Role of SST feedback in the prediction of the boreal summer monsoon intraseasonal oscillation

The Indian summer monsoon intraseasonal oscillations (MISOs) induce pronounced intraseasonal sea surface temperature (SST) variability in the Bay of Bengal (BoB), which has important feedbacks to atmospheric convection. An ocean general circulation model (OGCM) is employed to investigate the upper‐ocean processes affecting intraseasonal SST variability and its feedback to the MISO convection. In the BoB, the MISO induces intraseasonal SST variability predominantly through surface heat flux forcing with comparable contributions from shortwave radiation and turbulent heat flux, and to a much smaller extent through wind‐driven ocean mixed layer entrainment. The ocean salinity stratification, represented by mixed layer depth (MLD) and barrier layer thickness (BLT), has a strong control on SST but weak impact on convection of the MISO. The MLD is critical for the amplitude of SST response to various forcing processes, while the BLT mainly affects entrainment by determining the temperature difference between the mixed layer and the water below. From May to mid‐June, the shallow MLD and thin barrier layer greatly enhance intraseasonal SST anomalies, which can amplify convection fluctuations of the MISO through air‐sea interaction and leads to intense but short‐duration postconvection break spells. When either the MLD or the BLT is large, intraseasonal SSTs tend to be weak. Further investigation reveals that freshwater flux of the monsoon gives rise to the shallow MLD and thick barrier layer, and its overall effect on intraseasonal SSTs is a 20% enhancement. These results provide implications for improving the simulation and forecast of the MISO in climate models.

Active and break phases of the Indian summer monsoon are associated with sea surface temperature (SST) fluctuations at 30–90 days timescale in the Arabian Sea and Bay of Bengal. Mechanisms responsible for basin-scale intraseasonal SST variations have previously been discussed, but the maxima of SST variability are actually located in three specific offshore regions: the South-Eastern Arabian Sea (SEAS), the Southern Tip of India (STI) and the North-Western Bay of Bengal (NWBoB). In the present study, we use an eddy-permitting 0.25° regional ocean model to investigate mechanisms of this offshore intraseasonal SST variability. Modelled climatological mixed layer and upper thermocline depth are in very good agreement with estimates from three repeated expendable bathythermograph transects perpendicular to the Indian Coast. The model intraseasonal forcing and SST variability agree well with observed estimates, although modelled intraseasonal offshore SST amplitude is undere-stimated by 20–30 %. Our analysis reveals that surface heat flux variations drive a large part of the intraseasonal SST variations along the Indian coastline while oceanic processes have contrasted contributions depending of the region considered. In the SEAS, this contribution is very small because intraseasonal wind variations are essentially cross-shore, and thus not associated with significant upwelling intraseasonal fluctuations. In the STI, vertical advection associated with Ekman pumping contributes to ∼30 % of the SST fluctuations. In the NWBoB, vertical mixing diminishes the SST variations driven by the atmospheric heat flux perturbations by 40 %. Simple slab ocean model integrations show that the amplitude of these intraseasonal SST signals is not very sensitive to the heat flux dataset used, but more sensitive to mixed layer depth.

It has been suggested that the intraseasonal sea surface temperature (SST) variability in the tropical oceans can be amplified by the diurnal cycle of SST (dSST). Here, by analyzing the global tropical moored buoy array for the first time, we find that the intraseasonal SST variability is indeed amplified by the dSST in most of the tropical oceans, especially in the Indo‐Pacific warm pool, but weakened in the equatorial cold tongues of the Pacific and Atlantic Oceans. Such a divergent response is associated with the difference in atmosphere‐ocean interaction processes over these two regions. In the warm pool region, SST responds to the intraseasonal atmospheric variability, resulting in in‐phase intraseasonal fluctuations between SST and dSST that amplify the intraseasonal SST variability. However, in the cold tongue region, SST drives the atmospheric changes, which leads to out‐of‐phase intraseasonal fluctuations between SST and dSST and thus the inhibition of the intraseasonal SST variability.

This study investigates the variability and relationship between intraseasonal sea surface temperature (SST) and surface net heat flux (NHF) variations in the South China Sea and western North Pacific regions. It is shown that the intraseasonal SST variations and their coherence with surface heat flux variations display large differences between winter and summer and between 10–20 day and 30–60 day time scales. The intraseasonal SST variability is comparable on 10–20 day and 30–60 day time scales but larger during summer than during winter. The NHF variability is much larger on the 10–20 day time scale and during winter. The coherence between intraseasonal SST and NHF variations is higher during summer than during winter due to the seasonal change in the mixed‐layer depth. During summer, coherent intraseasonal SST and NHF variations are identified in a southwest‐northeast tilted region from the South China Sea to the subtropics on the 10–20 day time scale but within a broad zonal band from the South China Sea to the Philippine Sea on the 30–60 day time scale. Such difference is not discernable during winter. The contribution of NHF to the SST tendency is larger on the 30–60 day time scale than on the 10–20 day time scale and during summer than during winter. Latent heat flux provides a much larger contribution than shortwave radiation to intraseasonal SST variations in most regions except for the South China Sea during summer on the 30–60 day time scale.

The effect of high-frequency (period <20 days) wind on the intraseasonal (period 20–100 days) sea surface temperature (SST) anomalies over the mid-latitude North Pacific region (35°–45°N, 160°E–170°W) during boreal summer was examined through the diagnosis of reanalysis data and numerical experiments. The reanalysis data diagnosis shows that the near-surface high-frequency (HF) wind is weaker (stronger) during the intraseasonal SST warming (cooling) phase. The phase-dependent amplitude of HF wind is controled by the strength change of the upper-tropospheric westerly jet stream. Because the magnitude of the HF wind over the target region shows significantly positive correlation with the total wind speed, a weaker (higher) HF wind in the SST warming (cooling) phase tends to decrease (increase) total wind speed, which can further suppress (enhance) upward surface latent heat and sensible heat fluxes and strengthen the intraseasonal SST variability. The numerical experiments with an oceanic general circulation model demonstrate that the HF wind can amplify the intraseasonal SST variability over the target region mainly through nonlinearly rectifying intraseasonal surface latent and sensible heat fluxes. The HF wind can explain about 20 % of the intraseasonal SST variability in the target region.

The strongest large-scale intraseasonal (30–110 day) sea surface temperature (SST) variations in austral summer in the tropics are found in the eastern Indian Ocean between Australia and Indonesia (North-Western Australian Basin, or NWAB). TMI and Argo observations indicate that the temperature signal (std. ~0.4 °C) is most prominent within the top 20 m. This temperature signal appears as a standing oscillation with a 40–50 day timescale within the NWAB, associated with ~40 Wm−2 net heat fluxes (primarily shortwave and latent) and ~0.02 Nm−2 wind stress perturbations. This signal is largely related to the Madden-Julian Oscillation. A slab ocean model with climatological observed mixed-layer depth and an ocean general circulation model both accurately reproduce the observed intraseasonal SST oscillations in the NWAB. Both indicate that most of the intraseasonal SST variations in the NWAB in austral winter are related to surface heat flux forcing, and that intraseasonal SST variations are largest in austral summer because the mixed-layer is shallow (~20 m) and thus more responsive during that season. The general circulation model indicates that entrainment cooling plays little role in intraseasonal SST variations. The larger intraseasonal SST variations in the NWAB as compared to the widely-studied thermocline-ridge of the Indian Ocean region is explained by the larger convective and air-sea heat flux perturbations in the NWAB.

Impacts of the ocean surface on the representation of the northward-propagating boreal summer intraseasonal oscillation (NPBSISO) over the Indian monsoon region are analyzed using the National Centers for Environmental Prediction (NCEP) coupled atmosphere–ocean Climate Forecast System (CFS) and its atmospheric component, the NCEP Global Forecast System (GFS). Analyses are based on forecasts of five strong NPBSISO events during June–September 2005–07. The inclusion of an interactive ocean in the model is found to be necessary to maintain the observed NPBSISO. The atmosphere-only GFS is capable of maintaining the convection that propagates from the equator to 12°N with reasonable amplitude within the first 15 days, after which the anomalies become very weak, suggesting that the atmospheric internal dynamics alone are not sufficient to sustain the anomalies to propagate to higher latitudes. Forecasts of the NPBSISO in the CFS are more realistic, with the amplitude of precipitation and 850-mb zonal wind anomalies comparable to that in observations for the entire 30-day target period, but with slower northward propagation compared to that observed. Further, the phase relationship between precipitation, sea surface temperature (SST), and surface latent heat fluxes associated with the NPBSISO in the CFS is similar to that in the observations, with positive precipitation anomalies following warm SST anomalies, which are further led by positive anomalies of the surface latent heat and solar radiation fluxes into the ocean. Additional experiments with the atmosphere-only GFS are performed to examine the impacts of uncertainties in SSTs. It is found that intraseasonal SST anomalies 2–3 times as large as that of the observational bulk SST analysis of Reynolds et al. are needed for the GFS to produce realistic northward propagation of the NPBSISO with reasonable amplitude and to capture the observed phase lag between SST and precipitation. The analysis of the forecasts and the experiments suggests that a realistic representation of the observed propagation of the oscillation by the NCEP model requires not only an interactive ocean but also an intraseasonal SST variability stronger than that of the bulk SST analysis.

During boreal winter, there is a prominent maximum of intraseasonal sea-surface temperature (SST) variability associated with the Madden–Julian Oscillation (MJO) along a Thermocline Ridge located in the southwestern Indian Ocean (5°S–10°S, 60°E–90°E; TRIO region). There is an ongoing debate about the relative importance of air-sea heat fluxes and oceanic processes in driving this intraseasonal SST variability. Furthermore, various studies have suggested that interannual variability of the oceanic structure in the TRIO region could modulate the amplitude of the MJO-driven SST response. In this study, we use observations and ocean general circulation model (OGCM) experiments to quantify these two effects over the 1997–2006 period. Observational analysis indicates that Ekman pumping does not contribute significantly (on average) to intraseasonal SST variability. It is, however, difficult to quantify the relative contribution of net heat fluxes and entrainment to SST intraseasonal variability from observations alone. We therefore use a suite of OGCM experiments to isolate the impacts of each process. During 1997–2006, wind stress contributed on average only about 20% of the intraseasonal SST variability (averaged over the TRIO region), while heat fluxes contributed about 70%, with forcing by shortwave radiation (75%) dominating the other flux components (25%). This estimate is consistent with an independent air-sea flux product, which indicates that shortwave radiation contributes 68% of intraseasonal heat flux variability. The time scale of the heat-flux perturbation, in addition to its amplitude, is also important in controlling the intraseasonal SST signature, with longer periods favouring a larger response. There are also strong year-to-year variations in the respective role of heat fluxes and wind stress. Of the five strong cooling events identified in both observations and the model (two in 1999 and one in 2000, 2001 and 2002), intraseasonal-wind stress dominates the SST signature during 2001 and contributes significantly during 2000. Interannual variations of the subsurface thermal structure associated with the Indian Ocean Dipole or El Niño/La Niña events modulate the MJO-driven SST signature only moderately (by up to 30%), mainly by changing the temperature of water entrained into the mixed layer. The primary factor that controls year-to-year changes in the amplitude of TRIO, intraseasonal SST anomalies is hence the characteristics of intraseasonal surface flux perturbations, rather than changes in the underlying oceanic state.

The Bay of Bengal (BoB) receives a large amount of freshwater from rains and rivers, resulting in large upper-ocean stratification due to the freshening effect. This salinity stratification has been theorized to impact sea-surface temperature (SST) and convection on intra-seasonal time scales by affecting the ocean mixed layer and the barrier layer. This article aims to quantify the impact of salinity stratification on the sub-seasonal variability in SST and convection by using in-situ ocean observations and coupled model experiments. It is shown that monsoon intra-seasonal oscillations (MISOs) exhibit varied levels of intra-seasonal variability in SST and rainfall based on the underlying ocean conditions. The largest intra-seasonal variability in SST does not cause the largest convection variability in the north-western BoB. Instead, moderate variability in SST and rainfall associated with MISOs co-occur with deep mixed layer and thick barrier layer conditions. Realistic representation of river freshwater fluxes in a coupled ocean-atmosphere model leads to improved intra-seasonal SST and rainfall variability. Thick barrier layers in the north-western Bay attenuates the entrainment cooling of the mixed layer, and the high mixed layer heat content provides conducive oceanic conditions for the genesis of monsoon low-pressure systems (LPS), thereby affecting rainfall over India. This study has important implications for operation forecasting using coupled models.

During summer, the northern Indian Ocean exhibits significant atmospheric intraseasonal variability associated with active and break phases of the monsoon in the 30–90 days band. In this paper, we investigate mechanisms of the Sea Surface Temperature (SST) signature of this atmospheric variability, using a combination of observational datasets and Ocean General Circulation Model sensitivity experiments. In addition to the previously-reported intraseasonal SST signature in the Bay of Bengal, observations show clear SST signals in the Arabian Sea related to the active/break cycle of the monsoon. As the atmospheric intraseasonal oscillation moves northward, SST variations appear first at the southern tip of India (day 0), then in the Somali upwelling region (day 10), northern Bay of Bengal (day 19) and finally in the Oman upwelling region (day 23). The Bay of Bengal and Oman signals are most clearly associated with the monsoon active/break index, whereas the relationship with signals near Somali upwelling and the southern tip of India is weaker. In agreement with previous studies, we find that heat flux variations drive most of the intraseasonal SST variability in the Bay of Bengal, both in our model (regression coefficient, 0.9, against ~0.25 for wind stress) and in observations (0.8 regression coefficient); ~60% of the heat flux variation is due do shortwave radiation and ~40% due to latent heat flux. On the other hand, both observations and model results indicate a prominent role of dynamical oceanic processes in the Arabian Sea. Wind-stress variations force about 70–100% of SST intraseasonal variations in the Arabian Sea, through modulation of oceanic processes (entrainment, mixing, Ekman pumping, lateral advection). Our ~100 km resolution model suggests that internal oceanic variability (i.e. eddies) contributes substantially to intraseasonal variability at small-scale in the Somali upwelling region, but does not contribute to large-scale intraseasonal SST variability due to its small spatial scale and random phase relation to the active-break monsoon cycle. The effect of oceanic eddies; however, remains to be explored at a higher spatial resolution.

This study investigates sea surface temperature (SST) and precipitation variations in the eastern Arabian Sea (EAS) induced by the northward-propagating Indian summer monsoon (ISM) intraseasonal oscillations (MISOs) through analyzing satellite observations and the Climate Forecast System Reanalysis (CFSR) and performing ocean general circulation model (OGCM) experiments. MISOs in the EAS achieve the largest intensity in the developing stage (May–June) of the ISM. The MISOs induce intraseasonal SST variability primarily through surface heat flux forcing, contributed by both shortwave radiation and turbulent heat flux, and secondarily through mixed layer entrainment. The shallow mixed layer depth (MLD &amp;lt; 40 m) in the developing stage and decaying stage (September–October) of the ISM significantly amplifies the heat flux forcing effect on SST and causes large intraseasonal SST variability. Meanwhile, the high SST (&amp;gt;29°C) in the developing stage leads to enhanced response of MISO convection to SST anomaly. It means that the ocean state of the EAS region during the developing stage favors active two-way air–sea interaction and the formation of the strong first-pulse MISO event. These results provide compelling evidence for the vital role played by the ocean in the MISO mechanisms and have implications for understanding and forecasting the ISM onset. Compared to satellite observation, MISOs in CFSR data have weaker SST variability by ~50% and biased SST–precipitation relation. Reducing these biases in CFSR, which provides initial conditions of the National Centers for Environmental Prediction (NCEP) Climate Forecast System version 2 (CFSv2), may help improve the ISM rainfall forecast.

This paper evaluates the intraseasonal variability of sea surface temperature (SST) along the Sumatra‐Java southern coast using available satellite‐derived oceanic and atmospheric data combined with output from a numerical model. The result reveals that the intraseasonal variability of SST is greater during boreal summer–fall (June–October) than during boreal winter–spring (November–May). Composite analysis shows a correlation between positive/negative intraseasonal SST variabilities and coastal downwelling/upwelling, as well as onshore/offshore Ekman transport during summer–fall. During this period, with the significantly increasing role of oceanic advection, oceanic processes are evidently enhanced and dominate the intraseasonal variability of SST. Meanwhile, the contribution of atmospheric processes drops by 67%. During winter–spring, the intraseasonal SST is primarily contributed by atmospheric processes but has a nonsignificant relationship with sea level anomalies. Intraseasonal SST anomalies vary out of phase with surface wind anomalies. The result also shows a relatively small contribution by vertical processes throughout the year, with the maximum in April and the minimum during August–September. Further analysis reveals that the alternating dominance of atmospheric and oceanic processes on intraseasonal variability of SST is responsible for the seasonality along the Sumatra‐Java southern coast. Moreover, the result indicates that the seasonality in intraseasonal SST is different in the eastern Indonesian Seas, which tends to be relatively strong in boreal winter. Distinct dominance of atmospheric and oceanic processes in intraseasonal SST is the main reason for these differences in seasonal variation characteristics.

The Sea Surface Temperature (SST) intraseasonal variability (40–90 days) along the coast of Peru is commonly attributed to the efficient oceanic connection with the equatorial variability. Here we investigate the respective roles of local and remote equatorial forcing on the intraseasonal SST variability off central Peru (8°S–16°S) during the 2000–2008 period, based on the experimentation with a regional ocean model. We conduct model experiments with different open lateral boundary conditions and/or surface atmospheric forcing (i.e., climatological or not). Despite evidence of clear propagations of coastal trapped waves of equatorial origin and the comparable marked seasonal cycle in intraseasonal Kelvin wave activity and coastal SST variability (i.e., peak in Austral summer), this remote equatorial forcing only accounts for ∼20% of the intraseasonal SST regime, which instead is mainly forced by the local winds and heat fluxes. A heat budget analysis further reveals that during the Austral summer, despite the weak along‐shore upwelling (downwelling) favorable wind stress anomalies, significant cool (warm) SST anomalies along the coast are to a large extent driven by Ekman‐induced advection. This is shown to be due to the shallow mixed layer that increases the efficiency by which wind stress anomalies relates to SST through advection. Diabatic processes also contribute to the SST intraseasonal regime, which tends to shorten the lag between peak SST and wind stress anomalies compared to what is predicted from an advective mixed‐layer model.

Atmospheric dynamical mechanisms have been prevalently used to explain the characteristics of the sum- mer monsoon intraseasonal oscillation (MISO), which dictates the wet and dry spells of the monsoon rainfall. Recent studies show that ocean-atmosphere coupling has a vital role in simulating the observed amplitude and rela- tionship between precipitation and sea surface temperature (SST) at the intraseasonal scale. However it is not clear whether this role is simply 'passive' response to the atmospheric forcing alone, or 'active' in modulating the northward propagation of MISO, and also whether the extent to which it modulates is considerably noteworthy. Using coupled NCEP-Climate Forecast System (CFSv2) model and its atmospheric component the Global Forecast System (GFS), we investigate the relative role of the atmospheric dynamics and the ocean-atmosphere coupling in the initiation, maintenance, and northward propagation of MISO. Three numerical simulations are performed including (1) CFSv2 coupled with high frequency inter- active SST, the GFS forced with both (2) observed monthly SST (interpolated to daily) and (3) daily SST obtained from the CFSv2 simulations. Both CFSv2 and GFS simulate MISO of slightly higher period (*60 days) than observations (*45 days) and have reasonable seasonal rainfall over India. While MISO simulated by CFSv2 has realistic northward propagation, both the GFS model experiments show standing mode of MISO over India with no northward propagation of convection from the equator. The improvement in northward propagation in CFSv2, therefore, may not be due to improvement of the model physics in the atmospheric component alone. Our analysis indicates that even with the presence of conducive vertical wind shear, the absence of meridional humidity gradient and moistening of the atmosphere column north of con- vection hinders the northward movement of convection in GFS. This moistening mechanism works only in the pres- ence of an 'active' ocean. In CFSv2, the lead-lag rela- tionship between the atmospheric fluxes, SST and convection are maintained, while such lead-lag is unreal- istic in the uncoupled simulations. This leads to the con- clusion that high frequent and interactive ocean- atmosphere coupling is a necessary and crucial condition for reproducing the realistic northward propagation of MISO in this particular model.

The effect of high-frequency (< 20 days) wind on the intraseasonal sea surface temperature (SST) anomaly associated with the Madden-Julian oscillation (MJO) is examined by diagnosing reanalysis and outputs from a set of oceanic general circulation model (OGCM) experiments. Warm SST anomaly (SSTA) ahead of MJO convective center induces anomalous boundary-layer convergence, favoring the eastward propagation of the MJO. To understand the key physical processes contributing to the warm SSTA, the mixed-layer heat budget equation is diagnosed. The time change of SSTA ($$\partial \left\langle T \right\rangle /\partial t$$) mostly comes from shortwave radiative heating, while latent heat flux (LHF) plays the secondary role. Due to the strong nonlinearity of LHF, the high-frequency (< 20 days) wind may affect the intraseasonal LHF variability via interacting with the background state, resulting in changes in intraseasonal SSTA. Our diagnosis shows that the upscale feedback associated with high-frequency wind variability accounts for around 23% of the intraseasonal LHF in the intraseasonal SST warming region, supporting the growth of $$\partial \left\langle T \right\rangle /\partial t$$. Sensitivity experiments are then designed using an OGCM that simulates the upper-ocean temperature well, to verify the high-frequency wind effect on the intraseasonal SST variability. Once the high-frequency component of surface winds is removed in the model integration, the amplitudes of intraseasonal LHF and $$\partial \left\langle T \right\rangle /\partial t$$ are decreased, leading to reduced SSTA. The modeling results confirm the positive role of high-frequency wind in supporting the tropical intraseasonal SST variation. The findings of this study suggest that an accurate representation of high-frequency disturbances and their interaction with other components are crucial for MJO simulation and prediction.