This session explores the need, challenges, opportunities, drawbacks, and obstacles in applying and explaining the special world of compositional data spaces.

Compositional data is a special type of non-negative data that carries relevant information in the ratios among the variables rather than in the actual data values. Considering the metric of compositional data may conflict with familiar geoscientific approaches: single chemical elements or mineralogical concentrations as proxies for environmental changes, counts of individuals as indicators of ecological states, etc.
Felix Chayes (1960) and John Aitchison (1981) began to establish a parallel universe of statistical methods beyond normality and linearity to gain sustainable insights into Compositional Data Analysis (CoDA) in the geosciences. Since then, the geoscientific community has expanded the range of methods.
However, even in the era of rapidly expanding big data methods, sustainable CoDA remains underrepresented in recent research.

We welcome contributions from all disciplines to answer the following questions: (i) How does CoDA transform our view of the earth? (ii) How can we gain new sustainable insights by applying CoDA? (iii) How can CodA increase the scalability of numeric models? (iv) How to deal with zeros and missing data? (vi) How to deal with low-count compositions? (v) How should we convincingly communicate CoDA?
This session addresses researchers from all quantitative fields to exchange ideas and foster interdisciplinary collaboration.