Fruit growing
and viticulture of South Russia

Annotation

The issues of global geographical identification of wine products are considered. It has been found that for most modern studies, the use of chemometric processing of data obtained by promising methods for controlling macro- and microelements in wines is characteristic: atomic absorption spectroscopy, atomic emission spectroscopy, mass spectrometry with inductively coupled plasma and atomic emission spectroscopy with inductively coupled plasma. Analysis of macro - and microelements of wines from different countries of the world showed that a large variety of elemental composition of wine products and significant variability caused by technological treatments require the use of modern mathematical processing methods to establish the relationship of the composition of wines with a certain terroir. It is shown that on the basis of analysis of studies of the distribution and relationship of the content of micro and macro elements in wines and soils on which the corresponding grape varieties were grown, can be used to establish the geographical identification of wines. Studies aimed at assessing the relationship of the geographical location of the plots and cultivated grape varieties were carried out according to the correlations between the content of a number of macro- and microelements in wine and soils and the main parameters (grape variety, crop year). The results of the assessment of the relationship between individual samples of wines and soils using correlation, cluster analysis, the �random forest� algorithm and multi-label classification allowed us to evaluate the relationship of the geographical location of the plots and cultivated grape varieties with the content of elements and their pairwise ratios in the wines. Rb, Al, and Na were selected as markers for variety identification for the first year of research. Positive results for Cabernet Sauvignon, Riesling and Pinot Noir are obtained. Applying of selective indexes based on cross-validation allowed to reduce a set of significant indicators to 4: Ca, Ca / Sr, Sr and Ca / Al.