Modulation of monsoon intraseasonal ... - Wiley Online Library

18 downloads 1847 Views 4MB Size Report
May 7, 2014 - region has been influenced by the recent warming in the Indian Ocean. .... Arkin, 1997] provided by the NOAA via http://www.esrl.noaa.gov/psd/.
PUBLICATIONS Journal of Geophysical Research: Atmospheres RESEARCH ARTICLE 10.1002/2013JD021261 Key Points: • MISO variance has increased over ISMR region in the warming period • Northward propagation has slowed down in the warming period • Mean SST increase over the Indian Ocean are responsible for these change

Correspondence to: S. A. Rao, [email protected]

Citation: Sabeerali, C. T., S. A. Rao, G. George, D. N. Rao, S. Mahapatra, A. Kulkarni, and R. Murtugudde (2014), Modulation of monsoon intraseasonal oscillations in the recent warming period, J. Geophys. Res. Atmos., 119, 5185–5203, doi:10.1002/ 2013JD021261. Received 26 NOV 2013 Accepted 15 APR 2014 Accepted article online 22 APR 2014 Published online 7 MAY 2014

Modulation of monsoon intraseasonal oscillations in the recent warming period C. T. Sabeerali1, Suryachandra A Rao1, Gibies George1, D. Nagarjuna Rao1, S. Mahapatra1, A. Kulkarni1, and Raghu Murtugudde2 1

Indian Institute of Tropical Meteorology, Pune, India, 2Earth System Science Interdisciplinary Center, University of Maryland, College Park, Maryland, USA

Abstract The Indian Ocean sea surface temperature (SST) during the boreal summer has shown a significant warming of 0.3°C in the recent decade (2001–2010) compared to a former decade (1979–1988), and it is most pronounced in the central tropical Indian Ocean. By using reanalysis and satellite-derived data sets, we investigated how the monsoon intraseasonal oscillation (MISO) over the South Asian summer monsoon (ASM) region has been influenced by the recent warming in the Indian Ocean. It is found that the MISO variance has increased over the ASM region in the recent period compared with the earlier decade. It is also noted that the characteristic northward propagation of the MISO has slowed over 2001–2010, resembling more of a standing oscillation near the equator. Mechanisms implicated in the observed MISO changes are explored by conducting several model sensitivity experiments with an atmospheric general circulation model. The model experiments suggest that the mean SST increase over the Indian Ocean, and the associated changes in the air-sea interaction, the increased mean moisture convergence, and changes in the large-scale circulation are responsible for the changes in the characteristics of the MISO. The influence of the recent Indian Ocean warming on the MISO characteristics must be understood fully since they determine the seasonal amount of rainfall over the Indian subcontinent. An examination of future projections of the MISO using the MPI-ESM-LR model from the Coupled Model Intercomparison Project phase 5 archive also gives consistent result. 1. Introduction The summer monsoon rainfall over South Asia is not temporally homogeneous, and it consists of periods of abundant and scanty rainfall; these are the manifestations of a relatively high-frequency westward propagating mode with a 10–20 day period [Krishnamurti and Bhalme, 1976; Chen and Chen, 1993; Annamalai and Slingo, 2001; Chatterjee and Goswami, 2004] and a lower frequency northeastward propagating mode with a period of 30–60 days [Yasunari, 1979; Sikka and Gadgil, 1980; Annamalai and Sperber, 2005; Goswami, 2005]. The active and break phases of the South Asian summer monsoon (ASM) are closely connected to these fluctuations in the tropical atmosphere [Yasunari, 1980; Goswami and Shukla, 1984; Gadgil and Asha, 1992; Lawrence and Webster, 2002] and thus have received wide attention in the literature. It has been shown that prolonged dry spells during crop growth periods can have a severe impact on crop growth and hence yields [Lal et al., 1999]. Reliable prediction of seasonal mean Indian summer monsoon rainfall is of critical importance because millions of people across large parts of India depend upon rain-fed agriculture for their existence. Earlier studies have shown that the dominant mode of monsoon intraseasonal oscillation (MISO) has a large spatial structure similar to that of the seasonal mean and its interannual variability [Sperber et al., 2000; Goswami and Ajaya Mohan, 2001], and therefore, the MISO has a crucial role in deciding the seasonal monsoon strength and its predictability [Sperber et al., 2000; Goswami and Ajaya Mohan, 2001; Gadgil, 2003]. Several studies have found a link between the intraseasonal variations in the convection, and large-scale dynamics [Goswami and Shukla, 1984; Sperber et al., 2000; Goswami and Ajaya Mohan, 2001]. Wang and Xie [1997] showed the importance of large-scale mean monsoon flows and moisture distribution in the structure, development, and maintenance of the MISO. Using an atmospheric general circulation model, Krishnan and Venkatesan [1997] showed that the amplitude of the MISO increases (decreases) when the moisture availability in the model increases (decreases). One of the most striking features of the MISO is its northward propagation from the equator to about 25°N [Yasunari, 1979, 1980; Sikka and Gadgil, 1980; Webster et al., 1998]. Several different theories have been

SABEERALI ET AL.

©2014. American Geophysical Union. All Rights Reserved.

5185

Journal of Geophysical Research: Atmospheres

10.1002/2013JD021261

proposed to identify the processes responsible for the northward propagation of the MISO using theoretical and modeling frameworks [Webster, 1983; Goswami and Shukla, 1984; Gadgil and Srinivasan, 1990; Wang and Xie, 1997; Jiang et al., 2004; Wang and Xie, 1997; Lawrence and Webster, 2002; Kemball-Cook and Wang, 2001, and several others]. While a single mechanism has yet to emerge, a conceptual picture of the poleward propagation of the MISO was introduced by Jiang et al. [2004]. It is shown that the strong easterly vertical wind shear over the ASM region couples the baroclinic and barotropic modes in the free atmosphere and creates barotropic vorticity to the north of maximum convection and as a result, divergence in the free atmosphere is established to the north of the maximum convection. This leads to the northward shift of boundary layer moisture convergence, and thus, the MISO convection and facilitates the northward propagation of the MISO, and this mechanism is applicable only in the off-equatorial region [Jiang et al., 2004]. However, close to the equator, the northward propagation of the MISO is mainly controlled by the lowlevel moisture gradient in the north-south direction [Jiang et al., 2004]. A significant slowdown of the first episode (May–June) of northward propagating MISO due to the weakening of vertical easterly wind shear and north-south gradient of the mean low-level humidity around the equator during the recent decades has also been reported [Goswami et al., 2010]. Annamalai and Slingo [2001] noted that the regional Hadley circulation and the Walker circulation are closely connected to the MISO. Therefore, any changes to the circulation pattern will also be reflected in the characteristics of the MISO. Some recent studies have highlighted that increased tropical sea surface temperature (SST) will strengthen ocean-atmosphere coupling and thereby modulate the regional Hadley circulation [Neena and Goswami, 2010; Rao et al., 2010]. Interestingly, a weakening of the atmospheric Walker circulation is claimed in a warmer climate [Vecchi et al., 2006; Vecchi and Soden, 2007; Zhang and Song, 2006]. The importance of the Indian Ocean SSTs for the MISO characteristics has been explored using observations and general circulation models [Sengupta et al., 2001; Vecchi and Harrison, 2002; Fu et al., 2003; Klingaman et al., 2008; Wang et al., 2009; Krishnamurthy and Kirtman, 2009; Achuthavarier and Krishnamurthy, 2011; Lin et al., 2011]. A consistent relationship between convection, SST, and surface heat fluxes in the Indian Ocean has been established on the MISO time scale during the boreal summer [Sengupta et al., 2001; Fu et al., 2003]. During the northward propagation of the MISO, warm (cold) SST leads (lags) the maximum convection by about 10 days in the Indian Ocean [Sengupta et al., 2001; Vecchi and Harrison, 2002; Goswami, 2005]. When convection reaches its maximum (wet phase), the incoming solar radiation decreases, followed by an increase in latent heat release due to enhanced evaporation associated with an enhanced westerly wind. This leads to cold SST anomalies about 10 days after peak convection. Similarly, during the suppressed phase of convection, the incoming solar radiation increases followed by a decrease in latent heat release associated with decreased westerly wind and, finally, leads to warm SST anomalies about 10 days before the maximum convection. The Indian Ocean has been continuously warming for the last three decades, displaying a strong linear trend in the SST [Rao et al., 2012], and the warming trend is most pronounced in the tropical Indian Ocean compared to the other tropical ocean basins. It is found that the ongoing warming of the tropical Indian Ocean in the recent decades has strengthened the convection over the west central equatorial Indian Ocean and led to anomalous easterlies along the equator [Rao et al., 2012], consequently favoring frequent positive Indian Ocean Dipole (IOD) events. In another study, Rao et al. [2010] showed that the combined influence of linear warming trend in the tropical Indian Ocean and the warming associated with the positive IOD created a strong anomalous warming in the southern tropical Indian Ocean region in the 2008 summer monsoon period and caused the abnormal drought in the central Indian region. Recently, Wu et al. [2012] studied the global SST trend for the 1900–2008 period, and they found that the mean SST over the global ocean has warmed with a trend of 0.62°C/century. We also carried out similar trend analysis. The boreal summer SST warming trend of the global (60°S–60°N, 0°–360°E) and the central tropical Indian Ocean (15°S–6°N, 60°E–95°E) over two different periods (1900–2010 and 1979–2010) for two distinct data sets are given in Table 1. The magnitude of the warming trends is different for the two data sets. However, both data sets show that the warming is more rapid in the recent three decades (1979–2010) compared to the whole 1900–2010 period, and it is most pronounced in the central tropical Indian Ocean compared to the global ocean (Table 1). All the trend values showed in Table 1 are significant at 99% confidence level. To understand the future change in the MISO, it is essential to fully understand the response

SABEERALI ET AL.

©2014. American Geophysical Union. All Rights Reserved.

5186

Journal of Geophysical Research: Atmospheres

10.1002/2013JD021261

Table 1. The Boreal Summer (JJAS) SST Warming Trend of Global (60°S–60°N, 0°–360°E) and Central Tropical Indian a Ocean (15°S–6°N, 60°E–95°E) Over Two Different Periods (1900–2010 and 1979–2010) for Two Distinct Data Sets HadISST Period 1900–2010 1979–2010 a

ERSST V3b

Global

Central Tropical Indian Ocean

Global

Central Tropical Indian Ocean

0.58°C/century 0.96°C/century

0.65°C/century 1.32°C/century

0.74°C/century 1.22°C/century

0.85°C/century 1.61°C/century

Global: 60°S–60°N, 0°–360°. Central Indian Ocean: 15°S–6°N, 60°E–95°E.

of the MISO to the recent warming. Therefore, in the present study, we investigate the influence of the recent Indian Ocean warming on the MISO characteristics by using the daily/pentad observational and reanalysis data sets. The factors responsible for this change are also analyzed. In addition, a set of atmospheric general circulation model (AGCM) sensitivity experiments is carried out to understand the impact of the SST trend in modulating the MISO. The rest of the paper is organized as follows. Section 2 describes the data sets, analysis procedures, and experiment setup. The changes in the characteristics of the MISO in the recent decade along with possible reasons are discussed in section 3. In section 4, the mechanism responsible for the MISO change is described. Modulation of MISO under the future warming scenarios is discussed in section 5. The summary and conclusions are presented in section 6.

2. Data Sets, Methods, and Experiment Design 2.1. Data Sets To study the space-time characteristics of the MISO, the pentad mean Climate Prediction Center Merged Analysis of Precipitation (CMAP) [Xie and Arkin, 1997] provided by the NOAA via http://www.esrl.noaa.gov/psd/ is used in this study. The daily averaged outgoing longwave radiation (OLR) from the NOAA advanced very high resolution radiometer [Liebmann and Smith, 1996] is also employed to verify the result. The horizontal resolution for the CMAP and OLR data sets are 2.5° × 2.5°. The specific humidity, pressure vertical velocity, and the meridional and zonal components of wind are obtained from National Centers for Environmental Prediction (NCEP)-National Center for Atmospheric Research (NCAR) reanalysis I [Kalnay et al., 1996]. Additional verification is carried out with the high-resolution (1° × 1°) daily gridded rainfall data set (version 2) prepared by the India Meteorological Department (IMD) [Rajeevan et al., 2006], the European Centre for Medium-Range Weather Forecasts (ECMWF) Re-Analysis (ERA-40) [Uppala et al., 2005], the ERA-interim Re-analysis [Dee et al., 2011], and the pentad Global Precipitation Climatology Project (GPCP) [Xie et al., 2003]. In addition, the daily GPCP data [Huffman et al., 1997] are also used in the MISO analysis. To quantify Indian Ocean warming, the monthly NOAA Extended Reconstructed SST version 3 (ERSST V3) [Smith et al., 2008] data set (2° × 2° grid) and the Hadley Centre Global Sea Ice and SST (HadISST) from the Hadley center [Rayner et al., 2003] are used. We also use the highresolution (1° × 1°) optimum interpolated daily SST for the period 1982 to 2010 [Reynolds et al., 2007] to analyze the air-sea coupling over the Indian Ocean. In addition to the observational data sets, we have also used model output from MPI-ESM-LR, which is one of the CMIP5 (Coupled Model Intercomparison Project Phase 5) models developed by the Max Planck Institute for Meteorology (http://pcmdi9.llnl.gov/). The atmospheric component of this coupled model is European Centre/Hamburg version 6 (ECHAM6), the latest version of the AGCM that is used in this study (ECHAM5). The RCP4.5 output for the 2081–2100 period in comparison to the last 20 years of the historical simulations is used to assess the future change in the MISO characteristics under the global warming scenario. In the RCP4.5 run, the radiative forcing will increase and then stabilize at about 4.5 W m2 in the year 2100 [Taylor et al., 2012]. 2.2. Methods To study the recent changes in the characteristics of the MISO, we have selected two decades, viz., 1979–1988 and 2001–2010. We carried out the analysis separately for each decade. We should point out that the time series of the SST over the central tropical Indian Ocean clearly shows a strong interannual variation in the period 1979–1988 compared to 2001–2010 (Figure 2b). This is the motivation for selecting these two periods. To examine the sensitivity of our result to the choice of the periods, we repeated our analysis for different periods. For example, we compared 2001–2010 against 1982–1991, divided the period since 1979 into two

SABEERALI ET AL.

©2014. American Geophysical Union. All Rights Reserved.

5187

Journal of Geophysical Research: Atmospheres

10.1002/2013JD021261

halves (1979–1994 and 1995–2010), and compared the result. We found that the results presented in this study are not sensitive to the time periods selected, even if we change the period of study. To focus on the intraseasonal variability, a 20–100 day band Lanczos filter [Duchon, 1979] is applied to the OLR and rainfall anomalies. The daily anomalies are the departures from the climatological annual cycle (sum of annual mean and first three harmonics). Because our focus is on the MISO, all the analysis is carried out for the period from June to September (JJAS). To quantify the change in mean moisture transport over the ASM region for recent period, we computed the divergent component of the moisture transport over the ASM region following previous studies [Chen, 1985; Behera et al., 1999; Rao et al., 2010]. The vertically integrated water vapor transport is computed as: Q ¼

1 g

Psurf



Pt

qVdP

where g is the gravitational acceleration and Psurf is the pressure at the surface; Pt is the pressure at the top of the atmosphere, q is the specific humidity, and V is the horizontal wind vector. In the present study, the 300 hPa pressure level is considered as the top of the atmosphere. By using the Helmholtz theorem, the water vapor transport can be decomposed into a rotational and a divergent component. The significance of the various statistical analyses has been quantified using the Student’s t test. 2.3. The Model Description and Experiment Design The present study uses the Max Planck Institute for Meteorology atmospheric general circulation model, ECHAM5, to understand the importance of SST trend in modulating the MISO characteristics. A detailed description of the model is given in Roeckner et al. [2003]. In this study, we ran the model at T106 resolution with 19 vertical layers. Some previous studies [Annamalai et al., 2005; Rao et al., 2010; Sabeerali et al., 2012] have used this model to understand the impact of the Indo-Pacific SST on the characteristics of the Indian summer monsoon. They showed that this model reproduces the impact of the Indian/Pacific Ocean SST on the Indian summer monsoon well, and the model is able to reproduce the large-scale pattern of precipitation and circulation reasonably well albeit with an overestimation of boreal summer rainfall (JJAS mean) over the equatorial tropical Indian Ocean and the coasts of India. Recently, Abhik et al. [2013] conducted a detailed study on the simulation of MISO in ECHAM5, and they showed that this AGCM shows a reasonable skill in simulating some of the salient features of MISO. Although, the model overestimates the MISO variance over the eastern equatorial Indian Ocean, it faithfully captures the large-scale MISO variance pattern [Abhik et al., 2013]. However, there are some limitations in the northward propagation of the MISO in the model, and these limitations should be viewed by keeping in mind the crucial role of air-sea interaction in the northward propagation and that the SST-forced AGCM simulations tend to be affected by the lack of coupling. Despite the drawback in simulating the MISO amplitude over certain areas, MISO periodicity and thus the period of active and break events are well captured in this model [Liess et al., 2005]. Therefore, this model can be used to study the impact of the SST trend on the MISO characteristics. Three types of AGCM experiments are carried out in this study to understand the role of SST trend in modulating the MISO characteristics. The first run is termed the warming run, where the model is forced with the observed monthly SST and the sea ice as the lower boundary condition and integrated from 1979 to 1998 (20 years). The second run is termed the nonwarming run, in which the trend is removed from the observed global SST for the period 1956–1999 (for each month the trend was removed separately), and the second half of the trend removed SST (1978–1999) is used as the lower boundary condition to integrate the model from 1979 to 1998. The third run is called the Indian Ocean nonwarming run (IONW), which is similar to the nonwarming run, except that the SST trend is removed only from the Indian Ocean (30°S–30°N, 40°E–120°E), and the model is forced with observed SST elsewhere. The difference in the JJAS climatology of SST used in the warming run and nonwarming run are shown in Figure 1a. The SST prescribed in the warming run is warmer compared to the SST prescribed in the nonwarming run in most parts of the tropics, and it shows that a pronounced warming occurs in central tropical Indian Ocean, which agrees with the result of Rao et al. [2010, 2012]. The impact of global ocean SST trends on the MISO characteristics can be studied by comparing the warming run and nonwarming run. Similarly, the difference in the JJAS climatology of SST used in the warming run and Indian Ocean nonwarming run is shown in Figure 1b. Clearly, the comparison of the MISO

SABEERALI ET AL.

©2014. American Geophysical Union. All Rights Reserved.

5188

Journal of Geophysical Research: Atmospheres

10.1002/2013JD021261

Figure 1. Difference in the climatological JJAS mean SST (°C) used to force the (a) warming run and the nonwarming run and (b) warming run and Indian Ocean nonwarming (IONW) run.

simulated in the warming run and Indian Ocean nonwarming run isolates the impact of Indian Ocean SST trends on the MISO characteristics.

3. Observational Results 3.1. The Change in the Mean SST Over the Indian Ocean in the Recent Decade Before discussing the MISO change in the recent decades, we must understand how the basic mean state over the ocean has changed in the recent time. To exhibit the Indian Ocean warming in the recent decade, the difference in the climatological boreal summer (JJAS) SST between a recent decade (2001–2010) and a former decade (1979–1988) has been calculated (Figure 2a). It shows that the whole tropical Indian Ocean has warmed in the recent decade relative to the former decade. However, the most pronounced increase in SST occurred over the central tropical Indian Ocean and is consistent with the recent studies [Ajayamohan and Rao, 2008; Rao et al., 2010, 2012]. Boreal summer SST over the central tropical Indian Ocean (15°S–6°N, 60°E–95°E) for the period 1979–2010 shows a linear warming trend (0.2°C/decade; Figure 2b) which is significant at 99% confidence level. The mean temperature over the central tropical Indian Ocean has shown a significant rise of 0.3°C in the recent decade compared to the former decade. It is also noted that the interannual variations in SSTs are strong in the former decade compared to the recent decade (Figure 2b). When compared to the warming in other regions and other periods, this warming is very rapid (Table 1). Since the rapid warming takes place over the warm pool region of the Indian Ocean, it is expected to have strong influence on atmospheric convection and circulation [Rao et al., 2010]. The changes in the atmospheric circulation are certain to add to changes in the MISO characteristics. Furthermore, several studies have noted that even though the MISO is due to internal variability of the atmosphere, air-sea interaction modifies the characteristic of the MISO [Fu et al., 2003; Goswami, 2005]. 3.2. Modulation of the MISO Variance in the Recent Warming Decade The Arabian Sea, the Bay of Bengal, and the eastern equatorial Indian Ocean are the major regions of MISO activity [Hsu et al., 2004]. The intensity of the MISO in the eastern equatorial Indian Ocean is an indication of

SABEERALI ET AL.

©2014. American Geophysical Union. All Rights Reserved.

5189

Journal of Geophysical Research: Atmospheres

10.1002/2013JD021261

the strength of northward propagation [Wang et al., 2006]. The difference in the intraseasonal variance of band passfiltered (20–100 days) precipitation anomalies between the recent decade (2001–2010), and the former decade (1979–1988) is computed to quantify the change in the MISO variance in the recent decade (Figure 3a). It is found that the MISO variance is increased over the ASM region, particularly over the eastern tropical Indian Ocean, the head Bay of Bengal (~15°N to 22°N) and the adjoining region and the northwestern part of the Indian subcontinent, and the nearby northern Arabian Sea in the recent decade compared to the former decade (Figure 3a). It means that the MISO strength has increased in these regions during the recent period. The pentad GPCP precipitation data and daily OLR field also yield an identical result (figure not shown). Previous studies have shown that in response to enhanced surface warming and associated convection, the MISO intensity increases over the equatorial Indian Ocean after 1975/1976 climate shift [Zveryaev, 2002; Kulkarni et al., 2009]. To confirm our results, the longterm trend of the JJAS mean variance of the filtered (20–100 days) precipitation anomalies is also taken, which also Figure 2. (a) Difference in the JJAS climatological SST (°C) between the illustrates the increased MISO variance recent decade (2001–2010) and a former decade (1979–1988); (b) the time series of JJAS SST averaged over the central Indian Ocean (15°S–6°N, over the eastern equatorial Indian 60°E–95°E) for the period 1979 to 2010. The straight line shows the linear Ocean, the head Bay of Bengal, and the trend (significant at 99% confidence level) for the period 1979 to 2010. northwestern part of India (Figure 3b). The red lines show the mean SST for the two periods (1979–1988 and The long-term trend of the MISO 2001–2010) considered in this study. Statistically significant values are variance in the high-resolution IMD stippled in Figure 2a. precipitation [Rajeevan et al., 2006] for the 1951–2008 period also confirmed the increased variance over the northeastern and the northwestern part of the Indian subcontinent (figure not shown). 3.3. Changes in the Space-Time Characteristics of the MISO This section discusses in detail the space-time characteristics of the MISO in the ASM region and compares the space-time structure of the MISO between the former (1979–1988) and the recent decade (2001–2010). To explore the space-time characteristics of the MISO, a finite domain wave number frequency analysis for June–September (JJAS) is carried out [Teng and Wang, 2003; Fu et al., 2003]. This technique decomposes a field in the time and space domain into a wave number and frequency component [Hayashi, 1971]. The present study assumes that the dominant mode of the MISO is restricted to the latitudinal domain from 10°S to 30°N and longitudinal domain from 65°E to 95°E, as adopted in an earlier study [Fu et al., 2003]. However, modest changes in this domain do not make significant changes in the space-time characteristics of the MISO. During the boreal summer, the dominant mode of intraseasonal oscillation in the ASM domain is in the northward propagating band with a maximum variance around 40 days and wave number 1 [Goswami, 2005].

SABEERALI ET AL.

©2014. American Geophysical Union. All Rights Reserved.

5190

Journal of Geophysical Research: Atmospheres

10.1002/2013JD021261

Figure 3. (a) Difference in the JJAS climatological variance of 20 to 100 day band pass-filtered CMAP precipitation anoma2 2 lies (mm d ) between the recent decade (2001–2010) and a former decade (1979–1988); (b) the long-term trend of JJAS mean variance of 20 to 100 day band pass-filtered CMAP precipitation anomalies from 1979 to 2010. Statistically significant values are stippled.

For both the former and the recent decade, there is a concentration of maximum variance in the northward propagating mode, around wave number 1 (corresponds to a wavelength of 40° in latitude between 10°S and 30°N) and the periodicity 30–60 days (Figures 4a and 4b). It is also observed that the amplitude of the northward propagating band dominates its southward counterpart in both the decades (Figures 4a and 4b). However, the amplitudes of both northward and southward propagating modes have increased in the recent decade compared to the former decade. Though the amplitude of the northward propagating MISO has changed, the meridional wave number of the MISO remains unchanged in the recent decade. The finite domain space-time spectra of the daily OLR and 850 hPa zonal wind fields are also computed to show the robustness of the result (figure not shown), which also confirm the increased MISO amplitude in the recent decade compared to the former decade. 3.4. Change in the Northward Propagation Characteristics in the Recent Decade This section explains in detail the northward propagation characteristics of the MISO in the two periods (1979–1988 and 2001–2010), using the filtered (20–100 days) precipitation anomalies. A significant slowdown

Figure 4. North-south space-time spectra for 10 boreal summer seasons (June–September) calculated over the region 10°S–30°N, 65°E–95°E using the rainfall anomalies from CMAP (a) for the former decade (1979–1988) and (b) for the recent decade (2001–2010).

SABEERALI ET AL.

©2014. American Geophysical Union. All Rights Reserved.

5191

Journal of Geophysical Research: Atmospheres

10.1002/2013JD021261

1

Figure 5. Lag-latitude plot of regressed anomalies of 20–100 day filtered CMAP rainfall data (in mm day ; shaded) and SST (in °C; contour) averaged over 70°–95°E (a) 1979–1988 period and (b) 2001–2010 period. The filtered rainfall anomalies averaged over the monsoon core region (12°N–22°N, 70°E–95°E) are used as the reference time series to compute the regression. Statistically significant contours are shaded (90% confidence interval).

in the phase speed of the first phase (May–June) of the northward propagating MISO during the recent decades is reported [Goswami et al., 2010; Suhas and Goswami, 2008]. The lead/lag regression of 20–100 day band pass-filtered precipitation anomalies during the JJAS season are used to understand the propagation characteristics of the MISO over the ASM region for both periods (1979–1988 and 2001–2010). The filtered JJAS precipitation anomalies averaged over the monsoon core region (12°N–22°N, 70°E–95°E) are used as a reference time series for the regression. The propagation characteristics of the MISO can be most clearly observed from the latitude-lag regression plot (Figures 5a and 5b). During the former decade (1979–1988), organized northward propagation from the equatorial Indian Ocean is evident in the ASM region (Figure 5a). However, during the recent decade (2001–2010) the organized northward propagation of convection is evident only from 5°N onward, and the phase speed of the northward propagation of the MISO slightly slowed down in 2001–2010 (phase speed 1.14 latitude/day) compared to 1979–1988 (phase speed 1.34 latitude/day) period, resembling more of a standing oscillation around the equatorial region. To show the phase relationship between the convection and SST, the regressed anomaly of filtered SST is overlaid in Figure 5. It is found that the warm (cold) SST always leads (lags) the maximum rainfall. However, over the north Indian Ocean (beyond 15°N), a notable difference in the phase relationship between the SST and the rainfall is observed between the recent and the former decade (1979–1988 and 2001–2010). We return to this point in section 3.5. In order to understand how the northward propagation of the MISO has changed during the recent decade, we should know the current understanding and mechanisms explaining the poleward propagation of the MISO as discussed in section 1. Jiang et al. [2004] have shown that the meridional phase speed of the northward propagating MISO is linearly proportional to the easterly vertical wind shear. The vertical easterly shear of zonal wind between 200 hPa and 850 hPa during June–September period over the Asian monsoon region is examined from NCEP/NCAR reanalysis data sets. It is found that the mean easterly vertical shear over the ASM region during the boreal summer season shows a decreasing trend (significant at 99% confidence level) in the recent decade (Figure 6a). The ERA-40 data also confirm the decreasing easterly shear of zonal wind over the ASM region (figure not shown). The poleward propagation of the MISO over the equatorial

SABEERALI ET AL.

©2014. American Geophysical Union. All Rights Reserved.

5192

Journal of Geophysical Research: Atmospheres

10.1002/2013JD021261

Figure 6. (a) The time series of boreal summer (June–September) vertical shear of zonal wind (U850–U200) averaged over 1 the domain 60°E–100°E, EQ to 20°N, for the period 1979 to 2010 (unit: m s ). (b) The time series of JJAS mean moisture gradient over the equatorial region from 1979 to 2010 (unit: g/kg). The moisture gradient over the equatorial region is computed as the difference between vertically integrated (1000 to 300 hPa) moisture averaged over two boxes; 60°–90°E, EQ to 10°N and 60°–90°E, 10°S to EQ. Both these trend lines are significant at 99% confidence level.

region is mainly controlled by the moisture gradient across the equator [Jiang et al., 2004]. It is computed as the difference between the integrated moisture (1000–300 hPa) over two boxes; 60–90°E, equator (EQ) to 10°N and 60°–90°E, EQ to 10°S. The year-wise plot of the moisture gradient during JJAS over the equatorial region shows a decreasing trend (significant at 99% confidence level) during recent decades (Figure 6b). The moisture gradient shown here is computed from the NCEP/NCAR reanalysis. In short, the decreasing wind shear over the ASM domain and weakening of the moisture gradient across the equator might be the main factors responsible for the change in the propagation characteristics of the MISO during the recent decade. The meridional vertical structure of composite MISO in both the periods (1979–1988 and 2001–2010) is shown in Figure 7. The vertical structure of the composite MISO is created based on the methodology adopted by previous studies [Jiang et al., 2004; Ajayamohan et al., 2008]. First, a latitude time diagram of positive MISO precipitation anomalies averaged over 70°E–95°E is plotted for each year (figure not shown). From this plot, the MISO which show coherent northward propagation from the equator to 20°N are selected (31 events during 1979–1988 and 32 events during 2001–2010) and the latitude and time of the maximum convection (maximum positive rainfall anomalies) are noted for each northward propagating MISO pulse. A phase composite of the MISO is computed at each reference latitude over which the maximum positive precipitation anomalies occur. Since these composite structures bear a close resemblance, we further made a composite of these structures with respect to the maximum convection center [Jiang et al., 2004; Ajayamohan et al., 2008]. From Figure 7, clearly, during the recent decade, the amplitude of vorticity, divergence, specific humidity, and vertical velocity

SABEERALI ET AL.

©2014. American Geophysical Union. All Rights Reserved.

5193

Journal of Geophysical Research: Atmospheres

10.1002/2013JD021261

4

1

Figure 7. The meridional vertical structure of the northward propagating MISO in the former decade (1979–1988) for (a) specific humidity (10 kg kg ), (b) vertical 1 6 1 7 1 velocity (Pa S ), (c) vorticity (10 S ), and (d) divergence (10 S ). (e–h) Same as Figures 7a–7d but for the recent decade (2001–2010). The filtered anomalies are averaged over 70°–95°E. The vertical axis represents the pressure levels, and the horizontal axis denotes the meridional distance (in degree latitude) with respect to the convection center (0°). The negative (positive) values denote the south (north) of the convection center.

have strengthened to the north of the convection. At the same time, due to the decrease in the easterly shear of zonal wind, and the weakening of the low-level monsoon wind over the ASM in the recent decade, the maximum amplitudes of these parameters are slightly closer to the center of the convection. It facilitates the increased amplitude and decreased phase speed of the MISO over the ASM region during the recent decade. The analysis using the ERA-interim data sets also confirm the increased amplitude of these parameters to the north of the maximum convection center (figure not shown). The strengthened amplitude of vorticity, divergence, specific humidity, and vertical velocity to the north of the convection might be due to the enhanced air-sea interaction over the Indian Ocean driven by the recent increase in the SST. 3.5. Air-Sea Interaction in the Indian Ocean Although, the basic properties of the MISO emerge from the internal atmospheric variability, air-sea interaction in the Indian Ocean has the potential to modify the characteristics of the MISO [Fu et al., 2003; Jiang et al., 2004; Ajayamohan et al., 2008; Wang et al., 2009; Lin et al., 2011].

Figure 8. Lead-lag correlation of filtered anomalies of rainfall and SST over the head Bay of Bengal (15°–20°N, 85°–95°E). The blue line represents the recent decade, and the red dotted line denotes the former decade. The negative (positive) values in the X axis denotes that the SST leads (lags) the rainfall.

SABEERALI ET AL.

©2014. American Geophysical Union. All Rights Reserved.

The recent increase in the SST over the Indian Ocean has the potential to alter the air-sea interaction [Neena and Goswami, 2010]. The phase lag between SSTs and rainfall anomalies computed in the head Bay of Bengal (between 15°N and 20°N) is shown in Figure 8, which indicates that the air-sea interaction has strengthened in this region in the recent

5194

Journal of Geophysical Research: Atmospheres

10.1002/2013JD021261

2

4

Figure 9. (a) Difference in the mean variance of 20 to 100 day band pass-filtered June–September OLR anomalies (W m ) between the (a) warming run and the nonwarming run and (b) warming run and the IONW run. Statistically significant values are stippled.

decade compared to the former decade (Figure 8) as measured by the SST-rainfall correlation [also see Johnson and Xie, 2010]. This strengthened air-sea interaction over the head Bay of Bengal causes an increased specific humidity convergence into this region of convection and maybe one candidate factor contributing to the increased MISO amplitude. It is also observed that over the head Bay of Bengal, the phase lag between the SST and the rainfall has changed. During the recent decade, the maximum SST leads the rainfall by about 10 days, while during the former decade this lead was only about 5 days (Figure 8). It suggests that atmosphere takes more time to respond to a SST anomaly in the recent decade, and therefore, MISO gets enough time to build up its amplitude. The increased moisture-holding capacity of the atmosphere in the recent decade due to warming and associated increase in the residence time of the atmospheric moisture may be one possible reason to increase the maximum SST lead in the recent decade. The increase in the residence time of the atmospheric moisture, in fact, delays the triggering of the convection and hence provides more time for continued warming of SST [Trenberth, 2011]. Hence, the atmosphere effectively takes more time to respond to a SST anomaly in the recent decade. However, further studies with numerical experiments are required to confirm this conjecture, and the findings will be reported elsewhere. A set of AGCM experiments, using the ECHAM5 model, is carried out to understand the impact of increased SST in modulating the MISO characteristics, and the results are discussed in the next section.

4. The AGCM Experiments In the earlier sections, using observation, we showed that the MISO characteristics have changed in the context of the recent Indian Ocean warming. To test the role of SST trend on the characteristics of the MISO, we have carried out a set of AGCM experiments with ECHAM5. The details of each run are given in section 2.3. To understand the model response to the global SST trend on the MISO characteristics, we compare the MISO simulated between the warming run and the nonwarming run. The difference in the simulated MISO variance (20–100 day filtered) between the warming and the nonwarming run is given in Figure 9a. Consistent with observation, the MISO variance has increased over the ASM region in the warming run, in response to the global SST trend. However, a substantial difference is found in the Arabian Sea in the model simulation compared to observations, and it may be due to model bias in simulating the MISO variance. In ECHAM5, the location of maximum MISO variance which is seen in the eastern equatorial Indian Ocean region is extended too far to the western Indian Ocean compared to observations (figure not shown). To understand the relative role of the Indian Ocean SST on the characteristics of the MISO, we compared the MISO simulated in the warming run and the Indian Ocean nonwarming run. The differences in MISO variance between these two runs show that the Indian Ocean warming alone can explain most of these changes (Figure 9b). The simulated space-time characteristics of the MISO in the three different experiments show that the amplitude of northward propagating mode has increased in the warming run (Figures 10a–10c). Further, the simulated amplitude of northward propagating mode in the nonwarming run and the Indian Ocean nonwarming run is similar (Figures 10a and 10c). This implies that the Indian Ocean warming is the major player in modulating the amplitude of the northward propagating mode.

SABEERALI ET AL.

©2014. American Geophysical Union. All Rights Reserved.

5195

Journal of Geophysical Research: Atmospheres

10.1002/2013JD021261

Figure 10. The north-south space-time spectra for the JJAS season calculated over the region 10°S–30°N, 65°E–95°E using the OLR anomalies from (a) nonwarming run, (b) warming run, and (c) Indian Ocean nonwarming (IONW) run.

Although, the Indian Ocean SST trend plays a dominant role in modulating the MISO characteristics, the SST trend in other regions is also important, particularly the SST trend in western Pacific [Hoerling and Kumar, 2003]. Since the MISO spatial structure extends from the India land mass to the equatorial western Pacific [Annamalai and Slingo, 2001] in a tilted rain/OLR anomaly band, the role of western Pacific SST cannot be completely ignored. Hence, in the remaining discussions, we compare the warming run and the nonwarming run. We also analyzed how the propagation characteristics of the MISO changes between the warming and nonwarming run (Figure 11). Consistent with observations, the model also showed a decreased phase speed signal in the warming run, compared to the nonwarming run and it showed some tendency to become stationary over the equatorial region in the warming run. In the warming run, the easterly wind shear over the ASM domain and the specific humidity gradient over the equator are reduced compared to the nonwarming run (Figures 12a and 12b). These results are consistent with observations (Figures 6a and 6b). It is also noted that the region south of the equator becomes more moist compared to the northern latitudes in the warming run (Figure 12b), and that may be due to the pronounced warming over the southern tropical Indian Ocean and related feedbacks between SST and humidity. This decreased wind shear and reduced humidity gradient across the equator have the potential to affect the northward propagation of the MISO as shown by Jiang et al. [2004]. The following section

2

Figure 11. Lag-latitude plot of regressed anomalies of 20–100 day filtered OLR data (W m ) averaged over 70°E–95°E (a) nonwarming run and (b) warming run. The dotted lines represent the slope of the northward propagation in the nonwarming run, and the solid lines represent the slope of the propagation in the warming run. The filtered rainfall anomalies averaged over the region 10°S–5°N, 75°E–100°E are used as the reference time series to compute the regression.

SABEERALI ET AL.

©2014. American Geophysical Union. All Rights Reserved.

5196

Journal of Geophysical Research: Atmospheres

10.1002/2013JD021261

1

Figure 12. (a) Difference in JJAS seasonal mean easterly wind shear (U850–U200, m s ) between the warming run and the nonwarming run. (b) Difference in the vertically integrated (1000 to 300 hPa) JJAS seasonal mean moisture (kg/kg) between the warming run and the nonwarming run. Statistically significant values are stippled.

investigates mechanisms responsible for the decreased wind shear, reduced moisture gradient across the equator, and the increased amplitude of the northward propagating mode by comparing the warming and the nonwarming runs. 4.1. Mechanisms The change in the position of the anomalous moisture convergence over the Indian summer monsoon region has the potential to influence the regional Hadley circulation [Rao et al., 2010]. The difference in the moisture transport between the warming run and the nonwarming run are shown in Figure 13a. As a response to the pronounced warming over the central tropical Indian Ocean, an anomalous moisture convergence occurs over the tropical Indian Ocean region and an anomalous moisture divergence occurs over the Indian land mass in the warming run, compared to the nonwarming run (Figure 13a). As a result, the regional Hadley circulation over the Indian summer monsoon region is also modulated in such a way that an anomalous upward motion is observed over the equatorial region, and an anomalous downward motion is

7

1

1

Figure 13. Difference in the boreal summer (JJAS) (a) divergent component of moisture transport (10 kg S m ) and the related moisture transport vectors between the warming run and nonwarming run (b) regional Hadley circulation (averaged over 65°E–95°E) between the warming run and nonwarming run (c) upper tropospheric temperature (averaged 1 over 600–200 hPa) between the warming run and nonwarming run (d) 850 hPa wind (m s ) between the warming run and nonwarming run.

SABEERALI ET AL.

©2014. American Geophysical Union. All Rights Reserved.

5197

Journal of Geophysical Research: Atmospheres

10.1002/2013JD021261

Figure 14. Schematic representation of the mechanism proposed for the changes in the characteristics of the MISO.

observed over the Indian land mass (Figure 13b). This result is consistent with previous studies [Rao et al., 2010; Neena and Goswami, 2010], where they showed that the increased tropical SSTs would modify the regional Hadley circulation through the enhanced atmosphere ocean coupling. The anomalous moisture convergence and the upward motion over the equatorial and southern tropical Indian Ocean will enhance the upper tropospheric temperature over that region, due to the enhanced convection and associated increase in latent heat release (Figure 13c). As a result, the upper tropospheric temperature gradient over the ASM domain decreases (Figure 13c), thereby reducing the mean low-level monsoon wind during the boreal summer (Figure 13d) through the thermal wind balance, finally reducing the mean easterly wind shear over the ASM domain (Figure 12a). However, some previous studies have questioned the direct role of tropospheric temperature gradient in changing the monsoon winds [Lindzen and Hou, 1988; Bordoni and Schneider, 2008; Molnar et al., 2010], and it has been shown that a weakened regional Hadley circulation with a southward displaced ascending branch can itself lead to a decrease in monsoon winds without invoking changes in the meridional upper tropospheric temperature gradient and hence a weakened wind shear [Lindzen and Hou, 1988; Bordoni and Schneider, 2008]. In contrast to these studies, Abish et al. [2013] have shown that the rapid Indian Ocean warming and the associated convection have warmed the upper troposphere over the central equatorial Indian Ocean and thereby weakened the meridional tropospheric temperature gradient and this has reduced the monsoon winds through the thermal wind balance. However, our discussion on the Indian Ocean warming and associated changes in the regional Hadley circulation and its impact on the MISO propagation are not dependent on any of these arguments. The decreased mean monsoon wind also weakens the mean moisture transport across the equator and decreases the crossequatorial moisture gradient (Figure 12b). The decreased wind shear over the ASM domain together with the decreased moisture gradient across the equator slowdown the phase speed of the northward propagating mode over the ASM domain as we noticed in observations. Earlier studies have argued that the air-sea interaction over the Indian Ocean will enhance the amplitude of the MISO [Fu et al., 2003; Lin et al., 2011]. Using an AGCM experiment, Krishnan and Venkatesan [1997] showed that the amplitude of the MISO depends on moisture availability. They found that the amplitude of MISO increases (decreases) when the moisture availability in the model increases (decreases). By comparing the warming and nonwarming run, we found that the SST trend, particularly over the Indian Ocean, is responsible for the change in the MISO intensity. The enhanced moisture convergence over the ASM domain as a response to increased air-sea interaction, driven by the Indian Ocean warming, causes the increased amplitude of the MISO. The proposed mechanism for the total MISO change is represented schematically in Figure 14.

SABEERALI ET AL.

©2014. American Geophysical Union. All Rights Reserved.

5198

Journal of Geophysical Research: Atmospheres

10.1002/2013JD021261

Figure 15. (a) Difference in the JJAS climatological variance of 20 to 100 day band pass-filtered precipitation anomalies 2 2 (mm d ) between the historical (1986–2005) and the RCP4.5 (2081–2100) run of the MPI-ESM-LR model. Statistically significant values are stippled. (b) The meridional profile of the boreal summer intaseasonal variance of precipitation averaged between 70°E and 100°E for the historical and RCP4.5 run.

5. Modulation of MISO Under Future Warming Scenario Here we shall present the future projections of MISO characteristics using the MPI-ESM-LR model. By analyzing the multiple aspects of MISO simulated in the 32 CMIP5 models, Sabeerali et al. [2013] have shown that the MPI-ESM-LR model is better suited to represent the MISO characteristics. The difference in the intraseasonal variance of band pass-filtered (20–100 days) precipitation anomalies between the RCP4.5 (2081–2100) and the historical (1986–2005) run of the MPI-ESM-LR model is computed to quantify the projected change in the MISO variance under the warming scenario (Figure 15). Under the global warming scenario, MPI-ESM-LR projects an increase in the MISO variance over the Indian monsoon region (Figure 15a). The meridional profile of the MISO variance shows that the MISO variance is increased all the way from the equator to about 25°N under the warming scenario (Figure 15b). However, the projected increase in the MISO variance is most pronounced over the head Bay of Bengal, the central equatorial Indian Ocean, the west coast of India, and the northwest Pacific (Figure 15a). These are the regions where the MISO activity is strong. To explore the projected change in the space-time structure of the MISO, the simulated space-time structure of the MISO between the historical and the RCP4.5 run are compared (Figure 16). In the meridional direction, both the historical and RCP4.5 run simulate a dominant northward propagating mode around wave number 1

Figure 16. North-south space-time spectra for the boreal summer seasons (June–September) calculated over the region 10°S –30°N, 70°E–100°E using the precipitation anomalies from the MPI-ESM-LR model (a) for the historical run (1986–2005) and (b) for the RCP4.5 run (2081–2100).

SABEERALI ET AL.

©2014. American Geophysical Union. All Rights Reserved.

5199

Journal of Geophysical Research: Atmospheres

10.1002/2013JD021261

1

Figure 17. Lag-latitude plot of regressed anomalies of 20–100 day filtered precipitation data (mm d ) averaged over 70°E –95°E (a) historical run and (b) RCP4.5 run. The filtered rainfall anomalies averaged over the region 10°S–5°N, 75°E–100°E are used as the reference time series to compute the regression. The dotted lines represent the slope of the northward propagation in the historical run, and the solid lines represent the slope of the propagation in the RCP4.5 run.

and periodicity between 30 and 60 days. However, the amplitude of the northward propagating mode has increased and the maximum variance is shifted toward the low-frequency (higher periodicity) side, under the RCP4.5 scenario (Figures 16a and 16b). It is also found that the northward propagating phase speed of the MISO slows down significantly in the RCP4.5 scenario compared to the historical period (Figure 17). However, the MISO signals near the equator are almost same between the historical and RCP4.5 scenario (Figures 17a and 17b). The decreased wind shear over the ASM domain is responsible for the projected change in the phase speed of the northward propagation (figure not shown). These changes in the MISO characteristics under the warming scenario are consistent with the observational findings discussed above.

6. Discussion and Conclusions The mean SSTs during the boreal summer season over the Indian Ocean have displayed a significant rise in the recent decade (2001–2010) compared to a former decade (1979–1988), and it is most pronounced in the central (southern) tropical Indian Ocean. The current study investigated the change in the space-time characteristics of the MISO in the recent decade, in the context of this Indian Ocean warming, using the reanalysis, and the satellite-derived global data sets. The MISO variance in these two decades showed a significant difference. It is found that the MISO variance is increased over the Indian summer monsoon region, particularly over the head Bay of Bengal region, the northwestern part of the Indian subcontinent and the nearby northern Arabian Sea, and the eastern tropical Indian Ocean in the recent decade compared to the former decade. This implies that the MISO has strengthened in these regions during the recent decade. The long-term trend of the MISO variance also confirmed the result discussed above. The finite domain space-time spectrum of the precipitation and 850 hPa zonal wind anomalies showed an increase in the amplitude of the northward propagating mode in the recent decade compared to the former decade. However, the meridional wave number of the MISO remains essentially unchanged. The regressed anomalies of filtered rainfall showed a significant difference in the propagation characteristics of the MISO in the recent decade. In the former decade, organized northward propagation from the equatorial Indian Ocean is evident in the ASM region. However, during the recent decade, the organized northward propagation of convection is apparent only from 5°N latitude northward and the phase speed of this propagation is slightly slower. More precisely, during the recent decade, the convection over the equatorial Indian Ocean remains stationary for a longer time and resembles more of a standing oscillation around the equatorial region. To explore the factors responsible for this recent change in the northward propagating characteristics of MISO, we examined how the easterly shear of zonal wind over the ASM region and moisture gradient across

SABEERALI ET AL.

©2014. American Geophysical Union. All Rights Reserved.

5200

Journal of Geophysical Research: Atmospheres

10.1002/2013JD021261

the equator has changed in the recent decade. These are the factors mainly responsible for the northward propagation of the MISO. The long-term trend of the easterly shear of zonal wind over the Indian monsoon region and the moisture gradient across the equator showed a decreasing trend. The weakened zonal wind shear over the ASM region together with the weakened boreal summer monsoon winds facilitated a reduced meridional asymmetry in the maximum amplitude of vorticity, divergence, specific humidity, and vertical velocity with respect to the convection center. In short, a reduction in easterly zonal wind shear and north-south moisture gradient are responsible for the observed change in the propagation characteristics of the MISO in the recent period. A set of AGCM experiments is carried out to understand the model response to the SST trend on the characteristics of the MISO. It is found that the MISO variance has increased over the ASM domain in response to the SST trend, and this result agrees with the observations. Also consistent with observations, the model showed a decreased wind shear and a weakened moisture gradient across the equator. It is also found that Indian Ocean warming alone can explain most of these MISO changes. Model experiments suggest that the mean SST increase over the Indian Ocean and the associated changes in the air-sea interaction, the mean moisture convergence, and the large-scale circulations are together responsible for the change in characteristics of the MISO. An examination of the future projections of the MISO with the MPI_ESM-LR model also confirms most of the observational findings presented above. Clearly, the tropical Indian Ocean is the location for the genesis of Madden-Julian Oscillations (MJOs) and their interactions with MISO have been a focus of numerous studies. Our focus here is solely on the impact of warming on MISO even though they may be intricately interconnected with MJO. A separate study is being carried out to explore the MJO-MISO response to warming and the findings will be reported elsewhere. Also being investigated are the impacts of MISO changes on the trends in monsoon onset in light of a recent study that posits that enhanced MISO activities are typically associated with an early onset of the monsoon [Zhou and Murtugudde, 2014]. Acknowledgments IITM is supported by the Ministry of Earth Sciences, Government of India, New Delhi. The authors thank B.N. Goswami, the Director of the IITM for extending all support for this research work. C.T. Sabeerali, G. George, and D.N. Rao acknowledge the CSIR for research fellowship. NCAR Command Language is extensively used in this study. We would like to acknowledge the NCEP/NCAR and ECMWF for providing the reanalysis data. We also acknowledge the NOAA high-resolution SST data and NOAA_ERSST_V3 provided by the NOAA/OAR/ESRL, PSD, Boulder, Colorado, USA. The HadISST obtained from the website www.metoffice.gov.uk/ hadobs is also acknowledged. We also acknowledge the IMD for providing the rainfall data over the Indian region. We acknowledge the World Climate Research Programme’s Working Group on Coupled Modeling, which is responsible for CMIP5, and we thank the Max Planck Institute for Meteorology for producing and making available their model. R.M. acknowledges his NASA PO and JPL grants for Intraseasonal Bio-Physical Feedback Studies and his ONR DYNAMO and NASA CYGNSS grant. Authors also acknowledge the grant from National Monsoon Mission project. We would like to thank the Editor and three anonymous reviewers for their time and valuable comments.

SABEERALI ET AL.

References Abhik, S., P. Mukhopadhyay, and B. N. Goswami (2013), Evaluation of mean and intraseasonal variability of Indian summer monsoon simulation in ECHAM5: Identification of possible source of bias, Clim. Dyn., doi:10.1007/s00382-013-1824-7. Abish, B., P. V. Joseph, and O. M. Johannessen (2013), Weakening trend of the tropical easterly jet stream of the boreal summer monsoon season 1950–2009, J. Clim., 26, 9408–9414. Achuthavarier, D., and V. Krishnamurthy (2011), Role of Indian and Pacific SST in Indian summer monsoon intraseasonal variability, J. Clim., 24(12), 2915–2930, doi:10.1175/2010JCLI3639.1. Ajayamohan, R. S., and S. A. Rao (2008), Indian Ocean Dipole modulates the number of extreme rainfall events over India in a warming environment, J. Meteorol. Soc. Jpn., 86(1), 245–252, doi:10.2151/jmsj.86.245-252. Ajayamohan, R. S., S. A. Rao, and T. Yamagata (2008), Influence of Indian Ocean dipole on poleward propagation of boreal summer intraseasonal oscillations, J. Clim., 21(21), 5437–5454, doi:10.1175/2008JCLI1758.1. Annamalai, H., and J. M. Slingo (2001), Active/break cycles: Diagnosis of the intraseasonal variability of the Asian summer monsoon, Clim. Dyn., 18(1-2), 85–102, doi:10.1007/s003820100161. Annamalai, H., and K. R. Sperber (2005), Regional heat sources and the active and break phases of boreal summer intraseasonal (30–50 day) variability, J. Atmos.Sci, 62, 2726–2748. Annamalai, H., S. P. Xie, J. P. McCreary, and R. Murtugudde (2005), Impact of Indian Ocean sea surface temperature on developing El Niño, J. Clim., 18(2), 302–319. Behera, S. K., R. Krishnan, and T. Yamagata (1999), Unusual ocean-atmosphere conditions in the tropical Indian Ocean during 1994, Geophys. Res. Lett., 26(19), 3001–3004, doi:10.1029/1999GL010434. Bordoni, S., and T. Schneider (2008), Monsoons as eddy-mediated regime transitions of the tropical overturning circulation, Nat. Geosci., 1, 515–519. Chatterjee, P., and B. N. Goswami (2004), Structure, genesis and scale selection of the tropical quasi-biweekly mode, Q. J. R. Meteorol. Soc., 130(599), 1171–1194, doi:10.1256/qj.03.133. Chen, T.-C. (1985), Global water vapor flux and maintenance during FGGE, Mon. Weather Rev., 113(10), 1801–1819. Chen, T.-C., and J.-M. Chen (1993), The 10–20-day mode of the 1979 Indian monsoon: Its relation with the time variation of monsoon rainfall, Mon. Weather Rev., 121(9), 2465–2482. Dee, D. P., et al. (2011), The ERA-Interim reanalysis: Configuration and performance of the data assimilation system, Q. J. R. Meteorol. Soc., 137(656), 553–597. Duchon, C. E. (1979), Lanczos filtering in one and two dimensions, J. Appl. Meteorol., 18, 1016–1022. Fu, X., B. Wang, T. Li, and J. P. McCreary (2003), Coupling between northward-propagating, intraseasonal oscillations and sea surface temperature in the Indian Ocean, J. Atmos. Sci., 60(15), 1733–1753. Gadgil, S. (2003), The Indian monsoon and its variability, Annu. Rev. Earth Planet. Sci., 31(1), 429–467. Gadgil, S., and G. Asha (1992), Intraseasonal variation of the summer monsoon. Part I: Observational aspects, J. Meteorol. Soc. Jpn., 70(1B), 517–527. Gadgil, S., and J. Srinivasan (1990), Low frequency variation of tropical convergence zones, Meteorol. Atmos. Phys., 44(1), 119–132. Goswami, B. N. (2005), South Asian monsoon, in Intraseasonal Variability of the Atmosphere-Ocean Climate System, edited by W. K. M. Lau and D. E. Waliser, chap. 2, pp. 19–61, Springer-Praxis, Chichester, U. K. Goswami, B. N., and J. Shukla (1984), Quasi-periodic oscillations in a symmetric general circulation model, J. Atmos. Sci., 41(1), 20–37.

©2014. American Geophysical Union. All Rights Reserved.

5201

Journal of Geophysical Research: Atmospheres

10.1002/2013JD021261

Goswami, B. N., and R. S. Ajaya Mohan (2001), Intraseasonal oscillations and interannual variability of the Indian summer monsoon, J. Clim., 14, 1180–1198. Goswami, B. N., J. R. Kulkarni, V. R. Mujumdar, and R. Chattopadhyay (2010), On factors responsible for recent secular trend in the onset phase of monsoon intraseasonal oscillations, Int. J. Clim., 30(14), 2240–2246. Hayashi, Y. (1971), A generalized method of resolving disturbances into progressive and retrogressive waves by space Fourier and time cross-spectral analyses, J. Meteorol. Soc. Jpn., 49, 125–128. Hoerling, M., and A. Kumar (2003), The perfect ocean for drought, Science, 299, 691–694. Hsu, H.-H., C.-H. Weng, and C.-H. Wu (2004), Contrasting characteristics between the northward and eastward propagation of the intraseasonal oscillation during the boreal summer, J. Clim., 17(4), 727–743. Huffman, G. J., R. F. Adler, P. Arkin, A. Chang, R. Ferraro, A. Gruber, J. Janowiak, A. McNab, B. Rudolf, and U. Schneider (1997), The Global Precipitation Climatology Project (GPCP) combined precipitation dataset, Bull. Am. Meteorol. Soc., 78(1), 5–20. Jiang, X., T. Li, and B. Wang (2004), Structures and mechanisms of the northward propagating boreal summer intraseasonal oscillation, J. Clim., 17(5), 1022–1039. Johnson, N. C., and S.-P. Xie (2010), Changes in the sea surface temperature threshold for tropical convection, Nat. Geosci., 3(12), 842–845. Kalnay, E., et al. (1996), The NCEP/NCAR 40-year reanalysis project, Bull. Am. Meteorol. Soc., 77(3), 437–471. Kemball-Cook, S., and B. Wang (2001), Equatorial waves and air-sea interaction in the boreal summer intraseasonal oscillation, J. Clim., 14(13), 2923–2942. Klingaman, N. P., P. M. Inness, H. Weller, and J. M. Slingo (2008), The importance of high-frequency sea surface temperature variability to the intraseasonal oscillation of Indian monsoon rainfall, J. Clim., 21(23), 6119–6140. Krishnamurti, T. N., and H. N. Bhalme (1976), Oscillations of a monsoon system. Part I. Observational aspects, J. Atmos. Sci., 33, 1937–1954. Krishnamurthy, V., and B. P. Kirtman (2009), Relation between Indian monsoon variability and SST, J. Clim., 22(17), 4437–4458. Krishnan, R., and C. Venkatesan (1997), Mechanisms of low frequency intraseasonal oscillations of the Indian summer monsoon, Meteorol. Atmos. Phys., 62(1), 101–128. Kulkarni, A., S. S. Sabade, and R. H. Kripalani (2009), Spatial variability of intra-seasonal oscillations during extreme Indian monsoons, Int. J. Clim., 29(13), 1945–1955. Lal, M., K. K. Singh, G. Srinivasan, L. S. Rathore, D. Naidu, and C. N. Tripathi (1999), Growth and yield responses of soybean in Madhya Pradesh, India to climate variability and change, Agric. For. Meteorol., 93(1), 53–70. Lawrence, D. M., and P. J. Webster (2002), The boreal summer intraseasonal oscillation: Relationship between northward and eastward movement of convection, J. Atmos. Sci., 59(9), 1593–1606. Liebmann, B., and C. A. Smith (1996), Description of a complete (interpolated) outgoing longwave radiation dataset, Bull. Am. Meteorol. Soc., 77, 1275–1277. Liess, S., D. E. Waliser, and S. D. Schubert (2005), Predictability studies of the intraseasonal oscillation with the ECHAM5 GCM, J. Atmos. Sci., 62, 3320–3336. Lin, A., T. Li, X. Fu, J.-J. Luo, and Y. Masumoto (2011), Effects of air-sea coupling on the boreal summer intraseasonal oscillations over the tropical Indian Ocean, Clim. Dyn., 37(11), 2303–2322. Lindzen, R. S., and A. Y. Hou (1988), Hadley circulations for zonally averaged heating centered off the equator, J. Atmos. Sci., 45, 2416–2427. Molnar, P., W. R. Boos, and D. S. Battisti (2010), Orographic controls on climate and paleoclimate of Asia: Thermal and mechanical roles for the Tibetan Plateau, Annu. Rev. Earth Planet. Sci., 38, 77–102. Neena, J. M., and B. N. Goswami (2010), Extension of potential predictability of Indian summer monsoon dry and wet spells in recent decades, Q. J. Roy Meteorol. Soc., 136(648), 583–592. Rajeevan, M., J. Bhate, J. D. Kale, and B. Lal (2006), High resolution daily gridded rainfall data for the Indian region: Analysis of break and active monsoon spells, Curr. Sci., 91(3), 296–306. Rao, S. A., H. S. Chaudhari, S. Pokhrel, and B. N. Goswami (2010), Unusual central Indian drought of summer monsoon-2008: Role of southern tropical Indian Ocean warming, J. Clim., 23(19), 5163–5174. Rao, S. A., A. R. Dhakate, S. K. Saha, S. Mahapatra, H. S. Chaudhari, S. Pokhrel, and S. K. Sahu (2012), Why is Indian Ocean warming consistently?, Clim. Change, 110(3), 709–719. Rayner, N. A., D. E. Parker, E. B. Horton, C. K. Folland, L. V. Alexander, D. P. Rowell, E. C. Kent, and A. Kaplan (2003), Global analyses of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century, J. Geophys. Res., 108(D14), 4407, doi:10.1029/ 2002JD002670. Reynolds, R. W., T. M. Smith, C. Liu, D. B. Chelton, K. S. Casey, and M. G. Schlax (2007), Daily high-resolution-blended analyses for sea surface temperature, J. Clim., 20(22), 5473–5496. Roeckner, E., et al. (2003), The atmospheric general circulation model ECHAM5: Part 1: Model description, Max-Planck-Institut für Meteorologie Rep. 349, Hamburg, Germany. Sabeerali, C. T., S. A. Rao, R. S. Ajayamohan, and R. Murtugudde (2012), On the relationship between Indian summer monsoon withdrawal and Indo-Pacific SST anomalies before and after 1976/1977 climate shift, Clim. Dyn., 39, 841–859. Sabeerali, C. T., A. Ramu Dandi, A. Dhakate, K. Salunke, S. Mahapatra, and S. A. Rao (2013), Simulation of boreal summer intraseasonal oscillations in the latest CMIP5 coupled GCMs, J. Geophys. Res. Atmos., 118, 4401–4420, doi:10.1002/jgrd.50403. Sengupta, D., B. N. Goswami, and R. Senan (2001), Coherent intraseasonal oscillations of ocean and atmosphere during the Asian summer monsoon, Geophys. Res. Lett., 28(21), 4127–4130, doi:10.1029/2001GL013587. Sikka, D. R., and S. Gadgil (1980), On the maximum cloud zone and the ITCZ over Indian, longitudes during the southwest monsoon, Mon. Weather Rev., 108(11), 1840–1853. Smith, T. M., R. W. Reynolds, T. C. Peterson, and J. Lawrimore (2008), Improvements to NOAA’s historical merged land-ocean surface temperature analysis (1880–2006), J. Clim., 21(10), 2283–2296. Sperber, K. R., J. M. Slingo, and H. Annamalai (2000), Predictability and the relationship between subseasonal and interannual variability, Q. J. R. Meteorol. Soc., 126, 2545–2574. Suhas, E., and B. N. Goswami (2008), Regime shift in Indian summer monsoon climatological intraseasonal oscillations, Geophys. Res. Lett., 35, L20703, doi:10.1029/2008GL035511. Taylor, K. E., R. J. Stouffer, and G. A. Meehl (2012), An overview of CMIP5 and the experiment design, Bull. Am. Meteorol. Soc., 93(4), 485–498. Teng, H., and B. Wang (2003), Interannual variations of the boreal summer intraseasonal oscillation in the Asian-Pacific region, J. Clim., 16(22), 3572–3584. Trenberth, K. E. (2011), Changes in precipitation with climate change, Clim. Res., 47, 123–138. Uppala, S. M., et al. (2005), The ERA-40 re-analysis, Q. J. R. Meteorol. Soc., 131(612), 2961–3012.

SABEERALI ET AL.

©2014. American Geophysical Union. All Rights Reserved.

5202

Journal of Geophysical Research: Atmospheres

10.1002/2013JD021261

Vecchi, G., and D. Harrison (2002), Monsoon breaks and subseasonal sea surface temperature variability in the Bay of Bengal, J. Clim., 15, 1485–1493. Vecchi, G. A., B. J. Soden, A. T. Wittenberg, I. M. Held, A. Leetmaa, and M. J. Harrison (2006), Weakening of tropical Pacific atmospheric circulation due to anthropogenic forcing, Nature, 441(7089), 73–76, doi:10.1038/nature04744. Vecchi, G. A., and B. J. Soden (2007), Global warming and the weakening of the tropical circulation, J. Clim., 20(17), 4316–4340, doi:10.1175/ JCLI4258.1. Wang, B., and X. Xie (1997), A model for the boreal summer intraseasonal oscillation, J. Atmos. Sci., 54(1), 72–86. Wang, B., P. Webster, K. Kikuchi, T. Yasunari, and Y. Qi (2006), Boreal summer quasi-monthly oscillation in the global tropics, Clim. Dyn., 27(7), 661–675. Wang, W., M. Chen, and A. Kumar (2009), Impacts of ocean surface on the northward propagation of the boreal summer intraseasonal oscillation in the NCEP climate forecast system, J. Clim., 22(24), 6561–6576. Webster, P. J. (1983), Mechanisms of monsoon low-frequency variability: Surface hydrological effects, J. Atmos. Sci., 40(9), 2110–2124. Webster, P. J., V. O. Magaña, T. N. Palmer, J. Shukla, R. A. Tomas, M. Yanai, and T. Yasunari (1998), Monsoons: Processes, predictability, and the prospects for prediction, J. Geophys. Res., 103(C7), 14,451–14,510. Wu, L., et al. (2012), Enhanced warming over the global subtropical western boundary currents, Nat. Clim. Change, 2(3), 161–166. Xie, P., and P. A. Arkin (1997), Global precipitation: A 17-year monthly analysis based on gauge observations, satellite estimates, and numerical model outputs, Bull. Am. Meteorol. Soc., 78(11), 2539–2558. Xie, P., J. E. Janowiak, P. A. Arkin, R. Adler, A. Gruber, R. Ferraro, G. J. Huffman, and S. Curtis (2003), GPCP pentad precipitation analyses: An experimental dataset based on gauge observations and satellite estimates, J. Clim., 16(13), 2197–2214. Yasunari, T. (1979), Cloudiness fluctuations associated with the Northern Hemisphere summer monsoon, J. Meteorol. Soc. Jpn., 57(3), 227–242. Yasunari, T. (1980), A quasi-stationary appearance of 30 to 40 day period in the cloudiness fluctuations during the summer monsoon over India, J. Meteorol. Soc. Jpn., 58, 225–229. Zhang, M., and H. Song (2006), Evidence of deceleration of atmospheric vertical overturning circulation over the tropical Pacific, Geophys. Res. Lett., 33, L12701, doi:10.1029/2006GL025942. Zhou, L., and R. Murtugudde (2014), Impact of northward-propagating intraseasonal variability on the onset of Indian summer monsoon, J. Clim., 27, 126–139. Zveryaev, I. I. (2002), Interdecadal changes in the zonal wind and the intensity of intraseasonal oscillations during boreal summer Asian monsoon, Tellus A, 54(3), 288–298.

SABEERALI ET AL.

©2014. American Geophysical Union. All Rights Reserved.

5203