Extreme rainfall propagation within Boreal Summer Intraseasonal Oscillation modulated by Pacific sea surface temperature

Felix Strnad, Jakob Schloer, Ruth Geen, Niklas Boers, Bedartha Goswami

TL;DR

Researchers have identified three distinct "modes" of how the Boreal Summer Intraseasonal Oscillation (BSISO) propagates, revealing that Pacific sea surface temperatures act as a remote control for extreme rainfall in South Asia. By understanding how El Niño and La Niña conditions "block" or "stall" these weather patterns, the study demonstrates the potential to predict extreme rainfall events up to four weeks in advance.

What Problem Does This Paper Address?

The South Asian Summer Monsoon (SASM) is the lifeblood of a region housing 2.5 billion people, but it is notoriously volatile. Much of this volatility is driven by the BSISO—a massive, quasi-periodic pulse of clouds and rain that typically moves northeast from the Indian Ocean toward the Western Pacific.

While meteorologists understand the "canonical" (standard) path of the BSISO, it often deviates from this script. Sometimes it stalls; other times, it hits a wall at the Maritime Continent (the islands of Southeast Asia). These deviations lead to unexpected "active" (flood) and "break" (drought) periods that wreak havoc on agriculture and infrastructure. This paper seeks to categorize these deviations and identify the global climate drivers—specifically Pacific sea surface temperatures—that cause them.

Methods

The authors move beyond traditional linear statistics by employing Climate Networks and Machine Learning:

  1. Event Synchronization & Community Detection: Instead of looking at raw rainfall averages, they used "event synchronization" to see how extreme rainfall events (EREs) at different locations "talk" to each other. They then applied a Stochastic Block Model (SBM)—a probabilistic clustering algorithm—to group geographical regions into six "communities" that experience synchronous extreme rain.
  2. Hovmöller Clustering: To track movement, they created Hovmöller diagrams (plots showing how a variable like outgoing longwave radiation changes over time and space). They used K-means clustering on these diagrams to objectively sort BSISO cycles into three distinct propagation modes.
  3. Dynamic Analysis: They analyzed atmospheric "overturning circulations" (Walker and Hadley cells) and wave dynamics (Kelvin and Rossby waves) to explain the physical mechanisms behind the clusters.

Datasets

  • MSWEP (Multi-Source Weighted-Ensemble Precipitation): High-resolution daily precipitation data from 1979–2020.
  • ERA5 Reanalysis: Used for atmospheric variables like wind fields, humidity, and sea surface temperatures (SST).
  • BSISO Index: A standardized index used to track the phase and intensity of the oscillation.
  • NINO3.4 Index: Used to categorize the state of the El Niño Southern Oscillation (ENSO).

Key Insights and Ideas

  • The Three Modes of BSISO: The study identifies that the BSISO isn't just one thing. It operates in three modes:
  • Canonical: Moves smoothly northeast (the "standard" behavior).
  • Eastward-blocked: Moves north but gets "stuck" before reaching the Pacific.
  • Stationary: Stays put in the equatorial Indian Ocean.
  • The Pacific "Gatekeeper": Pacific SSTs modulate the "gate" at the Maritime Continent. La Niña conditions favor the Eastward-blocked mode, while El Niño conditions favor the Stationary mode.
  • Wave Suppression: The "Aha!" moment is the discovery that during El Niño, descending dry air over the Maritime Continent creates opposing winds that effectively "choke" the Kelvin waves required for the BSISO to move eastward.
  • Predictive Power: Because these modes are tied to slowly-evolving ocean temperatures, they provide a "predictable" framework for weather that usually seems chaotic.

Key Results

  • Frequency: The Canonical mode occurs about twice as often as the other two modes.
  • ENSO Correlation: Normal Pacific conditions favor the Canonical mode. El Niño-like conditions significantly increase the likelihood of the Stationary mode, while La Niña-like conditions favor the Eastward-blocked mode.
  • Early Warning Potential: The researchers showed that extreme rainfall in specific regions follows a predictable sequence. For example, if a "trigger" event occurs in the Indian Ocean, there is a 66% probability of extreme rain in the Bay of Bengal 4–6 days later, and a 61% probability in South Asia 15–18 days later.
  • Lead Time: This approach demonstrates the potential for spatially resolved early warning signals up to 25 days (nearly 4 weeks) in advance, significantly outperforming current operational models that struggle beyond 14 days.

Limitations and Future Work

  • Sample Size: The study is limited by the observational record (roughly 40 years), which results in a relatively small number of BSISO events for clustering.
  • Model Representation: The authors note that many current Global Climate Models (GCMs) fail to simulate these diverse propagation modes accurately. A key next step is using these findings to improve how models handle the "Maritime Continent barrier."
  • Real-time Application: While the potential for a 4-week warning is shown, the study is a retrospective analysis. Future work involves turning this framework into a real-time forecasting tool using machine learning.

Summary updated

Abstract

Intraseasonal variations of the South Asian Summer Monsoon (SASM) contain alternating extreme rainfall (active) and low rainfall (break) phases impacting agriculture and economies. Their timing and spatial location are dominated by the Boreal Summer Intraseasonal Oscillation (BSISO), a quasi-periodic movement of convective precipitation from the equatorial Indian Ocean to the Western Pacific. However, observed deviations from the BSISO's canonical north-eastward propagation are poorly understood. Utilizing climate networks to characterize how active phases propagate within the SASM domain and using clustering analysis, we reveal three distinct modes of BSISO propagation: north-eastward, eastward-blocked, and stationary. We further show that Pacific sea surface temperatures modulate the modes - with El Niño- (La Niña-) like conditions favoring the stationary (eastward-blocked) - by changing local zonal and meridional overturning circulations and the BSISO Kelvin wave component. Using these insights, we demonstrate the potential for early warning signals of extreme rainfall until four weeks in advance.
BibTeX
@inproceedings{strnad2023extreme,
  author = {Felix Strnad and Jakob Schloer and Ruth Geen and Niklas Boers and Bedartha Goswami},
  title = {{Extreme rainfall propagation within Boreal Summer Intraseasonal Oscillation modulated by Pacific sea surface temperature}},
  booktitle = {Nature Communications},
  year = {2023},
  doi = {10.1038/s41467-023-41400-9},
}
Figure