Causal Drivers and Mechanisms of Monsoon Rainfall Over the Northern Indian Subcontinent

The Indian summer monsoon (ISM) features an intense rainy season typically lasting from June to September that is responsible for approximately 75% of the total annual rainfall on the Indian subcontinent. Specifically, the Western (WHF) and Eastern Himalayan foothills (EHF) receive very high amounts of precipitation during the ISM season while also being densely populated [1, 2]. Therefore, a better understanding of the processes controlling ISM intraseasonal variability are of great societal relevance.

In our present work we identify and quantify causal relationships at short lead-times (three to nine days) between characteristic remote and local climate patterns and the precipitation over the WHF and EHF in order to uncover the underlying mechanisms. For this purpose we  first apply the so-called response-guided causal precursor detection (RG-CPD) scheme, an algorithm designed to identify causal precursors of a variable of interest [3]. The employed method is based on concepts of information theory and statistical mechanics, and allows to identify strongly interdependent climate patterns associated with the ISM and to distinguish between spurious and truly causal links. Resulting from this, causal effect networks (CENs) summarize the relationships among different climate variables by visually representing the actual direct causal linkages between the different variables, their strength and directionality, and associated time-lag.

Our analysis reveals that WHF rainfall variability is influenced by mid-latitude teleconnections such as the circumglobal teleconnection index. This can be seen in the analysis of the geopotential height at 200 hPa and in the 2m temperature. In addition the mean sea-level pressure of the Indian Ocean and the outgoing longwave radiation act as causal precursors to the rainfall. In general the WHF rainfall seems to be driven by similar precursors as the precipitation on the monsoon trough, which corresponds to a large region on the Indian subcontinent receiving the highest amounts of rainfall [4, 5]. By contrast, EHF rainfall is driven by a different set of atmospheric processes. Specifically, we find a causal driver in the eastern equatorial Pacific manifesting itself in the geopotential height at 500 hPa and the mean sea-level pressure, potentially indicating that intraseasonal tropical variability patterns associated with the Madden-Julian oscillation and/or the Walker circulation might exert a significant influence on EHF rainfall. The obtained results by this study may be relevant for designing improved (statistical) forecasts of monsoonal rainfall activity in the different regions beyond synoptic time scales.

References
[1] Vellore, R., et al., On the anomalous precipitation enhancement over the Himalayan foothills during monsoon breaks, Clim. Dynam., 43, 2009-2031 (2014).
[2] Vellore, R., et al., Monsoon { extratropical circulation interactions in Himalayan extreme rainfall, Clim. Dynam., 46, 3517-3564 (2016).
[3] Runge, J., Causal network reconstruction from time series: From theoretical assumptions to practical estimation, Chaos, 28, 075310 (2018).
[4] Di Capua, G., et al., Tropical and mid-latitude teleconnections interacting with the Indian summer monsoon rainfall: a theory-guided causal effect network approach, Earth Syst. Dyn., 11, 17-34 (2020).
[5] Di Capua, G., et al., Long-Lead Statistical Forecasts of the Indian Summer Monsoon Rainfall Based on Causal Precursors, Weather Forecast., 34, 1377-1394 (2019).