The predictability of global land dry-wet condition contributed by sea surface temperature of Pacific: A new perspective from nonlinear dynamical system

Ocean states significantly influence the temporal variability of dry-wet conditions (DWC). Quantifying their contribution to DWC predictability is crucial for drought and extreme pluvial events early-warning and prevention. Due to the complex physical mechanism, the relationship between ocean states and DWC is inherently nonlinear, necessitating a statistical tool capable of addressing such nonlinearity for better quantification. This study employs Nonlinear Dynamical System (NDS) theory to understand the predictability of DWC, and Convergent Cross-mapping (CCM) to assess the contribution of Sea Surface Temperature (SST) indices—Niño1 + 2 and Niño3.4—to the predictability of precipitation and DWC globally. Using the Pearl River Basin (PRB) as an example case, we illustrate the calculation of CCM and the interpretation of its results. For instance, indicated by CCM skills, the Niño1 + 2 accounts for approximately 0.78 of the predictability of precipitation in PRB, whereas the Niño3.4 contributes around 0.46. After that, we apply CCM to map the spatial distribution of SST’s contribution to the global land surface precipitation and DWC, highlighting regions where SST has the highest and lowest contributions. More importantly, we confirm the applicability of CCM and NDS theory for SST and DWC, validating our findings and signifying that the predictability in the NDS is short-term but has a deterministic nature. As a first attempt, this study provides additional avenues to understand the predictability of DWC and to quantify the contribution of ocean states to it, from a global perspective based on NDS theory.