New study extracts dominant influence routes from massive climate datasets. It shows causal pathways between teleconnections.
The computing framework avoids ML. Phases include:
• Gridded data is transformed and stored in relational and graph databases for fast geospatial retrieval.
• The pipeline attacks dimensionality through spatial clustering.
• Causal processing: one representative time series per cluster is computed by the SLARAC method.
• A Yen’s k-shortest path algorithm extracts the most relevant causal chains.
Cruz, J. E. C., da Rocha, V. R., de Andrade, F. M., & Quiles, M. G. (2026). "A multivariate causal chain detection methodology assessed on the ENSO modulation of the Northeastern Brazil precipitation" https://doi.org/10.1007/s41060-026-01322-3
#ENSO #ElNiño #causality #climate #weather #SLARAC #networks #dataEngineering #Brazil
