework for exploring complex patterns in temporal datasets, thereby supporting informed decision-making and long-term ecological research. This is a preliminary analysis of a methodology that could be applied as a service for the sites involved in eLTER-RI, to the data series collected as part of the eLTER Data Mobilization or within the scope of the Standard Observations (SOs).