Boreal Summer Intraseasonal Oscillation (BSISO)

• BSISO is a boreal-summer intraseasonal convective mode characterized by a southeast–northwest-tilted precipitation structure and a northward or northeastward propagation over the Asian monsoon region.
• The topic employs multiple diagnostic frameworks such as multivariate phase–amplitude indexing, event synchronization, and EEOF filtering to capture its complex variability.
• BSISO influences extreme rainfall, humid heatwaves, and monsoon predictability, with its behavior critically modulated by background SST, land–ocean coupling, and regional teleconnections.

Searching arXiv for recent BSISO-related papers and the cited IDs to ground the article. The Boreal Summer Intraseasonal Oscillation (BSISO) is a pronounced mode of tropical variability that organizes boreal-summer convection, rainfall, and circulation across the Asian monsoon domain and into the western Pacific. In the supplied literature, it is described as the boreal-summer manifestation of the broader MJO-related tropical intraseasonal mode family, with a southeast–northwest-tilted precipitation structure from South Asia to the western Pacific, northeastward and northward propagation, and characteristic time scales described as $30$–$45$ days or $30$–$60$ days depending on the framework adopted (Rocuet et al., 22 Sep 2025). It is central to active–break variability of the Indian Summer Monsoon, to extreme-rainfall organization, and to subseasonal hydroclimatic risk, but recent work also emphasizes that canonical propagation is not universal and that BSISO impacts depend strongly on background state, regional coupling, and the diagnostic lens used (Strnad et al., 2023).

1. Definition, scope, and relation to adjacent intraseasonal modes

BSISO is treated in the literature as a boreal-summer intraseasonal convective mode over the Asian monsoon region, but its conceptual framing is not uniform. One line of work states that the BSISO “can be seen as a seasonal variation of the same underlying dynamics as the MJO,” while emphasizing that boreal summer shifts the mode toward the Northern Hemisphere and gives it a southeast–northwest-tilted precipitation structure from South Asia to the western Pacific with northeastward propagation on a $30$–$45$-day timescale (Rocuet et al., 22 Sep 2025). Another line of work retains the more monsoon-centered term MISO and explicitly notes that “forecasting MISO or Boreal Summer Intraseasonal Oscillations (BSISO) by the S2S systems has stagnated,” thereby treating Indian-sector MISO as the operationally relevant branch of BSISO (Banerjee et al., 2 Jul 2026).

The relationship between BSISO and the “boreal summer MJO” is also handled differently across studies. Over Indonesia, one study defines the mode using the RMM framework and states that “the main difference between boreal winter and boreal summer MJO is its propagation,” specifically that boreal summer shows “more northward tilt of the propagation when passing the Maritime Continent,” while also acknowledging that further work may need to consider interactions with “summer monsoon (also known as BSISO)” (Muhammad et al., 2021). This framing implies strong overlap in observed behavior, but it does not constitute a dedicated BSISO diagnosis.

A further distinction is required between the canonical BSISO and the boreal-summer quasi-biweekly oscillation. The latter is treated as a westward-propagating, $10$–$25$-day, planetary-scale tropical intraseasonal mode during JJAS, and the supplied analysis is explicit that its conclusions apply most directly to the westward QBWO/CCER-like branch rather than to the full BSISO spectrum, especially not to the canonical $30$–$60$-day northward-propagating mode (Biswas et al., 11 Dec 2025). This suggests that “BSISO” is best understood as a family of boreal-summer monsoon-coupled intraseasonal disturbances with multiple observational and theoretical projections rather than as a single universally diagnosed object.

2. Diagnostic frameworks and observational identification

The modern literature uses several parallel frameworks to identify BSISO. A common approach is multivariate phase–amplitude indexing. In an Indo–African rainfall teleconnection study, BSISO is diagnosed with the daily Kikuchi et al. index using the first two leading principal components and the amplitude

$45$0

with $45$1 defined as active and $45$2 as inactive, and the $45$3–$45$4 phase space divided into eight equal sectors (Strnad et al., 2024). A separate study of extreme-rainfall propagation within the South Asian Summer Monsoon instead uses the daily Lee et al. BSISO index and classifies days with $45$5 as active (Strnad et al., 2023). A global humid-heat study does not separate MJO and BSISO with a dedicated summer index; instead it uses the all-season OLR-based MJO Index, divides the combined two-dimensional phase space into eight phases, and composites only days with amplitude $45$6, arguing that OMI is “more all-round” because it better captures boreal-summer north-eastward propagation (Rocuet et al., 22 Sep 2025).

A second major framework is event synchronization and climate-network analysis of extreme rainfall. In the North India–Sahel study, daily MSWEP precipitation over 1980–2022 is converted into local extreme rainfall event series by defining wet days as days with rainfall at least $45$7 mm and EREs as exceedances of the local $45$8, then constructing a directed climate network with an event-synchronization algorithm allowing a dynamical delay up to $45$9 days and retaining links that exceed the $30$0th percentile from $30$1 surrogate event-sequence pairs (Strnad et al., 2024). An analogous SASM-domain study identifies community structure with a Bayesian stochastic block model and uses EIO synchronization maxima as day 0 for propagation analysis, rather than beginning from the BSISO index itself (Strnad et al., 2023). On a broader scale, a global study of boreal-summer extreme-rainfall interdependence also uses event synchronization, now with TRMM and GPM daily rainfall, adaptive delays capped at $30$2 days, and consensus clustering plus mutual correspondences to recover robust synchronized structures of the global monsoon (Su et al., 2021).

A third framework emphasizes filtered anomaly fields and EEOFs. Over South Asia, one process study uses ERA5 and NOAA OLR filtered to $30$3–$30$4 days, then defines propagating and non-propagating BSISO events from box-mean OLR anomalies over the equatorial Indian Ocean and South Asian land (Ghatak et al., 2023). In monsoon-model evaluation, rainfall-based MISO diagnostics are built from Wheeler–Kiladis spectra, regional meridional wavenumber–frequency spectra, and an eight-phase MISO composite based on rainfall EEOFs over $30$5–$30$6–$30$7 (Banerjee et al., 2 Jul 2026). The coexistence of these frameworks is methodologically consequential: some studies diagnose BSISO as a filtered, propagating mode; others diagnose its impacts through conditioned extremes, synchronization, or phase-dependent risk.

3. Canonical structure, propagation diversity, and physical mechanisms

The canonical BSISO structure in the supplied literature is a convective/rainfall system that initiates over the equatorial Indian Ocean, develops a northwest–southeast-tilted band, propagates northward over South Asia, and also projects eastward toward the Maritime Continent and western Pacific (Strnad et al., 2023). A South Asian process study sharpens the mechanism by contrasting northward-propagating and non-propagating events. It argues that robust propagation begins when convection over the equatorial Indian Ocean moistens the lower free troposphere of the southern Arabian Sea through horizontal advection of the background moisture field by anomalous BSISO winds; this permits convection to extend into the Arabian Sea and form a NW–SE convective band (Ghatak et al., 2023). In a second stage, the presence of background easterly vertical shear of monsoon winds and a meridional gradient of anomalous vertical velocity generates a vortex-tilting tendency, especially through

$30$8

which reorients the Rossby gyre, produces south-easterly BSISO winds over the South Asian landmass, and enables further northward propagation over land (Ghatak et al., 2023). Non-propagating events fail because the initial Rossby response is too weak to moisten the southern Arabian Sea sufficiently, so the tilting mechanism never develops and convection stalls south of about $30$9.

Recent work also shows that the canonical pathway is not unique. A network-based propagation study identifies three modes: north-eastward, eastward-blocked, and stationary (Strnad et al., 2023). The canonical cluster contains $60$0 events and exhibits a zonal phase speed of about $60$1 and a northward phase speed of about $60$2. The eastward-blocked cluster contains $60$3 events; it still moves northward at about $60$4 but its eastward branch weakens to about $60$5 and does not cross the Maritime Continent. The stationary cluster contains $60$6 events and remains confined to the equatorial Indian Ocean (Strnad et al., 2023). The proposed mechanism is that Pacific SST background state modulates local zonal and meridional overturning circulations and the BSISO Kelvin wave component: El Niño-like conditions favor stationary behavior, while La Niña-like conditions favor eastward-blocked behavior (Strnad et al., 2023). This directly challenges the widespread shorthand that BSISO propagation is always canonical.

Minimal-theory work supplies a complementary viewpoint. A moist shallow-water study finds that when the saturation field varies in both latitude and longitude, the long-time response includes not only westward Rossby modes but also a slow eastward-propagating, predominantly rotational mode best explained by moist potential vorticity conservation,

$60$7

rather than by a sustained Kelvin-wave interpretation (Suhas et al., 2019). In realistic July saturation fields, this eastward response is “restricted to the northern hemisphere,” passes over the Indian landmass into the subtropics, reaches across the Pacific to North America, and produces wet and dry oscillations over India with a time period of about $60$8 to $60$9 days (Suhas et al., 2019). This does not amount to a complete BSISO theory, but it suggests that BSISO-like eastward or northeastward propagation can emerge from moist rotational dynamics in an inhomogeneous summer moisture background.

The quasi-biweekly branch exhibits a different but related sensitivity to background state. A QBWO study shows that in dry and moderately moist regions with easterly mean flow, the mode behaves like a mean-flow-driven linear mode, whereas in very moist regions with westerly flow, eddy advection of background vorticity and moisture becomes dominant (Biswas et al., 11 Dec 2025). A plausible implication is that boreal-summer intraseasonal propagation mechanisms are state dependent even within the broader BSISO family.

4. Hydroclimatic expression and regional impacts

BSISO organizes not only convective propagation but also phase-dependent hydroclimatic risk. Over the South Asian monsoon domain, a climate-network study shows a typical sequence of synchronized extreme-rainfall communities from the equatorial Indian Ocean to the Bay of Bengal, Maritime Continent, South Asia, and western Pacific over about $30$0 days, with the canonical propagation mode providing conditional lead times of roughly $30$1–$30$2 days for the Bay of Bengal, $30$3–$30$4 days for the Maritime Continent, $30$5–$30$6 days for South Asia, and $30$7–$30$8 days for the western Pacific after EIO synchronization peaks (Strnad et al., 2023). This framing links BSISO directly to early warning of coherent extreme-rainfall episodes rather than only to filtered anomaly composites.

Over Indonesia, summertime intraseasonal variability identified in an RMM framework strongly modulates the probability of $30$9th-percentile rainfall exceedance. The strongest positive signal occurs in western and northern Indonesia, with regional-average increases of around $45$0 during phase $45$1 and $45$2 during phase $45$3 in the west, and approximately $45$4, $45$5, and $45$6 during phases $45$7, $45$8, and $45$9 in the north; the south shows only up to about $10$0 increase (Muhammad et al., 2021). The paper interprets this north–south contrast as consistent with the northward tilt of boreal-summer propagation over the Maritime Continent, making the Indonesian response strongly BSISO-relevant even though the analysis is framed as “boreal summer MJO.”

The influence extends beyond rainfall alone. A global humid-heat study shows that the MJO/BSISO phase framework strongly modulates wet-bulb temperature and humid heatwave occurrence across much of the tropics and subtropics in MJJASO, with humid heatwave likelihood able to double or halve depending on phase and reaching almost triple the seasonal mean in the western Pacific during phases $10$1&$10$2 (Rocuet et al., 22 Sep 2025). Its core mechanistic conclusion is that summer $10$3 anomalies arise primarily through specific humidity rather than dry-bulb temperature anomalies, with horizontal moisture advection in the planetary boundary layer playing a leading role (Rocuet et al., 22 Sep 2025). Over India, warm $10$4 anomalies coincide with rainfall peaks, reinforcing the view that BSISO hydroclimate impacts are often moisture loaded rather than simply temperature driven.

BSISO also conditions oceanic extremes that feed back onto monsoon rainfall. In the Arabian Sea, short-lived summer marine heatwaves are initiated by enhanced surface shortwave radiation and reduced latent heat loss associated with the suppressed convection phase of BSISO, with favorable phases in the eastern Arabian Sea corresponding to phases $10$5–$10$6 and in the northern Arabian Sea to phases $10$7–$10$8 (Suhas et al., 18 Mar 2026). Once these MHWs develop, they induce low-level cyclonic circulation, moisture convergence, and higher extreme-precipitation risk. Kerala experiences a $10$9–$25$0 times higher risk of extreme precipitation during BSISOs accompanied by MHWs than during BSISOs without MHWs, and northwestern India and Pakistan show a similar $25$1–$25$2 times risk increase in the northern Arabian Sea influence region (Suhas et al., 18 Mar 2026).

At the synoptic scale, BSISO phase also organizes the genesis pathways of Middle Tropospheric Cyclones over the Arabian Sea and western India. Objective tracking and clustering show that Type $25$3 and Type $25$4 systems occur primarily in phases $25$5 and $25$6, whereas Type $25$7 and Type $25$8 favor phases $25$9 and $30$0; Type $30$1 is associated with rain rates exceeding $30$2 mm/day (Kushwaha et al., 2022). This places BSISO as an environmental conditioner of synoptic heavy-rain systems, not merely a backdrop to seasonal monsoon variability.

5. Background-state modulation, land–ocean coupling, and teleconnections

A recurring result across the supplied literature is that BSISO does not operate in isolation from the slowly varying background state. In the Indo–African extreme-rainfall teleconnection, La Niña-like SST conditions form the slowly varying background, strengthen the Indian Summer Monsoon, favor northward BSISO propagation, and intensify the Tropical Easterly Jet (Strnad et al., 2024). In that framework, BSISO is the intraseasonal trigger or modulator of convection over northwest India, but the TEJ provides the upper-level propagation corridor and La Niña-like conditions intensify both the Indian monsoon and the BSISO/TEJ configuration (Strnad et al., 2024). The extreme-rainfall linkage is sharply lagged: North India EREs are typically followed by Sahel EREs after about $30$3 days, with one stronger propagation mode around $30$4–$30$5 days and a weaker one around $30$6–$30$7 days (Strnad et al., 2024). The authors repeatedly emphasize that this is not a direct trans-African propagation of a canonical BSISO wave; rather, BSISO initiates the Indian convection, while the TEJ advects upper-tropospheric moist anomalies westward and the Sahel response emerges after transport and local convective amplification (Strnad et al., 2024).

Land coupling provides a second type of background modulation. A study of land ITCZ maintenance during active phases of the Indian Summer Monsoon shows that soil moisture itself has significant intraseasonal variability, with rainfall, evaporation, and soil moisture all showing strongest variance in the $30$8–$30$9-day band (Gautam et al., 2023). During active monsoon phases, maximum positive soil-moisture anomaly appears over northwest India rather than in the seasonal-mean wettest regions, and shallow soil moisture in the northern monsoon core zone leads rainfall by a few days (Gautam et al., 2023). The proposed pathway is soil moisture $60$0 evaporation $60$1 boundary-layer humidification $60$2 higher moist static energy and instability $60$3 stronger land ITCZ and rainfall, implying that land-surface moisture acts as a local intraseasonal memory reservoir over India (Gautam et al., 2023). This does not replace large-scale BSISO dynamics, but it does provide a mechanism for sustaining the land branch of the active phase once the convective envelope reaches India.

The global-monsoon context further supports the importance of subseasonal reorganization. A network study of global extreme-rainfall interdependence identifies two major boreal-summer synchronized structures with independent temporal and spatial characteristics: an early-summer regime from early June to mid-July and a later-summer regime from mid-July to late August (Su et al., 2021). These structures “manifest the primary intraseasonal variability in the context of the global monsoon,” including the monsoon jump over East Asia and West Africa and the mid-summer drought over Central America and southern Mexico (Su et al., 2021). This is not an explicit BSISO diagnosis, but it suggests that BSISO should be situated within a broader network of synchronized monsoon subseasonal transitions.

6. Modeling, predictability, and unresolved issues

Modeling work indicates that reproducing BSISO requires accurate representation of both mean state and coupled subseasonal dynamics. In a deep-learning coupled Earth system emulator, SamudrACE simulates a recognizable Indian-sector MISO/BSISO with realistic northward migration of the monsoon rain band up to about $60$4 and a qualitatively correct eight-phase northward progression (Banerjee et al., 2 Jul 2026). However, its dominant period is biased: the observed MISO peak occurs at $60$5 days, whereas CM4 and SamudrACE peak at $60$6 days; modeled power at $60$7 days is about $60$8 of observed, while power at $60$9 days is $45$00–$45$01 times larger than observed (Banerjee et al., 2 Jul 2026). The propagating rain band is weaker than observed, too zonal, and initiated from a mislocated oceanic rain band near $45$02 rather than the observed $45$03, while Kelvin waves with $45$04–$45$05-day periods and antisymmetric MRG waves are also underestimated (Banerjee et al., 2 Jul 2026). The paper therefore treats current AI/ML coupled emulators as promising but not yet faithful enough for robust BSISO/MISO S2S use.

The predictability implications of the broader literature are nevertheless substantial. The humid-heat study argues that because dynamical forecasts of MJO/BSISO are skillful at roughly $45$06 to $45$07 weeks, the strong phase dependence of humid-heatwave risk implies a real subseasonal forecasting opportunity (Rocuet et al., 22 Sep 2025). The South Asian propagation study argues that identifying whether an equatorial Indian Ocean initiation will evolve into a canonical, eastward-blocked, or stationary mode could extend useful anticipation of regionally coherent extreme-rainfall episodes to roughly four weeks (Strnad et al., 2023). The MTC classification study similarly suggests that BSISO phase, combined with Bay of Bengal disturbance monitoring, provides advance warning of favored synoptic pathways and heavy rain over western India (Kushwaha et al., 2022).

Several unresolved issues recur. One is index choice: some studies use dedicated BSISO indices, others use RMM, OMI, or rainfall EEOFs, and the resulting “BSISO” is not always the same dynamical object (Strnad et al., 2024). Another is process separation: the Indonesian precipitation-extremes study is highly relevant to BSISO interpretation but remains within an MJO/RMM lens (Muhammad et al., 2021), while the global humid-heat study explicitly states, “Here we do not differentiate between MJO and BSISO” (Rocuet et al., 22 Sep 2025). A third is the limits of causal attribution. In the North India–Sahel study, the network-detected synchronization, the $45$08-day lag, and the TEJ composites are directly shown, but the full causal chain linking La Niña, BSISO, TEJ, and Sahel convection remains a physically argued synthesis rather than a strict causal attribution experiment (Strnad et al., 2024). More generally, the supplied literature suggests that BSISO should not be reduced to a single universal mechanism. Depending on region and time scale, the relevant balances may involve free-tropospheric moisture advection, vortex tilting in a vertically sheared monsoon background, Kelvin-wave modulation by Pacific SST, land-surface preconditioning, TEJ-mediated moisture transport, or eddy advection of background moisture and vorticity.

Taken together, these studies define BSISO as a monsoon-coupled boreal-summer intraseasonal mode whose canonical north-eastward propagation remains foundational, but whose observed behavior is conditional on background SST, land and ocean memory, vertical shear, moisture gradients, and the diagnostic framework used to isolate it. The contemporary literature therefore treats BSISO not merely as a recurrent convective pattern, but as a dynamically heterogeneous organizing mode of boreal-summer hydroclimate across South Asia, the Maritime Continent, the western Pacific, the Arabian Sea, and, through teleconnected pathways, regions as remote as the Sahel.