Dynamics of 'Internal' Interannual Variability of the Indian ... - CiteSeerX

Climate Dynamics manuscript No. (will be inserted by the editor)

Dynamics of ’Internal’ Interannual Variability of the Indian Summer Monsoon in a GCM B. N. Goswami, Prince K. Xavier Centre for Atmospheric and Oceanic Sciences Indian Institute of Science, Bangalore, 560012, INDIA

Abstract

The interannual variability (IAV) of the tropical climate is partially driven by anomalous

boundary conditions as well as by the internal low frequency feedbacks within the system. As the internal IAV is less predictable, predictability of the system depends on the relative contributions of these two components, that are difficult to separate. Predictability studies of the Indian summer monsoon indicate that at least 50% of IAV of the Indian monsoon is of internal origin. However, mechanisms responsible for the internal IAV of the monsoon have not been clearly identified. Here, using a general circulation model (GCM) driven by fixed annual cycle boundary forcing, the ’internal’ variability of the Indian summer monsoon is simulated and compared with the observed interannual variability of the seasonal (JuneSeptember) mean. It is shown that the internal interannual variability of the summer monsoon simulated by the model is comparable in amplitude and similar in spatial structure to those in observations. An attempt has been made to determine the causes of these ’internal’ interannual variability. The hypothesis that the monsoon intraseasonal oscillations (ISO) are responsible for the IAV, is tested in this study. The spatial and temporal characteristics of simulated intraseasonal variability are examined and are in good agreement to those observed. The spatial patterns of ISOs and IAV are found to be similar and it is shown that higher frequency of occurrance of active (break) phase of ISOs gives rise to strong

2

B. N. Goswami, Prince K. Xavier

(weak) monsoons. This hypothesis is confirmed by a thorough examination of the anomalies of seasonal mean and the seasonal mean of intraseasonal anomalies. It is concluded that the internal IAV of the Indian summer monsoon is governed primarily by projection of ISOs on the IAV mode. However, this contribution explains only about 50% of the simulated IAV, implying that the nonlinear processes, such as the soil moisture feedback, also play an equally important role in amplifying the internal interannual variations of the Indian summer monsoon driven by the monsoon ISOs.

1 Introduction

With a large fraction of world population depending on the monsoon rainfall, prediction of summer monsoon rainfall at least one season in advance assumes great significance. For over a century, attempts have been made to predict seasonal mean monsoon rainfall using empirical techniques involving local and global antecedent parameters that correlate with the monsoon rainfall (Blandford, 1884; Walker, 1923, 1924; Gowarikar et al., 1989; Sahai et al., 2003). The linear and nonlinear regression models as well as the neural network based models (e.g. Goswami and Srividya (1996)) perform reasonably well when the monsoon is close to normal but fails to predict the extremes with useful skill. This was illustrated by the failure of almost all empirical models in predicting the 2002 drought. Another intrinsic limitation of the empirical techniques arises due to interdecadal variation of the correlations between predictors and monsoon rainfall (Kriplani and Kulkarni, 1997; Kumar et al., 1997; Krishnamurthy and Goswami, 2000). Dynamical prediction of the seasonal mean monsoon using state of the art climate models, therefore, offers a logical alternative to empirical forecasting. Unfortunately, however, the dynamical models have not demonstrated skill better than the empirical techniques so far. Multi-model super-ensemble forecasting (Krishnamurti et al., 1999, 2000, 2001) seem to show promise of improving the dynamical forecasts beyond the skill of individual model’s skill. The inability of the state of the art climate model’s in predicting the seasonal mean monsoon appears to be due to two factors. Firstly, even though climate models have improved over the last three decades

Dynamics of ’Internal’ Interannual Variability of the Indian Summer Monsoon in a GCM

3

in simulating the global climate in general, most climate models still have major systematic errors in simulating the seasonal mean Indian summer monsoon and its interannual variability (Sperber and Palmer, 1996; Saji and Goswami, 1997; Gadgil and Sajani, 1998; Kang et al., 2002). Secondly, following the seminal work of Charney and Shukla (1981), even though it has been shown (Shukla and Wallace, 1983; Lau, 1985; Kumar and Hoerling, 1995; Shukla, 1998; Anderson et al., 1999; Fennessy and Shukla, 1999) that the tropical climate is largely driven by anomalous boundary conditions (ABC) and much less sensitive to initial conditions, the Indian summer monsoon seems to be an exception within the tropics and seems to be quite sensitive to initial conditions (Sperber and Palmer, 1996; Sperber et al., 2000; Krishnamurthy and Shukla, 2000; Cherchi and Navarra, 2003).

What makes the simulation of the Indian summer monsoon so sensitive to initial conditions and physical parameterizations and hence probably the most difficult tropical system to predict? Answer to this question lies in the fact that the interannual variability of the tropical climate is partially governed by internally generated low frequency (LF) oscillations in addition to being forced by slowly varying ABC (such as the sea surface temperature, soil moisture, sea ice etc). Such internal LF variability could, in principle, be generated through nonlinear scale interactions between high frequency oscillations, interactions between flow and topography, feedback between organized convection and large scale dynamics etc. The predictability of the climate depends on relative contributions of slowly varying boundary forcing or ’external’ component and LF ’internal’ component to the interannual variability. Goswami (1998) made an estimate of contributions from the two components to the interannual variability in the tropics within the framework of a general circulation model (GCM) and shown that as large as 50% of the model’s monsoon interannual variability could be governed by the ’internal’ component. Could this estimate be influenced by the model biases? How much of the observed monsoon interannual variability is governed by ’internal’ processes? Using long daily observations, Ajayamohan and Goswami (2003) show that both for monthly mean as well as seasonal mean summer monsoon climate about 50% of the observed variability is governed by ’internal’ processes. Both these studies also show that there are certain regions in the

4

B. N. Goswami, Prince K. Xavier

tropics, such as central and eastern equatorial Pacific, central equatorial Atlantic etc., where contribution of ’internal’ variability is small and interannual variability is largely governed by ’external’ boundary forcing. These regions are the regions where GCM experiments also show very little sensitivity to initial conditions (e.g. Shukla (1998)).

What is responsible for the ’internal’ interannual variability in the GCM’s and in observations? Clear understanding of mechanisms responsible for the ’internal’ interannual variability is imperative for making progress in seasonal monsoon prediction. Circumstantial evidence of connection between monsoon intraseasonal oscillations (ISO’s) and interannual variability (IAV) was presented by Fennessy and Shukla (1994) and Ferranti et al. (1997) when they showed similarity in spatial structures of ISO and IAV of the monsoon. Indian summer monsoon is known to have vigorous ISO’s (Yasunari, 1979, 1980; Sikka and Gadgil, 1980; Krishnamurti and Bhalme, 1976; Krishnamurti and Subrahmanyam, 1982; Krishnamurti and Ardunay, 1980; Murakami et al., 1984) having time scales of 10-20 days and 30-60 days. Goswami (1994) speculated that interannual variations of statistics of monsoon ISO (such as preferred periodicity or frequency of occurrence etc) could give rise to interannual variability of the seasonal mean. Palmer (1994) proposed a paradigm in which frequency of occurrence of chaotic ISO could result in interannual variability of the seasonal mean. Using long records of daily circulation and convection data, Goswami and Ajayamohan (2001) showed that the intraseasonal and interannual variability of the summer monsoon are governed by a common mode of spatial variability. With this establishment of similarity in spatial structure between the two, frequency of occurrence of ISO could influence the seasonal mean and its interannual variability. They further showed that strong (weak) monsoon is characterized by higher probability of occurrence of active (break) conditions. Sperber et al. (2000) also showed that the ISO and IAV of monsoon are governed by a common spatial mode of variability and frequency of occurrence of the ISO could influence the seasonal mean and its IAV. The IAV of the observed monsoon precipitation or circulation arise from a combined effect of boundary forced variability and ’internal’ LF variability. While the studies just mentioned establish that frequency of occurrence of ISO and IAV of monsoon are

Dynamics of ’Internal’ Interannual Variability of the Indian Summer Monsoon in a GCM

5

related, a part of frequency change of ISO could arise due to regime changes in boundary forced climate. Mechanism for the ’internal’ component of the LF variability could not be clearly identified from these studies. The objective of the present study is to unravel mechanism for pure ’internal’ interannual variability of the Indian summer monsoon. This is best done within the context of a GCM. For this purpose we integrate a GCM for a reasonably long period with no external interannual forcing. The model simulations show significant IAV of the summer monsoon similar in structure to those of observed monsoon IAV. Careful analysis demonstrate that the simulated ’internal’ IAV is generated by frequency changes of the monsoon ISO. The model experiment and data used for verification are described in section 2. The model climatology is discussed in Section 3 and simulated IAV of Indian summer monsoon is discussed in Section 4. The nature of the simulated monsoon ISO is described in section 5 while relationship between ISO and ISV of simulated monsoon is discussed in Section 6. Results are summarized in section 7.

2 The Model Experiment and Data

The standard version of Community Climate Model version 3 (CCM3) (Kiehl et al., 1998) developed by National Centre for Atmospheric Research (NCAR) is used in this study. It is a global spectral GCM with T42 horizontal resolution (≈ 2.8◦ × 2.8◦ Gaussian grid) and 18 levels in the vertical with top of the model at 4.8 hPa. The model time step is 1200 s (20 min). Deep convection is simulated by the mass-flux scheme of Zhang and McFarlane (1995) and the triplet convection scheme of Hack (1994) simulates shallow convection. An explicit, non-local atmospheric boundary layer (ABL) parameterization is incorporated into the vertical diffusion parameterization (Holtslag and Boville, 1993). The absorptivity-emissivity formulation of Ramanathan and Downey (1986) is employed to represent the long wave radiative transfer. A control simulation for 25 years is conducted with December 1 initial condition and climatological monthly mean sea surface temperature (SST). The climatological mean SST data is derived from Reynolds and Smith (1994). Since solar forcing as well as the boundary forcing (SST) are only annually varying,

6

B. N. Goswami, Prince K. Xavier

no external forcing with interannual period forces the model simulations. So, the simulated interannual atmospheric variability may be regarded as variability arising due to internal feedbacks within the system. We refer to it as ’internal’ variability. The daily mean outputs are examined in this study. We have used low level circulation and precipitation data from observations. The circulation data is derived from National Centers for Environmental Prediction/National Center for Atmospheric Research (NCEP/NCAR) reanalysis (Kalnay et al., 1996; Kistler et al., 2001). The reanalysis data is generated using a ”frozen” state of the art data assimilation system and model. The rainfall data is taken from Climate Prediction Center Merged Analysis of Precipitation (CMAP) (Xie and Arkin, 1996). The pentad CMAP used same algorithm and data sources as the monthly CMAP and is obtained from Climate Prediction Center, Washington D.C., U.S.A. The pentad data is linearly interpolated to daily values to enable compatibility with daily circulation data and daily mean outputs from the model. Both circulation data and CMAP are taken for a period of 20 years (1979-1998). The observed data is used to compare the characteristics of the model simulated seasonal mean climate as well as the intraseasonal variability. The amplitudes of the interannual variability in both these parameters are compared. For this purpose, the interannual anomalies of the seasonal mean for both observations and simulations are derived in the following way. Let Pi,j (x, y), i = 1, 2, ..., N , j = 1, 2, ..., M be the variable of interest, where N =365 days and M =Number of years used in the analysis. The seasonal mean is defined as

P j (x, y) =

30Sep 1 X Pi,j (x, y), j = 1, 2, ...M 122

(1)

i=1Jun

The climatological seasonal mean may be calculated as

P (x, y) =

M 1 X P j (x, y) M j=1

(2)

and the interannual anomalies of the seasonal mean is defined as 0

P j (x, y) = P j (x, y) − P (x, y)

(3)

Dynamics of ’Internal’ Interannual Variability of the Indian Summer Monsoon in a GCM

7

The intraseasonal component of precipitation and low level circulation anomalies from the model and observations are calculated in the following manner. The daily anomalies may be calculated as deviation ∗ of the daily values from a climatological mean annual cycle. Let Pi,j (x, y), i = 1, 2, ..., 365, j = 1, 2, ..., M

be annual cycle (defined as annual mean + first 3 harmonics) for the j th year, where, M is the number of years used in the analysis (M =25 for model and M =20 for observations). The climatological mean annual cycle is Pi∗ (x, y) =

M 1 X ∗ P (x, y), i = 1, 2, ..., 365 M j=1 i,j

(4)

The daily anomalies are defined as 0 Pi,j (x, y) = Pi,j (x, y) − Pi∗ (x, y), i = 1, 2, ..., 365; j = 1, 2, ..., M

(5)

These daily anomalies are extracted for the period 1 June - 30 September and combined together. Power spectrum analysis was carried out on unfiltered anomalies over certain regions in order to identify the simulated intraseasonal modes. The daily anomalies are filtered using a 10-20 day as well as a 3090 day bandpass Lanczos filter (Duchon, 1979) to compare the characteristics of the different modes of monsoon intraseasonal variability simulated by the model to that is observed. Also a 10-90 day bandpass filter is applied to extract the total intraseasonal anomaly consisting of both 10-20 day and 30-90 day f0 k,j (x, y), k = 1, 2, ..., M × 122, j = 1, 2, ..., M be the 10-90 day filtered anomalies for JJAS modes. Let P for M years. The JJAS seasonal mean of intraseasonal anomalies may be written as f0 j (x, y) = 1 P 122

j×122 X

f0 k,j (x, y), P

j = 1, 2, ..., M

(6)

k=(j−1)×122+1

f0 j (x, y) explains the contribution from the intraseasonal timescale to the seasonal timescale which can P partially contribute to the interannual variability. This part is discussed in detail in the following sections.

3 Model Climatology of Indian Summer Monsoon The objective of the study is to use the simulated ’internal’ interannual variability of summer monsoon by the GCM to gain insight regarding the origin of observed ’internal’ interannual variability. For the exercise

8

B. N. Goswami, Prince K. Xavier

for our objective to be meaningful, CCM3 should simulate the summer monsoon climatology reasonably well. While Meehl and Arblaster (1998), described some broad aspects of simulated summer climatology of the CCM3, a detailed evaluation of simulation of summer monsoon climatology of the model has not been done. Before proceeding to investigate the cause of the simulated interannual summer monsoon variability, we examine the fidelity of CCM3 in simulating the summer monsoon climatology. JuneSeptember climatology of precipitation, 850 hPa winds and 200 hPa winds constructed from the 25-year simulations are shown in Fig.1A,B,C while similar climatology of observed precipitation from CMAP and winds from NCEP/NCAR reanalysis are shown in Fig.1D,E,F. With the simulation of the Bay-of-Bengal (BoB) maximum in precipitation located close to 20◦ N (Fig.1A) as in observations (Fig.1D), a major systematic error is overcome by CCM3. The western coast of India maximum in precipitation as well as the secondary maximum just south of the equator between 80◦ E and 100◦ E are also well simulated by the model. The model tends to produce higher precipitation (between 10% and 20%) in all three observed high precipitation regions. Significant systematic error in simulation of climatological mean summer precipitation, however, occurs in the equatorial region between 55◦ E and 70◦ E and 5◦ N and 5◦ S where the model simulated precipitation is nearly twice as large as the observed. The pattern correlation between the simulated and observed precipitation climatology over the region between 40◦ E -110◦ E and 20◦ S-25◦ N is 0.61.

The model simulates location of low level jet, the cross equatorial flow and the south equatorial easterlies well (Figs.1B and 1E). However, consistent with slightly stronger simulated monsoon precipitation, the simulated maximum wind speed of the low level jet (LLJ) is about 10% larger than that observed. The observed monsoon trough as derived from conventional radiosonde observations (Rao, 1976) is not well represented by the NCEP reanalysis. While the monsoon trough is better simulated by the model than in NCEP, the model simulated westerlies in south China Sea are much weaker than observed. The pattern correlation between the simulated and observed (NCEP) zonal and meridional wind climatology at 850 hPa are 0.88 and 0.85 respectively. The strength of the upper level easterly jet is close to the ob-

Dynamics of ’Internal’ Interannual Variability of the Indian Summer Monsoon in a GCM

9

served one over the equatorial Indian Ocean (Fig.1C, and 1F). The turning of the easterlies to westerlies (region of minimum wind speed) around 15◦ S is also well simulated. However, the maximum strength of the model simulated easterly jet tends to occur around 40◦ E compared to 75◦ E in observations. The westward shift of the maximum of the easterly jet speed in the model climate may be partly due to unrealistic summer precipitation simulated by the model over Saudi Arabia (Fig.1A). The local Hadley circulation with ascending motion in the region (around 20◦ N) would be associated with equator ward flow at upper level leading to strong easterly flow at these longitudes. The pattern correlation between simulated and observed zonal and meridional winds at 200 hPa are 0.94 and 0.71 respectively.

The onset of the summer monsoon with rather sharp increase of kinetic energy of the low level winds and rapid northward shift and establishment of the tropical convergence zone (TCZ) over the land are important component of annual evolution of the Indian monsoon. The model’s fidelity in simulating the annual evolution of the Indian monsoon is tested in Figures 2 and 3. Daily climatological mean precipitation simulated by the model averaged over 70◦ E-100◦ E as a function of latitude (Fig.2A) is contrasted with similar plot for observed precipitation (CMAP, Fig.2B). It is noteworthy that during the summer monsoon season (June-September), the model simulates the the two preferred latitudinal locations of precipitation with a minimum in between quite well except that the simulated precipitation is about 10-20% stronger. The simulated annual evolution of the precipitation belt has a significant systematic bias during northern winter (November-February) when it remains at around 5◦ N (Fig.2A) and do not move southward to about 5◦ S as is observed (Fig.2B). The model simulation of precipitation is characterized by a rather gradual advance of rain belt from beginning of May establishing the monsoon in June over the land (Fig.2A) and lacks the relatively sudden northward jump of the rain belt in the beginning of June as seen in observations (Fig.2B). This bias is also seen in the simulated KE of the low level jet as shown in Fig.3. The model monsoon onset takes place about 2 weeks earlier than observed and withdraws about one week prior to observed withdrawal. Consistent with stronger than observed low level winds (Fig.1B, 1E), the KE of the LLJ is much stronger than observed during June and July.

10

B. N. Goswami, Prince K. Xavier

4 Simulated Interannual variability (IAV) of Indian summer Monsoon

In the absence of any external interannual forcing, any significant IAV of the monsoon simulated by the model must arise purely from internal atmospheric dynamics. Does the model simulate significant IAV of the monsoon? Even if it does, does the spatial structure of the simulated IAV have any similarity with that of the observed IAV of the summer monsoon? In this section, we address these two questions and show that the model simulates significant IAV of the summer monsoon and that the spatial structure of this variability is quite similar to the observed IAV of the summer monsoon. For this purpose, we construct several indices representing the strength of the summer monsoon. First index is the extended Indian monsoon rainfall index (EIMR) represented by June-September mean precipitation averaged over 70◦ E110◦ E, 10◦ -25◦ N similar to the one suggested by Goswami et al. (1999). The second index is monsoon Hadley index (MH) defined as shear of the meridional winds between lower and upper atmosphere (V850 -V200 ) averaged over the region 70◦ E-110◦ E, 10◦ -30◦ N (Goswami et al., 1999). As the low level winds over the Arabian sea are strongly related to the precipitation over the monsoon region, a third summer monsoon index is constructed as the KE of the LLJ (KELLJ) defined as the seasonal mean KE of winds at 850 hPa level averaged over 50◦ E-65◦ E, 5◦ N-15◦ N. Long term mean and interannual standard deviation of these indices are shown in Table-I and compared with corresponding values from observations. The observed EIMR is calculated from CMAP data while the observed circulation indices are constructed from NCEP reanalysis. As could be expected from the discussions in the previous Section, the long term mean of the indices are close to the observed values. The internal interannual variance of EIMR and MH simulated by CCM3 represents 50%-60% of observed interannual variance. This is consistent with estimate of internal variance of monsoon using GFDL climate model (Goswami, 1998) and estimates made using observations (Ajayamohan and Goswami, 2003). Thus, the model does simulate significant IAV of the summer monsoon. The IAV of the KELLJ is, however, much stronger than that observed. We shall show little later that EIMR and KELLJ are well correlated in the model simulations. Therefore, it is a bit puzzling to note that IAV of KELLJ is stronger than observed while that of EIMR is weaker

Dynamics of ’Internal’ Interannual Variability of the Indian Summer Monsoon in a GCM

11

than observed. The reason for this discrepancy will be clearer when we examine the spatial structure of the IAV. The IAV of the indices are normalized by their own standard deviation are shown in Fig.4. It is clear that the three indices are correlated with each other. The correlation coefficient between EIMR and KELLJ is 0.81, while that between EIMR and MH is 0.59 and that between KELLJ and MH is 0.50. To examine the large scale structure associated with the simulated internal IAV of the summer monsoon, composite of precipitation and 850 hPa winds corresponding to simulated strong and weak monsoons are created. Strong (weak) monsoons are selected from the 25 year simulations depending on whether the normalized EIMR index is greater than 1 (less than -1) (Fig.4). Model simulated composites of strong and weak summer monsoon are compared with similar strong and weak monsoon composites of precipitation and 850 hPa winds from observations. Observed precipitation is taken from CMAP while observed winds are taken from NCEP reanalysis. Strong (weak) observed monsoon being defined by normalized observed EIMR being greater than 1 (less than -1). Such strong summer monsoon composites of precipitation and 850 hPa winds from model and observations are shown in Fig.5 while similar composites for weak monsoon are shown in Fig.6. To start with, we note that the internally generated IAV of the Indian summer monsoon has large spatial scale similar to the observed summer monsoon interannual variability. The simulated precipitation anomalies have a meridional bi-model structure similar to the one associated with the observed strong (weak) summer monsoon anomalies. We noted earlier (Fig.1A, 1D and Fig.2) that the location of the simulated northern climatological mean precipitation band correspond well with observations. However, the interannual precipitation anomalies associated with strong (weak) monsoon simulated by the model seem to move only up to 15◦ N as against about 25◦ N in observations. The low level winds associated with the internal strong (weak) monsoon (Fig.5B,6B) strengthen (weaken) the climatological summer mean flow (Fig.1B) as in observation. Consistent with the observation made in Table-I that amplitude of simulated IAV of KELLJ is larger than observed, we note that the strength of the wind anomalies over Arabian Sea LLJ area is stronger than that observed over the same area during strong as well as weak monsoon (Fig.5B,D and 6B,D).

12

B. N. Goswami, Prince K. Xavier

This feature can be related an anomalous feature of simulated rainfall anomalies. Note from Fig.5A and 6A that strong (weak) simulated monsoon is associated with a region of negative (positive) precipitation anomaly between 60◦ E-70◦ E centered around 5◦ S not seen in observations (Fig.5C, 6C). This essentially represents decrease (increase) of large simulated climatological precipitation in this region during strong (weak) monsoon. Linear response of the atmosphere to positive precipitation anomalies to the north (western coast of India and BoB) and negative anomalies to the south during a strong monsoon is consistent with stronger wind anomalies over the LLJ area. During simulated weak monsoon the northsouth dipole pattern of precipitation anomalies reverse in sign and gives rise to strong wind anomalies of opposite sign. This process leads to stronger IAV simulated wind anomalies in this region. Thus, even though the GCM simulations have some biases, climatology of GCM simulated monsoon is reasonable and it simulates internal IAV of summer monsoon comparable in amplitude to observed amplitude of IAV of summer monsoon. Further, the spatial structure of the simulated internal IAV is similar to that of the observed IAV of the summer monsoon. What is responsible for the significant internal IAV of the monsoon in the GCM? To gain insight on this question, we shall examine the nature of the simulated intraseasonal variability (ISV) and its relationship with simulated IAV. The working hypothesis is that if the spatial structure of ISV and the seasonal mean is similar, changes in frequency of occurrence of the intraseasonal oscillations could give rise to IAV.

5 Simulated Monsoon Intraseasonal Variability (ISV)

Both temporal and spatial characteristics of the simulated ISO’s during northern summer season (June - September) are examined here and compared with corresponding characteristics of observed ISO’s. To examine the temporal characteristics, a zonal wind time series is created with daily anomalies between June 1 and September 30 for all simulated years averaged between 80◦ E-90◦ E, 5◦ N-15◦ N. Observed zonal wind time series is obtained from daily anomalies of zonal winds at 850 hPa from NCEP reanalysis. To examine the intraseasonal temporal characteristics of simulated precipitation anomalies, a precipitation

Dynamics of ’Internal’ Interannual Variability of the Indian Summer Monsoon in a GCM

13

time series is created with daily anomalies between June 1 and September 30 for all years averaged between 85◦ E-95◦ E, 15◦ N-20◦ N. The power spectrum of the precipitation time series is shown in Fig.7A. The model simulated precipitation anomalies have two preferred periodicities, one around 10-20 days and another around 30-60 days. As CMAP is a pentad data set and as there is no good daily precipitation data over this region for a sufficiently long period of time, a similar spectra of observed rainfall is not shown. The power spectra of the zonal wind time series from CCM3 and from observation are shown in Fig.7B and 7C. The model captures the peak around 10-20 days as well as around 30-60 days. In order to study the spatial structure and other characteristics of the two range of preferred periodicities in some detail, both precipitation and zonal wind anomalies are band-pass filtered using a Lanczos filter to retain periodicities between 10 and 20 days and between 30 and 90 days. Percentage of total daily variance accounted by the 10-20 day mode and the 30-90 day mode for both precipitation and zonal wind during the summer monsoon season are shown in Fig.8. It is noted that each mode explains approximately 30% of the total daily variance in the core regions in both precipitation and zonal winds. The spatial distribution of variance explained by each mode in precipitation and low level wind field is similar.

To study the spatial structure of the 10-20 day mode, a reference time series is created by averaging 10-20 day filtered zonal winds over a box between 85◦ E-90◦ E and 5◦ N-10◦ N during the summer monsoon season (June 1 to 30 September) for all years of simulations as well as for 20 years (1979-1998) of NCEP . Lag regressions of 10-20 day filtered zonal and meridional winds are then constructed with respect to the reference time series both for model simulations as well as for observation (NCEP). Simultaneous regressed vector wind pattern from the model simulations and NCEP are shown in Fig.9B and 9E while those with 7 days lead with respect to the reference time series are shown in Fig.9A and 9D. The amplitude and spatial pattern of the simulated 10-20 day mode is quite similar to that observed. The westward propagation of the mode seen in a plot of regressed zonal winds averaged over 5◦ N-15◦ N as a function of longitudes and lags (Fig.9C) is quite similar to the westward propagation of the observed mode (Fig.9F). The spatial structure of the mode in model simulation as well as in observations resemble the

14

B. N. Goswami, Prince K. Xavier

gravest meridional mode equatorial Rossby wave shifted to north by about 5◦ (Chatterjee and Goswami, 2003, preprint available at http://caos.iisc.ernet.in/faculty/bng/qbm.pdf). The horizontal structure of the 30-90 day mode is studied in a similar manner by constructing a reference time series of 30-90 day filtered zonal winds averaged over 80◦ E-90◦ E and 10◦ N-15◦ N and calculating lag regressions of 30-90 day filtered zonal and meridional winds everywhere. The simultaneous vector wind pattern for the mode from model simulations as well as from observation are shown in Fig.10A and 10B respectively. Similar vector wind pattern with lead of 23 days with respect to the reference time series are shown in Fig.10C and 10D respectively. The model simulates the observed horizontal structure of the mode quite well which is much larger than that of the 10-20 day mode (see Fig.9). The correlation between model simulated and observed regressions patterns at zero lag for zonal and meridional winds are 0.78 and 0.72 respectively. The 30-90 day mode is known to have northward propagation over the Indian monsoon region. Laglatitude plot of regression of zonal winds averaged over 60◦ E-90◦ E are shown in Fig.10E and 10F from model simulations and NCEP respectively. The simulated 30-90 day mode has northward propagation in the Indian monsoon region similar to that observed. Thus, the model is successful in simulating the temporal and spatial characteristics of the observed ISO’s during northern summer over the Indian monsoon region reasonably well. In the next section, we examine if ISO’s could influence the seasonal mean monsoon and result in the simulated IAV of the seasonal mean monsoon.

6 Mechanism for Internal IAV of simulated Indian Monsoon

In this section, we try to understand the mechanism for the simulated internal IAV of the Indian summer monsoon. As we mentioned in the Section 4, the working hypothesis is that if the spatial structure of the ISO is similar to that of the seasonal mean, higher frequency of occurrence of active (weak) phase of the ISO during the monsoon season could give rise to a strong (weak) monsoon. If ISO’s were a single frequency sinusoidal oscillations, this could not happen. The observed and simulated ISO’s, however, have

Dynamics of ’Internal’ Interannual Variability of the Indian Summer Monsoon in a GCM

15

a broad band spectrum with periods ranging from 10 to 90 days. The temporal evolution is, therefore, aperiodic. This makes it possible for the probability of occurrence of active and weak phases to be different. In this section, we define ISV variability to contain all periods between 10 and 90 days. As a result, all data are filtered with 10-90 day band-pass Lanczos filter. All the results presented here are based on such a 10-90 day filtered data.

The large scale nature of the spatial structure if ISV was already noted in Fig.10. Here we draw attention to the fact that the structure of ISO mode is quite similar to the seasonal mean (Fig.1B). In the active phase (Fig.10A), the ISO strengthens the large scale monsoon circulation while in the weak phase (Fig.10B), it weakens the same. Hence, ISO’s could influence the IAV of the seasonal mean. If ISO’s lead to IAV of the seasonal mean, the spatial structure of IAV of the seasonal mean should also be similar to that of the ISO’s. Therefore, the spatial structure of the dominant mode of IAV is compared with that of the ISV. For this purpose, an empirical orthogonal function (EOF) analysis of interannual anomalies of the seasonal mean precipitation and zonal winds at 850 hPa (U850 ) were carried out. The first EOF of IAV for precipitation and U850 are shown in Fig.11A and Fig.11B. Similarly, the dominant mode intraseasonal variability of precipitation and U850 are found by carrying out EOF analysis of intraseasonal filtered anomalies during the summer season (1 June - 30 September) for all years taken together. The first EOF of ISV of precipitation and U850 are shown in Fig.11C and Fig.11B. It is clear that a common spatial mode governs both intraseasonal and interannual variability of precipitation as well as low level winds. The pattern correlation between the ISV and IAV for precipitation and U850 calculated between 50◦ E-110◦ E, 15◦ S-25◦ N are 0.82 and 0.80 respectively.

The above evidence supports the hypothesis that the ISO’s are responsible for IAV of the seasonal mean. This conclusion is further supported by the fact that the regions of high ISV are also regions of high IAV of the seasonal mean. This is shown in Fig.12 where the spatial distribution of variance of intraseasonal oscillations averaged over all 25 summer seasons for precipitation and U850 are compared with

16

B. N. Goswami, Prince K. Xavier

their counterpart of interannual variance of the seasonal mean. Good similarity between the intraseasonal and interannual variance is evident in both fields. To examine whether the cumulative effect of the ISO’s lead to the anomaly of the seasonal mean, seasonal mean (1 June - 30 September) of the ISO precipitation anomaly for the year 7 of simulations is plotted in Fig.13A and compared with anomaly of seasonal mean precipitation for the same year (Fig.13B). The two patterns are almost identical (pattern correlation is better than 0.99) although the seasonal mean of ISO anomalies are weaker than the anomaly of the seasonal mean by nearly a factor of two. The close correspondence between the pattern of seasonal mean of ISO anomalies and anomalies of seasonal mean is not unique to the particular year under consideration, but is characteristic feature of all years of simulations. The pattern correlation between the seasonal mean of ISO anomalies and anomaly of seasonal mean for precipitation and U850 over the full domain between 40◦ E-120◦ E, 20◦ S-30◦ N for all 25 years of simulations are shown in Fig.14. The mean pattern correlation of all years is approximately 0.98. Thus, everywhere in the domain, residual of intraseasonal oscillations result in variability of the seasonal mean. This is further illustrated by the difference in frequency of occurrence of intraseasonal precipitation anomalies during strong and weak monsoon years in the model simulations. The frequency distribution of intraseasonal precipitation anomalies averaged over 70◦ E-95◦ E, 10◦ N-25◦ N for three strong monsoon years (years 7, 14 and 15) and three weak monsoon years (years 12, 18 and 19) are shown in Fig.15. It is clear that strong (weak) monsoon in the model is characterized by higher frequency of occurrence of positive (negative) intraseaonal anomalies. The high pattern correlation between the seasonal mean of ISO anomalies and anomaly of seasonal mean is not confined to only the Indian monsoon region of the model but remains very high (0.96) even over the full tropical belt between 30◦ S-30◦ N, 0◦ -360◦ . The result is also true over the extratropical simulations. Thus, the primary mechanism for producing IAV of the model climate is essentially residual influence of the ISV. If this was the only mechanism, full interannual anomalies of seasonal mean should be accounted for by the seasonal mean of the intraseasonal anomalies. However, we have seen in Fig.13

Dynamics of ’Internal’ Interannual Variability of the Indian Summer Monsoon in a GCM

17

that the mean of intraseasonal anomalies seem to account for only about 50% of the anomalies of seasonal mean of that year. To investigate how the mean of intraseasonal anomalies fare in explaining the amplitude of anomalies of the seasonal mean for all simulated years, the averaged IS anomalies and anomalies of seasonal mean for precipitation and U850 averaged over the region bounded by 70◦ E-100◦ E and 10◦ N25◦ N for all years are shown in Fig.16. The correlation between the two curves is very high as could be expected from the pattern correlations (Fig.14), but the amplitude of mean of the IS anomalies is always about one half or less than the interannual anomalies of the seasonal mean. This indicates that the the ISO’s do control the nature of the IAV but the linear effect of projection of the IS mode onto the interannual mode is significantly amplified by some nonlinear interactions within the system.

7 Conclusions and Discussions

Seasonal mean prediction and predictability experiments with GCM’s indicate that a significant fraction of observed interannual variability (approximately 50% or more) of the Indian summer monsoon may be due to ’internal’ low frequency variability. Potential predictability studies with long observations also indicate similar contribution of ’internal’ LF variability to observed IAV of the Indian summer monsoon. The ’internal’ LF variability act as a background of unpredictable noise mixed with the predictable ’externally’ forced signal. Improvement in seasonal mean prediction would require successful extraction the ’signal’ from the background of noise of comparable magnitude. In order to develop such a technique, a clearer understanding of physical mechanism(s) responsible for the ’internal’ LF variability is required but is not currently available. The objective of the present study has been to attempt to unravel the mechanism responsible for the observed ’internal’ LF variability. Since clean separation between the ’internal’ and ’external’ component of interannual variability from observations is difficult, we take recourse to a GCM experiment to achieve the goal. The CCM3 model with T42 horizontal resolution and 18 levels in the vertical is integrated for 25 years without any external interannual forcing, with only annual cycle of solar forcing and annual cycle of SST as boundary forcing. Any interannual variability of the simulated

18

B. N. Goswami, Prince K. Xavier

Indian summer monsoon (or any other climate system for that matter) is, therefore, of internal origin. It is shown that model simulates interannual variability of the Indian monsoon comparable in amplitude and similar in spatial structure to the observed interannual variability. As the model simulates significant and realistic internal IAV of the summer monsoon, an understanding of the model internal IAV may provide insight towards understanding the observed internal IAV. Inspired by some of our earlier work, we start with the working hypothesis that monsoon ISO’s are responsible for generating the internal monsoon IAV. The first important element of this hypothesis are that the spatial structure of the ISO’s and IAV should be similar. Then, higher frequency of occurrence of active (weak) phase of the ISO in a given season could give rise to a strong (weak) monsoon. To test this hypothesis, first, the nature of the simulated ISO’s is studied in detail and compared with observed character of ISO’s. It is shown that the model simulates the spatial and temporal structures and the propagation characteristics of the summer ISO’s over the Indian monsoon region with good deal of fidelity. The spatial structure of the ISO’s is similar to that of the seasonal mean such that it strengthens (weakens) the seasonal mean in its active (weak) phase. It is further shown that dominant mode of ISV and that of IAV of the seasonal mean are governed by a common mode of spatial variability. The fact that the projection of the ISV mode on the IAV mode governs the IAV of the simulated monsoon is supported by the finding that strong (weak) simulated monsoon is associated with higher probability of occurrence of active (weak) ISO conditions. Strong support to this conclusion is provided when we show that the spatial pattern of seasonal mean of ISO anomalies is almost identical to that of anomaly of seasonal mean for every year of simulations with pattern correlations ranging between 0.96 and 0.99. Thus, the internal IAV of the Indian monsoon appears to be governed primarily by projection of the ISO’s on to the IAV mode. However, when the actual amplitude of IAV of the monsoon accounted by the projection of the ISO’s is examined, it is noted that the process accounts only about 50% of IAV of the monsoon even within the context of the GCM. It implies that while the ISO’s do command the nature of the IAV of the

Dynamics of ’Internal’ Interannual Variability of the Indian Summer Monsoon in a GCM

19

monsoon, the process of determining the amplitude of IAV is not entirely linear. Nonlinearities within the system (nonlinear interaction between synoptic disturbances and the frequencies within the ISO band) seem to amplify the ISO induced signal. The soil moisture in this run has been interactive. Therefore, soil moisture feedback could have also contributed to enhancement in the ISO induced IAV signal. Acknowledgments We are grateful to Department Ocean Development, Govt. of India, New Delhi and Indian National Centre for Ocean Information Services (INCOIS), Hyderabad for a grant and partial support for this work. UCAR/NCAR/CGD is greatfully acknowledged for making the NCAR-CCM3 available. Thanks are due to R. Vinay for help with the model.

References

Ajayamohan, R. S., and B. N. Goswami, 2003: Potential predictability of the Asian summer monsoon on monthly and seasonal time scales. Met. Atmos. Phys., pp. DOI 10.1007/s00,703–002–0576–4. Anderson, J. L. H., van den Dool, A. Barnston, W. Chen, W. Stern, and J. Ploshay, 1999: Present day capabilities of numerical and statistical models for atmospheric extratropical seasonal simulation and prediction. Bull. Amer. Meteor. Soc., 80, 1349–1361. Blandford, H. F., 1884: On the connection of the Himalaya snowfall with dry winds and seasons of drought in india. Proc. Roy. Soc. London, 37, 3–22. Charney, J. G., and J. Shukla, 1981: Predictability of monsoons. Monsoon Dynamics, J. Lighthill and R. P. Pearce, Eds., Cambridge University Press, pp. 99–108. Cherchi, A., and A. Navarra, 2003: Reproducibility and predictability of Asian summer monsoon in the ECHAM4-GCM. Climate Dyn., 20, 365–379. Duchon, C. E., 1979: Lanczos filtering in one and two dimensions. J. Appl. Meteor., 18, 1016–1022. Fennessy, M., and J. Shukla, 1994: Simulation and predictability of monsoons. Proceedings of the international conference on monsoon variability and prediction, WMO/TD 619, Trieste, pp. 567–575.

20

B. N. Goswami, Prince K. Xavier

Fennessy, M. J., and J. Shukla, 1999: Impact of initial soil wetness on seasonal atmospheric prediction. J. Climate, 12, 3167–3180. Ferranti, L., J. M. Slingo, T. N. Palmer, and B. J. Hoskins, 1997: Relations between interannual and intraseasonal monsoon variability as diagnosed from AMIP integrations. Quart. J. Roy. Meteor. Soc., 123, 1323–1357. Gadgil, S., and S. Sajani, 1998: Monsoon precipitation in the AMIP runs. Climate Dyn., 14, 659–689. Goswami, B. N., 1994: Dynamical predictability of seasonal monsoon rainfall: Problems and prospects. Proc. Ind. Nat. Acad. Sci., 60A, 101–120. Goswami, B. N., 1998: Interannual variations of Indian summer monsoon in a GCM: External conditions versus internal feedbacks. J. Climate, 11, 501–522. Goswami, B. N., and R. S. Ajayamohan, 2001: Intraseasonal oscillations and interannual variability of the Indian summer monsoon. J. Climate, 14, 1180–1198. Goswami, B. N., V. Krishnamurthy, and H. Annamalai, 1999: A broad-scale circulation index for interannual variability of the Indian summer monsoon. Quart. J. Roy. Meteor. Soc., 125, 611–633. Goswami, P., and Srividya, 1996: A novel neural network design for long-range prediction of rainfall pattern. Curr. Sci., 70, 447–457. Gowarikar, V., V. Thapliyal, R. P. Sarker, G. S. Mandel, and D. R. Sikka, 1989: Parametric and power regression models: New approach to long range forecasting of monsoon rain in India. Mausam, 40, 125–130. Hack, J. J., 1994: Parametrization of moist convection in the National Center for Atmospheric Research Community Climate Model (CCM2). J. Geophys. Res., 99, 5551–5568. Holtslag, A. A. M., and B. A. Boville, 1993: Local versus nonlocal boundary-layer diffusion in a global climate model. J. Climate, 6, 1825–1842. Kalnay, E., et al., 1996: The NCEP/NCAR 40-year reanalysis project. Bull. Amer. Meteor. Soc., 77, 437–471.

Dynamics of ’Internal’ Interannual Variability of the Indian Summer Monsoon in a GCM

21

Kang, I.-S., et al., 2002: Intercomparion of atomospheric GCM simulated anomalies associated with the 1997-98 El Nino. J. Climate, 15, 2791–2805. Kiehl, J. T., J. J. Hack, G. B. Bonan, B. A. Boville, D. L. Williamson, and P. J. Rasch, 1998: The National Center for Atmospheric Research Community Climate Model. J. Climate, 11, 1131–1150. Kistler, R., et al., 2001: The NCEP NCAR 50-year reanalysis: Monthly means cd-rom and documentation. Bull. Amer. Meteor. Soc., 82, 247–267. Kriplani, R. H., and A. Kulkarni, 1997: Climate impact of El Nino/La Nina on the Indian monsoon: A new perspective. Weather , 52, 39–46. Krishnamurthy, V., and B. N. Goswami, 2000: Indian monsoon-ENSO relationship on inter decadal time scales. J. Climate, 13, 579–595. Krishnamurthy, V., and J. Shukla, 2000: Intraseasonal and interannual variability of rainfall over India. J. Climate, 13, 4366–4377. Krishnamurti, T. N., and P. Ardunay, 1980: The 10 to 20 day westward propagating mode and ’breaks’ in the monsoons. Tellus, 32, 15–26. Krishnamurti, T. N., and H. N. Bhalme, 1976: Oscillations of monsoon system. part I: Observational aspects. J. Atmos. Sci., 45, 1937–1954. Krishnamurti, T. N., and D. Subrahmanyam, 1982: The 30-50 day mode at 850mb during MONEX. J. Atmos. Sci., 39, 2088–2095. Krishnamurti, T. N., C. M. Kishtawal, T. E. LaRow, D. R. Bachiochi, Z. Zhang, C. E. Williford, S. Gadgil, and S. Surendran, 1999: Improved weather and seasonal climate forecasts from multimodel superensemble. Science, 285, 1548–1550. Krishnamurti, T. N., C. Kishtawal, T. LaRow, D. Bachiochi, Z. Zhang, C. Williford, S. Gadgil, and S. Surendran, 2000: Multimodel superensemble forecasts for weather and seasonal climate. J. Climate, 13, 4196–4216.

22

B. N. Goswami, Prince K. Xavier

Krishnamurti, T. N., et al., 2001: Real-time multianalysis-multimodel superensemble forecasts of precipitation using TRMM and SSM/I products. Mon. Wea. Rev., 129, 2861–2883. Kumar, A., and M. P. Hoerling, 1995: Prospects and limitations of atmospheric GCM climate predictions. Bull. Amer. Meteor. Soc., 76, 335–345. Kumar, K. K., B. Rajagopalan, and M. A. Cane, 1997: On the weakening relationship between the Indian monsoon and ENSO. Science, 284, 2156–2159. Lau, N. C., 1985: Modelling the seasonal dependence of atmospheric response to observed El Nino in 1962-76. Mon. Wea. Rev., 113, 1970–1996. Meehl, G. A., and J. Arblaster, 1998: The Asian-Australian monsoon and El Nino-Southern Oscillation in the NCAR Climate System Model. J. Climate, 11, 1356–1385. Murakami, T., T. Nakazawa, and J. He, 1984: On the 40-50 day oscillation during 1979 northern hemisphere summer. part I: Phase propagation. J. Meteor. Soc. Japan, 62, 440–468. Palmer, T. N., 1994: Chaos and the predictability in forecasting monsoons. Proc. Ind. Nat. Acad. Sci., 60A, 57–66. Ramanathan, V., and P. Downey, 1986: A nonisothermal emissivity and absorptivity formulation for water vapor. J. Geophys. Res., 91, 8649–8666. Rao, Y. P., 1976: Southwest Monsoon, India Meteorological Department, New Delhi, meteorological Monograph. Reynolds, R. W., and T. M. Smith, 1994: Improved global sea surface temperature analyses using optimum interpolation. J. Climate, 7, 929–948. Sahai, A. K., A. M. Grimm, V. Satyan, and G. B. Pant, 2003: Long-lead prediction of Indian summer monsoon rainfall from global SST evolution. Climate Dyn., 20 (7/8), 855–863. Saji, N. H., and B. N. Goswami, 1997: An intercomparison of the seasonal cycle of tropical surface stress simulated by 17 AMIP GCMs. Climate Dyn., 13, 561–585.

Dynamics of ’Internal’ Interannual Variability of the Indian Summer Monsoon in a GCM

23

Shukla, J., 1998: Predictability in the midst of chaos: A scientific basis for climate forecasting. Science, 282, 728–731. Shukla, J., and J. M. Wallace, 1983: Numerical simulation of the atmospheric response to equatorial Pacific sea surface temperature anomalies. J. Atmos. Sci., 40, 1613–1630. Sikka, D. R., and S. Gadgil, 1980: On the maximum cloud zone and the ITCZ over Indian longitude during southwest monsoon. Mon. Wea. Rev., 108, 1840–1853. Sperber, K. R., and T. N. Palmer, 1996: Interannual tropical rainfall variability in General Circulation Model simulations associated with Atmospheric Model Intercomparison Project. J. Climate, 9, 2727– 2750. Sperber, K. R., J. M. Slingo, and H. Annamalai, 2000: Predictability and the relationship between subseasonal and interannual variability during the Asian summer monsoons. Quart. J. Roy. Meteor. Soc., 126, 2545–2574. Walker, G. T., 1923: Correlation in seasonal variations of weather. III: A preliminary study of world weather. Mem. Indian Meteorol. Dept, 24, 75–131. Walker, G. T., 1924: Correlation in seasonal variations of weather. IV : A further study of world weather. Mem. Indian Meteorol. Dept, 24, 275–332. Xie, P., and P. A. Arkin, 1996: Analyses of global monthly precipitation using guage observations, satellite estimates and numerical predictions. J. Climate, 9, 840–858. Yasunari, T., 1979: Cloudiness fluctuation associated with the northern hemisphere summer monsoon. J. Meteor. Soc. Japan, 57, 227–242. Yasunari, T., 1980: A quasi-stationary appearance of 30-40 day period in the cloudiness fluctuation during summer monsoon over India. J. Meteor. Soc. Japan, 58, 225–229. Zhang, G. J., and N. A. McFarlane, 1995: Sensitivity of climate simulations to the parameterization of cumulus convection in the Canadian Climate Centre General Circulation Model. Atmos.-Ocean, 33, 407–446.

24

B. N. Goswami, Prince K. Xavier

Table 1 Seasonal mean and its interannual standard deviation Mean

Standard Deviation

Indices

CCM3

Observation

CCM3

Observation

EIMR

8.96

8.20

0.61

0.76

MH

2.88

2.54

0.24

0.34

KELLJ

139.15

111.18

11.96

6.90

Dynamics of ’Internal’ Interannual Variability of the Indian Summer Monsoon in a GCM

25

Fig. 1 Comparison of climatological 1June-30September (JJAS) mean of monsoon precipitation and winds. Model precipitation (A) and CMAP (D). 850 hPa winds from CCM3 (B) and NCEP (E) plotted as vectors and magnitude is contoured with contour interval 2 m s−1 . Winds at 200 hPa from CCM3 (C) and NCEP (F) plotted as vectors and magnitude is contoured with contour interval 5 m s−1 .

26

B. N. Goswami, Prince K. Xavier

Fig. 2 Latitude-time plot of daily climatological mean precipitation (mm day−1 ) averaged over the longitudinal belt 70◦ -100◦ E from CCM3 (A) and from CMAP (B).

Fig. 3 Kinetic energy (m2 s−2 ) at 850 hPa from daily climatological winds averaged over the region 50◦ -65◦ E, 5◦ -15◦ N from CCM3 (dotted line) and from NCEP (solid line).

Dynamics of ’Internal’ Interannual Variability of the Indian Summer Monsoon in a GCM 3

27

EIMR KE−LLJ MHI

2 1 0 −1 −2 1

3

5

7

9

11

13

15

17

19

21

23

25

Fig. 4 Different indices for monsoon interannual variability. Construction of these indices is described in section 4.

Fig. 5 Composites of anomalies of seasonal mean precipitation (mm day−1 ) from CCM3 (A) and CMAP (C) and 850 hPa winds from CCM3 (B) and NCEP (D) for strong monsoon years. Strong monsoon years are identified based on normalized EIMR. Scale for wind anomalies is shown in the inset box.

28

Fig. 6 Same as Fig. 5, but for weak monsoon years.

B. N. Goswami, Prince K. Xavier

Dynamics of ’Internal’ Interannual Variability of the Indian Summer Monsoon in a GCM

29

8

A f × p(f)

6 4 2 0 4

0

20

40

60

80

100

B f × p(f)

3 2 1 0 8

0

20

40

60

80

100

C f × p(f)

6 4 2 0

0

20

40

60

Period (days)

80

100

Fig. 7 Power spectrum showing power×frequency (y-axis) versus period in days (x-axis). (A) for CCM3 JJAS precipitation anomalies averaged over 85◦ -95◦ E, 15◦ -20◦ N. (B) for CCM3 850 hPa zonal wind anomalies for the JJAS period averaged over the region 80◦ -90◦ E, 5◦ -15◦ N. (C) same as (B) but for zonal wind anomalies from NCEP.

30

B. N. Goswami, Prince K. Xavier

Fig. 8 Ratio of variance of 10-20 day (A) and 30-90 day modes (B) in precipitation to the total variance expressed as percentage at all grid points. Similar ratio for 10-20 day (C) and 30-90 day (D) modes in 850 hPa zonal wind.

Dynamics of ’Internal’ Interannual Variability of the Indian Summer Monsoon in a GCM

31

Fig. 9 Regression of 10-20 day band-passed 850 hPa zonal (u) and meridional (v) winds at all grid points with a reference time series defined as filtered u anomalies averaged over 85◦ -90◦ E, 5◦ -10◦ N with 7 days lead (A) and zero lag (B). Longitude versus lags (days) plot of regressed u averaged over 5◦ -15◦ N from CCM3 (C). D, E, and F are similar to A, B and C respectively, but from NCEP.

32

B. N. Goswami, Prince K. Xavier

Fig. 10 Regression of 30-90 day band-passed 850 hPa zonal (u) and meridional (v) winds at all grid points with a reference time series defined as filtered u anomalies averaged over 80◦ -90◦ E, 10◦ -15◦ N with 23 days lead (A) and zero lag (B). Latitude versus lags (days) plot of regressed u averaged over 70◦ -90◦ E from CCM3 (C). D, E, and F are similar to A, B and C respectively, but from NCEP.

Dynamics of ’Internal’ Interannual Variability of the Indian Summer Monsoon in a GCM

33

Fig. 11 EOF1 of interannual anomalies of seasonal mean of precipitation (A) and u at 850 hPa (C) respectively. B and D show EOF1 of 10-90 day bandpassed intraseasonal anomalies of precipitation and u at 850 hPa respectively.

Fig. 12 Variance of interannual anomalies of seasonal mean precipitation (A) and u at 850 hPa (C). Variance of the intraseasonal anomalies of precipitation (B) and u at 850 hPa (D).

34

B. N. Goswami, Prince K. Xavier

Fig. 13 Spatial pattern of anomaly of seasonal mean precipitation (A) and that of seasonal mean intraseasonal anomaly (B) for the year 7.

Fig. 14 Spatial correlation between anomaly of seasonal mean and seasonal mean intraseasonal anomaly of precipitation (open circle) and u at 850 hPa (filled circle) for each year.

Dynamics of ’Internal’ Interannual Variability of the Indian Summer Monsoon in a GCM

35

A

20

10

0

−2

−1

0

1

2

B

30 20 10 0

−2

−1

0

1

2

Fig. 15 Histogram of intraseasonal anomalies for 3 strong monsoon years (A) and 3 weak monsoon years. Strong and weak monsoon years are identified based on EIMR (see Fig 4.

36

B. N. Goswami, Prince K. Xavier

Fig. 16 Relationship between interannual anomalies of seasonal mean (open circle) and seasonal mean intraseasonal anomalies (filled circle) of u 850 hPa (A) and precipitation (B) for each year.