ABSTRACT
Identifying causal relationships among multiple variables and direction/strength of these causal links are key challenges in the analysis of complex dynamical systems. Statistical correlation and regression analysis have been important tools in understanding the relationships in social and environmental systems. However, these tools are not adequate to find stimulus–response mechanisms. Unlike medical or social systems, interventional experiments are not possible for real environmental systems to discover the causal links. Instead, the main focus of causality analysis is on the observational type of methods based on the data-driven approaches. The goal of this chapter is to review some common approaches of causality detection in the areas of climate and environmental science. Here, we focused on three classes of methods to demonstrate their application in the area and compare their performance on multiple generated and real-world time series.
