1. Introduction
Water management is a key issue affecting public health and sociopolitical stability in Saudi Arabia. Because the atmospheric moisture budget plays an important role in the hydrology of this region, understanding the seasonal moisture budget and its interannual variability is important to arid–semiarid zone management and weather-modification projects.
Aridity in a subtropical or midlatitude desert region such as Saudi Arabia (Fig. 1a) is often related to large-scale subsidence caused by Hadley circulations and a location far from an oceanic source of moisture. However, being close to the midlatitude storm track, wintertime synoptic disturbances from the Mediterranean Sea bring moisture to the region (Mujumdar et al. 2005, manuscript submitted to Geophys. Res. Lett., hereafter MBC). Local intensification of such disturbances is a major cause of rainfall there in winter. The monsoon system in the neighboring Indian Ocean, on the other hand, has a strong impact on the summer climate of this region when most of it is governed by a surface thermal low that extends from the Indian monsoon low. During this season, a lower-tropospheric intense jet [called the “Findlater jet” after Findlater (1969)] carrying moisture above the western Indian Ocean veers off the coast of the Arabian Peninsula. This change in the direction of the low-level jet along East Africa and Somalia reduces moisture supply in the interior of Saudi Arabia and toward its mountainous region of Asir Province.
Interannual variability of seasonal atmospheric moisture over Saudi Arabia is caused by variations of regional and global circulation patterns due to tropical and extratropical climatic events such as El Niño–Southern Oscillation (ENSO) and the Indian Ocean dipole (IOD; Saji et al. 1999; Webster et al. 1999). The impact of ENSO on the Indian summer monsoon is discussed by many authors (Walker 1923; Shukla and Paolino 1983; Joseph et al. 1994). Webster et al. (1998) explained how ENSO influences the Indian Ocean region through changes in the Walker circulation. Webster found that warming of the sea surface temperature (SST) in the central and eastern Pacific during an El Niño event diminished easterly trade winds and shifted the tropical convection eastward. Since the monsoon mechanism involves an interaction between the Walker and Hadley circulations, any change in the large-scale Walker circulation over the Indian Ocean can be expected to influence the moisture supply to the monsoon system as well as to the coast of Saudi Arabia adjacent to the Findlater jet. Like ENSO events, IOD events have a strong influence on the climate in the immediately neighboring East African and Indonesian regions (Saji et al. 1999) as well as in east Asia (Saji and Yamagata 2003; Guan et al. 2003), the Mediterranean, Australia, and Brazil (Saji and Yamagata 2003).
The principal objectives of our research were (i) to understand the seasonal characteristics of moisture flux, precipitable water content, and precipitation over Saudi Arabia using National Centers for Environmental Prediction–National Center for Atmospheric Research (NCEP–NCAR) reanalysis and Xie–Arkin datasets, and (ii) to understand the role of IOD and ENSO with respect to the interannual moisture budget by analyzing reanalysis data and conducting SST sensitivity experiments with a global general circulation model (GCM). The paper is organized as follows: A brief description of the model datasets and methodology is presented in section 2. In section 3, results and discussion of our analysis of the NCEP–NCAR reanalysis and our model results are presented. Conclusions appear in section 4.
2. Data, methodology, and model
a. Data and methodology
The NCEP–NCAR reanalysis monthly mean datasets (Kalnay et al. 1996) for the 43-yr period from 1958 through 2000 were employed to study the seasonal moisture-flux climatology, its origin, and its precipitable water content. These data have been widely used over the last few years in tropical climate research (Krishna Kumar et al. 1999; Wang et al. 2001; Xie et al. 2002). Though the NCEP–NCAR precipitation data are underestimated, other atmospheric quantities are reasonable (Annamalai et al. 1999). Since our research explored seasonal aspects of moisture transport, we used all levels of winds (zonal and meridional components) and specific humidity from the above reanalysis in our work. To study the interannual variations of moisture flux into Saudi Arabia and their relation to IOD and ENSO, we computed the dipole mode index (DMI; Saji et al. 1999) and the Niño-3 index using the Global Sea Ice and Sea Surface Temperature dataset (GISST; Rayner et al. 1996) from the same period as the NCEP–NCAR datasets. The DMI is obtained by computing the SST anomaly difference between the tropical western Indian Ocean (10°S–10°N, 50°–70°E) and the tropical southeastern Indian Ocean (10°S–equator, 90°–110°E). The Niño-3 index is obtained by taking the area average of SST anomalies over 5°N–5°S, 150°–90°W. Seasonal rainfall climatologies of Saudi Arabia were investigated by using rainfall data from Willmott et al. (1996), Xie and Arkin (1997), and station observations, which are 24-h total rain gauge observations obtained from the Meteorological and Environmental Protection Agency (MEPA) of Saudi Arabia. Domains and regions considered in our study are shown in Fig. 1b.
The vertical distribution of water vapor over Saudi Arabia is different from other regions of the globe in that it is a unique region of descending motion (Rodwell and Hoskins 1996). Keeping this in mind, we separated the whole troposphere into lower (1000–850 hPa), middle (700–500 hPa), and upper (400–100 hPa) layers in our study. Sen (1983), Nouh (1987), and Bazuhair et al. (1997) identified the temporal distribution of wet and dry periods in Saudi Arabia using monthly rainfall data from various rain gauge stations. They identified two distinct rainfall periods over this region, one in winter (November through April) and the other in summer (June through August). The spatial patterns of monthly rainfall climatology (Fig. 2a)—a detailed explanation is given in the results section of this paper—are also consistent with the above rainfall studies. We considered three seasons based on the above studies: November–January (NDJ), February–April (FMA), and June–August (JJA). Because summer rainfall was observed only in the mountainous region of Saudi Arabia compared to the winter period (Fig. 2b), we considered two domains (ABCD and EFGH) in our moisture-budget analysis.
The mathematical formulations used in our study are given in the appendix. The precipitable water content in an air column within any two pressure layers was calculated using Eq. (A1). To identify the moisture-source (divergence) and moisture-sink (convergence) regions of water vapor, we computed both the divergent and rotational components of moisture transport. To find the components of moisture vectors, first we calculated the total water vapor transport (Q) within any two pressure layers using Eqs. (A2) and (A3), then we used Eq. (A5) for streamfunction (ψ) and velocity potential (χ). Finally, we used Eq. (A6) because the streamfunction and velocity potential are already known.
b. Brief description of the model and experimental design
To understand IOD and ENSO influences on the Saudi Arabian climate, we conducted four sensitivity experiments using the University of Tokyo Community Model (UTCM; Chakraborty et al. 2003); its atmospheric component is an upgraded version of the Frontier Atmospheric General Circulation Model (AGCM) version 1.0 (FrAM-1.0; Guan et al. 2000; Ashok et al. 2001). Each experiment has five-member ensembles that differ from each other in initial conditions (conditions were 1–5 June daily climatology derived from the NCEP daily reanalysis). The same set of initial conditions was used for all experiments where the model was integrated for 1 yr from the starting initial-condition date for each experiment year. In the control experiment (CTL), we imposed the seasonally varying climatological SST (Fig. 3a) as the lower boundary condition. In the second and third sensitivity experiments (IOD_EXP1 and IOD_EXP2), we merged the tropical Indian Ocean SST anomalies of the 1961 and 1997 IOD years (Figs. 3b,c) with that of the CTL experiment as the lower boundary condition. In the fourth sensitivity experiment (ENSO_EXP), we merged the tropical Pacific Ocean SST anomalies of the 1987 warm ENSO event (Fig. 3d) with that of the CTL experiment as the lower boundary condition. The results for each experiment are the average of the simulations from its five-member ensembles. These ensemble averages were used to remove the portion of the intraseasonal variability internally generated by the model. The results from the CTL experiment were subtracted from the IOD_EXP1, the IOD_EXP2, or the ENSO_EXP experiment to obtain the various anomalies in response to the positive IOD SST anomalies or the warm ENSO SST anomalies, respectively.
3. Results and discussions
a. Diagnosis of Xie–Arkin rainfall and NCEP–NCAR reanalysis datasets
1) Rainfall climatology
To get a complete picture of the spatial and temporal evolution of monthly and seasonal rainfall climatology over Saudi Arabia, we present a map from the rainfall dataset (Fig. 2) of Xie and Arkin (1997), which is called the Climate Prediction Center (CPC) Merged Analysis of Precipitation (CMAP). These fields rely primarily on information from rain gauge observations and satellite estimates made with several different algorithms based on outgoing longwave radiation (OLR), and on scattering and emission of microwave radiation. This is the best available product since NCEP–NCAR precipitation is underestimated (Annamalai et al. 1999). The monthly rainfall climatology shows widespread rainfall (>5 mm) over Saudi Arabia during winter and spring, extending from November through April (Fig. 2a). A rainfall peak was observed during these months over the mountainous region of Asir Province (20°–2°N, 39°–41°E). Although the interior of Saudi Arabia is nearly dry from June through October, rainfall has been observed near the coastal Red Sea region toward Asir Province. The same characteristic has been observed in seasonal climatology with widespread rainfall during winter and spring (NDJ and FMA; Fig. 2b) but confined rainfall during summer (JJA; Fig. 2b). The seasonal maxima during winter and spring over the mountainous region of Asir Province (20°–22°N, 39°– 41°E) can been seen in Fig. 2b. Although Xie–Arkin precipitation for May is weak and similar to April, we did not consider May in our calculations because Bazuhair et al. (1997) do not identify it as a wet period in their study based on rain gauge stations.
2) Moisture flux and precipitable water content
The characteristics of moisture-flux components (divergent and rotational) along with precipitable water content for different tropospheric layers and for different seasons are presented in this section. The divergent component reflects the moisture-source/-sink regions; the rotational component describes the atmospheric water vapor transport. The calculations of these components are described in our data and methodology section. In the lower troposphere, eastern Saudi Arabia received more precipitable water than western and central areas during all three seasons (Fig. 4) of our study. The two moisture-source regions as observed from the divergent component, which are located over the Persian Gulf (including the adjacent Arabian Sea and the North Atlantic), supply moisture to Saudi Arabia during NDJ and FMA. The presence of a strong anticyclonic circulation associated with the Azores high over the North Atlantic, however, restricts moisture transport to the Saudi Arabian region in both winter and summer. Saudi Arabia receives precipitable water during NDJ and FMA mainly from the Red Sea and the Arabian Sea, and to some extent from the Mediterranean Sea. In summer (JJA) the main source regions of moisture fluxes are over the Indian Ocean, the northern part of the Red Sea, and the Mediterranean Sea. The rotational component of the moisture transport due to a weak cyclonic circulation over Saudi Arabia brings moisture from the Persian Gulf to the Saudi Arabian region in summer.
In the middle troposphere, rotational and divergent moisture fluxes reveal very important characteristics of rainfall over Saudi Arabia (Fig. 5). The divergent moisture flux during NDJ and FMA was observed over central and east Africa around 10°N with a strong core near the north Ethiopia–Somalia region; this is the main source of moisture flux. An elongated convergence zone was seen from the Black Sea to the Persian Gulf region. A strong moisture source with a large amount of precipitable water (>25 mm) was also observed in this area. The moisture feeding into the Saudi Arabian region comes from the Atlantic via the Mediterranean Sea, as well as from the Arabian Sea and the Red Sea due to the anticyclonic circulation over Yemen, Oman, and adjacent seas. We observed that the lower-tropospheric convergence (Fig. 4) and the middle-tropospheric divergence (Fig. 5) were located over eastern Africa and southern Saudi Arabia. Moisture convergence in the lower troposphere and divergence in the middle troposphere during winter are indicative of the upward motion of moist air, which in turn may release latent heat due to condensation. This available heat energy may be the source of low-level baroclinic instability. The divergent moisture flux was observed during JJA over the same region as that of NDJ and FMA, except that the flanks of moisture diverge over southeast Africa and the South Atlantic Ocean. The large amount of precipitable water over Saudi Arabia in the middle troposphere is transported by rotational winds from the North Atlantic and subtropical regions of Iran and Afghanistan. However, the strong divergence of available moisture around 20°N from the Sahara to western India and extending into the Middle East may explain the scarce rainfall in the interior of Saudi Arabia during summer (Fig. 2).
The upper-tropospheric distribution of the precipitable water content is low (<8 mm) during all three seasons, compared to that in the middle and lower troposphere (Fig. 6). The convergence over the Atlantic–African sector and the divergence over the India–Indian subcontinental sector are weak. Thus upper-level moisture transport may not have much influence on precipitation over Saudi Arabia during NDJ and FMA.
3) Seasonal variability in local and regional moisture flux
The distribution of net incoming moisture flux (Fig. 7a) vertically integrated from 1000 to 300 hPa showed a semiannual signal with a larger peak in winter (November–April) and another much smaller maximum in summer (June–August). This was observed in both ABCD (all of Saudi Arabia) and EFGH (only Abha and its surrounding region) domains (Fig. 1b). Moisture-flux peaks in summer for both domains are identical and collocated in time. Because most incoming moisture flux concentrates in the Abha region, the summer rain peak (Fig. 7b) is also expected to be near Abha. In contrast, the winter peak of the ABCD domain is much larger than that of the EFGH, indicating widespread rainfall over Saudi Arabia (Fig. 2). Wintertime moisture flux is associated with the enhancement of eastward-propagating Mediterranean disturbances near the Abha region by local effects of orography and moist-air advection from the Red Sea (MBC). For other months of the year, however, the net moisture flux over these domains is near zero or negative. In agreement with the net moisture flux, rainfall estimates over the mountainous region of the Asir Province (Abha) show similar semiannual signals (Fig. 7b).
To describe the seasonality of moisture transports to the Saudi Arabian regions precisely, we present the climatology of net flux of water vapor across the walls for three different seasons and three different tropospheric levels. These quantities were computed using Eqs. (A7) and (A8), as shown in the appendix. The moisture-flux vectors of different tropospheric levels through AB, BC, CD, and DA of the ABCD domain and through EF, FG, GH, and HE of the EFGH domain are shown in Fig. 8. The flux vectors are consistent with the large-scale circulation patterns (Figs. 4 –6), although a small difference in the mountainous Abha region (EFGH) was observed. We argue that this difference may be due to local small-scale circulation. There is an incoming flux from EF (the Red Sea side) in the lower troposphere in all three study seasons; however, outflux is observed in the middle and upper troposphere across EF during JJA. Moisture transport across FG and GH of the smaller domain is in phase with that across BC and CD of the larger domain (Fig. 8) during winter and spring. On the Red Sea side, the moisture transport is in phase across EF and AB during JJA (Fig. 8). The net moisture flux within ABCD shows seasonality similar to that within EFGH; it shows a stronger maximum during winter and spring and a weaker maximum during JJA. This is also consistent with Xie–Arkin rainfall (Fig. 2).
The net fluxes available within the ABCD (EFGH) region during NDJ and FMA for the whole troposphere are 1.28 × 106 (0.776 × 106) and 2.65 × 106 (1.756 × 106) kg s−1, respectively. The net flux reduces to 0.20 × 106 (0.2083 × 106) kg s−1 during JJA. The availability of higher moisture during winter and spring compared to summer provides a favorable condition for higher rainfall over Saudi Arabia (Fig. 2); this is also associated with the weather disturbances traveling from the Mediterranean region. MBC showed that the disturbances that propagate southwestward from the Mediterranean region enhance moisture transport from the Red Sea to the mountainous region of Asir Province due to an elongated low over southwest Saudi Arabia.
4) Interannual variability of moisture flux
We examined the interannual variability of the moisture flux. In particular, we focused on possible relations with tropical climate phenomena like IOD and ENSO. The interannual variability of the incoming moisture-flux indices across the walls of the ABCD and EFGH domains are shown in Fig. 9. The indices were calculated from the seasonal-climatology-removed anomalies of the time series data. The long-term time mean of the above anomalies was subtracted from each time series datum and normalized by dividing the standard deviation. For smoothing we used a 5-month running mean. These normalized moisture-flux indices show significant variability during the 1960–67 and the 1980–000 periods and weak variability during the 1968–79 period. During the positive IOD (1961, 1994, and 1997) and the warm ENSO (1963/64, 1982/83, 1986/87, and 1997/98) events, the incoming moisture flux is even larger than in normal years except for 1968–79. The weak variability during 1968–79 may be associated with the period of weak variation of IOD (Behera and Yamagata 2003). During positive IOD events the incoming moisture flux across the Red Sea side (AB or EF) increases (Fig. 9). This influx helps the net gain of moisture but is slightly offset by the weak outflux across the Persian Gulf side (CD/GH). The most interesting point is that during a positive IOD or a warm ENSO year, the net moisture flux through the Red Sea side is more significant compared to the other three sides. Because the peak of moisture flux through the Red Sea wall does not occur concurrently with the IOD or ENSO peaks, we have carried out a partial-lag correlation here. Again during some of the year both IOD and ENSO occur concurrently; thus we adopted a partial-lag correlation instead of a simple correlation. This partial-lag correlation (Fig. 10) with the moisture flux from the Red Sea side shows that a positive IOD (El Niño) event has a peak correlation coefficient of about 0.5 (0.6) with a 5- (2-) month lead. This indicates that IOD (ENSO) leads moisture flux from the Red Sea side by 5 (2) months or that the moisture flux through the Red Sea side will attain its maximum value after about 5 (2) months of the occurrence of IOD (ENSO) events.
During an El Niño event the characteristic of the moisture flux across the Red Sea and Arabian Sea sides is somewhat different as compared to a positive IOD event (Fig. 11). In the case of a pure IOD event (e.g., 1961 and 1994; Yamagata et al. 2004), the transport comes from the Sudan region and reaches the Saudi Arabian region through the Red Sea side. This type of transport is known locally as “the Sudanese wave” (Abdullah and Al-Mazroui 1998). The eastward tropical flow during a pure positive IOD event caused moisture convergence over the southwestern Arabian region (Fig. 11a). The local orographic ascent in Asir Province led to an increase in rainfall. The large-scale subsidence over the western Indian Ocean during the pure El Niño event (e.g., 1987) also caused the local convergence of moisture over the southwestern Arabian region (Fig. 11b). When positive IOD and warm ENSO events occur concurrently (Fig. 11c), the convergence of moisture flux is enhanced, perhaps owing to the combined contribution of large-scale circulation from positive IOD and warm ENSO. Because anomalous cyclonic circulation over northwestern Saudi Arabia blocks outflow toward the Persian Gulf, local moisture convergence over the southwestern Arabian region is enhanced; as a result, rainfall activity increases.
b. Model results
To understand the influences of IOD and ENSO events, we conducted sensitivity experiments using UTCM. The description of the model as well as the initial and boundary conditions for these sensitivity experiments is explained in section 2. The model has shown sensitivity in the moisture transport in relation to changes in the lower boundary condition associated with the prescribed SST anomalies. The rotational and divergent components of moisture-flux anomalies in the lower troposphere (1000–850 hPa) of IOD_EXP1, IOD_EXP2, and ENSO_EXP for September–October and NDJ are shown in Figs. 12 and 13, respectively. We selected September–October and NDJ results to understand the evolution mechanism of moisture transport and its impact on Saudi Arabia because the peak time of IOD is September–November; IOD leads moisture flux by 5 months (Fig. 10). During September–October for IOD_EXP1 (Fig. 12), the moisture-source regions were over the southeastern Indian Ocean, the Indonesian region, eastern China, and central Africa. The sink regions were over the southwestern and equatorial Indian Ocean. The southern Indian Ocean, the eastern India–Bangladesh region with its core in the Bay of Bengal, the north Mediterranean region, and central Africa became the source region; eastern Africa and Saudi Arabia became the sink during NDJ (Fig. 13); and most of the moisture was transported from the Indian and Mediterranean Oceans to Saudi Arabia.
In the IOD_EXP2 simulation, eastern Asia, Africa, the Arabian Sea around 10°N, and the south Indian Ocean around 10°S became the source regions of moisture at the lower-tropospheric level, whereas the equatorial Indian Ocean and southern Indian Ocean below 20°S became the sink regions (Fig. 12). The IOD_EXP2 results are different from the IOD_EXP1 results because 1997 was a year when both IOD and ENSO occurred concurrently. There was no moisture convergence over Saudi Arabia during September–October for either IOD_EXP1 or IOD_EXP2. However, both IOD_EXP1 and IOD_EXP2 showed moisture convergence in the lower troposphere over Saudi Arabia during NDJ (Fig. 13). The IOD_EXP2 simulation also showed moisture convergence over eastern Africa and the Middle East although it was confined over Saudi Arabia and East Africa in IOD_EXP1. The strong subsidence over the southeastern Indian Ocean in IOD_EXP2 intensified the local lower-tropospheric moisture divergence. This divergent center intensified the convergent centers over Saudi Arabia through the lower-tropospheric moisture transport and was weak in the IOD_EXP1 during NDJ. Saudi Arabia was supplied more moisture from the Indian Ocean, Arabian Sea, central Africa, and the Mediterranean Ocean via the Red Sea in both IOD_EXP1 and IOD_EXP2.
In ENSO_EXP, the moisture source and the sink regions were located mainly over the southern Indian Ocean around 10°S during September–October (Fig. 12). A convergence zone around the Gulf of Aden was noted in ENSO_EXP. Also, the moisture flow over the Indian subcontinent changed its direction as seen in the rotational component compared to the IOD_EXP1 (Fig. 12). The Saudi Arabian region continued to obtain moisture from the Arabian Sea and the Mediterranean Sea and became the moisture-sink region in NDJ. The moisture convergence was weak but widespread compared to IOD_EXP2 (Fig. 13). This region received moisture from the Indian Ocean and the Atlantic Ocean via the Red Sea during NDJ. In the ENSO_EXP simulation, the imposed SST anomalies in the tropical Pacific produced warming in the eastern and central Pacific accompanied by diminished easterly trade winds and an eastward shift of tropical convection. The eastward shift of the convection zone from the African continent to the East Africa–Indian Ocean area enhanced moisture flux from the south Indian Ocean toward Saudi Arabia via the Red Sea (Figs. 13 and 14). We calculated the net moisture flux by integrating vertically from the surface to 300 hPa for our four experiments (CTL, IOD_EXP1, IOD_EXP2, and ENSO_EXP). The moisture-flux anomalies were calculated by subtracting the vertically integrated moisture flux of the CTL simulation from that of the other sensitivity simulations. These vertically integrated moisture-flux anomalies increased toward Saudi Arabia across the Red Sea during a pure positive IOD year (Fig. 14). The transport reached its peak during NDJ. For a year with concurrent IOD and ENSO such as 1997, the simulation (IOD_EXP2) showed that the vertically integrated moisture-flux anomalies for Saudi Arabia were similar in character to that of a pure IOD experiment (IOD_EXP1). The results from ENSO_EXP and also from IOD_EXP2 show a stronger flux of vertically integrated moisture anomalies toward the Saudi Arabian region through the Red Sea (Fig. 14).
4. Conclusions
We carried out a diagnostic study of the climatology with regard to seasonal and interannual variability of moisture in the Saudi Arabian region by analyzing a 43-yr NCEP–NCAR reanalysis and other datasets. In the near-coastal lowlands and mountainous regions of the eastern and western provinces of Saudi Arabia, rain is almost exclusively confined to the winter and spring seasons from November through April. The analysis of lower-tropospheric moisture fluxes shows that adjacent seas are the main sources of moisture supply for this region during such periods (NDJ and FMA). In the middle troposphere, the primary moisture source is found over central Africa, which in turn is traced to the Atlantic Ocean, the Indian Ocean, and the Arabian Sea. The upper-level moisture fluxes are weak and play a minor role for convection over this region. The precipitable water content is higher in the middle troposphere (>25 mm) and the lower troposphere (>15 mm) when compared to the upper troposphere (<3 mm). In the summer (JJA), the main sources of lower-tropospheric moisture are located over the Red Sea, the Arabian Sea, and the Mediterranean Sea.
The seasonal variation of net moisture shows a prominent semiannual signal with its highest peak during November–April and another smaller peak during June–August. The seasonality of moisture in summer does not vary between the smaller region of Asir Province (EFGH) and the larger region containing all of Saudi Arabia (ABCD), thus explaining why rainfall during summer is confined only to Asir Province. In contrast, the moisture gain over a wider area (ABCD) compared to the smaller domain (EFGH) during winter and spring explains widespread rain during these seasons.
The interannual variation of moisture flux shows that it is strongly modified by tropical climate phenomenon like positive IOD and warm ENSO events. During these climate events, the transport of moisture from the Red Sea toward the mountainous region of Asir Province is enhanced, and, as a result, rainfall during November through April increases. The impact on the moisture transport to the Saudi Arabian region from the Red Sea side (AB and EF) for both positive IOD and the warm ENSO events is stronger than other sides. The positive IOD and warm ENSO show the highest correlations with moisture flux from the Red Sea side. Numerical experiments using AGCM confirm the findings from NCEP–NCAR, Xie–Arkin, and other datasets. In all GCM sensitivity experiments, the vertically integrated moisture transport anomalies are remarkable during NDJ. It is also evident that during pure positive IOD year both the South Atlantic Ocean and the North Atlantic Ocean contribute moisture transport to Saudi Arabia, while, during an ENSO year, the South Atlantic is the principal contributor.
The interannual variability of winter rainfall is related primarily to the teleconnections originating from tropical climate phenomena like IOD or El Niño. The significance of the impact of a concurrent positive IOD and a warm ENSO event on the interannual variability of the incoming moisture flux over Saudi Arabia is one of the main findings of our study. This is important for the understanding of climate impact in desert amelioration. An investigation of the mechanism of climatic events in Saudi Arabia together with the actual origin of water vapor (Numaguti 1993) is the goal of our current research project.
Acknowledgments
This study was supported by the Revolutionary Kyosei Research Project (PR2002) of the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan through Mitsubishi Heavy Industries Ltd. The authors are grateful to Dr. David Schultz, the anonymous reviewers, and to editorial assistant Mary Golden whose suggestions helped to improve the manuscript. We are thankful to Mr. M. Harada, Mr. K. Muta, Dr. K. Baba, Ms. N. Hoeller, and Ms. J. Moriyama for their support in carrying out this research. The discussions and suggestions provided by Drs. Tomoki Tozuka, R. Suzuki, H. Nakamura, J. Masumoto, H. Sakuma, and scientists from FRCGC are greatly appreciated.
REFERENCES
Abdullah, M. A., and M. A. Al-Mazroui, 1998: Climatological study of the southwestern region of Saudi Arabia. I. Rainfall analysis. Climate Res, 9 , 213–223.
Annamalai, H., J. Slingo, K. R. Sperber, and K. Hodges, 1999: The mean evolution and variability of the Asian summer monsoon: Comparison of ECMWF and NCEP/NCAR analysis. Mon. Wea. Rev, 127 , 1157–1186.
Ashok, K., Z. Guan, and T. Yamagata, 2001: Impact of the Indian Ocean dipole on the relationship between the Indian monsoon rainfall and ENSO. Geophys. Res. Lett, 26 , 4499–4502.
Bazuhair, A. S., A. Algohani, and Z. Sen, 1997: Determination of monthly wet and dry periods in Saudi-Arabia. Int. J. Climatol, 17 , 303–311.
Behera, S. K., and T. Yamagata, 2003: Influence of the Indian Ocean dipole on the Southern Oscillation. J. Meteor. Soc. Japan, 81 , 169–177.
Chakraborty, A., T. Tozuka, T. Miyasaka, M. Mujumdar, S. K. Behera, and T. Yamagata, 2003: Development of the University of Tokyo Community Model (UTCM) for climate system: Preliminary results. Tech. Rep., Mitsubishi Heavy Industries Ltd., 49 pp.
Chen, T. C., 1985: Global water vapor flux and maintenance during FGGE. Mon. Wea. Rev, 113 , 1801–1819.
Findlater, J., 1969: A major low-level air current near the Indian Ocean during the northern summer. Quart. J. Roy. Meteor. Soc, 95 , 362–380.
Guan, Z., S. Iizuka, M. Chiba, S. Yamane, K. Ashok, and T. Yamagata, 2000: Frontier Atmospheric General Circulation Model Version 1.0 (FrAM–1.0): Model climatology. Tech. Rep. FTR-1, Frontier Research for Global Change/JAMSTEC, 27 pp.
Guan, Z., K. Ashok, and T. Yamagata, 2003: Summertime response of the tropical atmosphere to the Indian Ocean dipole sea surface temperature anomalies. J. Meteor. Soc. Japan, 81 , 533–561.
Joseph, P. V., J. K. Eischeid, and R. J. Pyle, 1994: Interannual variability of the onset of the Indian summer monsoon and its association with atmospheric features, El Niño, and sea surface temperature anomalies. J. Climate, 7 , 81–105.
Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Yr Reanalysis Project. Bull. Amer. Meteor. Soc, 77 , 437–471.
Krishna Kumar, K., B. Rajagopalan, and M. A. Cane, 1999: On the weakening relationship between the Indian monsoon and ENSO. Science, 284 , 2156–2159.
Krishnamurti, T. N., H. S. Bedi, and V. M. Hardiker, 1998: An Introduction to Global Spectral Modeling. Oxford University Press, 253 pp.
Nouh, M. A., 1987: Analysis of rainfall in the south-west region of Saudi Arabia. Proc. Inst. Civ. Eng, 2 , 339–349.
Numaguti, A., 1993: Dynamics and energy balance of the Hadley circulation and the tropical precipitation zones: Significance of the distribution of evaporation. J. Atmos. Sci, 50 , 1874–1887.
Peixoto, J. P., and A. H. Oort, 1983: The atmospheric branch of the hydrological cycle and climate. Variations of the Global Water Budget, Reidel, 5–65.
Rayner, N. A., E. B. Horton, D. E. Parker, C. K. Folland, and R. B. Hackett, 1996: Version 2.2 of the global sea-ice and sea surface temperature data set, 1903–1994. Climate Research Tech. Note 74, Hadley Centre for Climate Prediction and Research, Bracknell, United Kingdom, 21 pp.
Rodwell, M. R., and B. J. Hoskins, 1996: Monsoons and the dynamics of deserts. Quart. J. Roy. Meteor. Soc, 122 , 1385–1404.
Rosen, R. D., D. A. Salstein, and J. P. Peixoto, 1979: Variability in the annual fields of large-scale atmospheric water vapor transport. Mon. Wea. Rev, 107 , 26–37.
Saji, N. H., and T. Yamagata, 2003: Possible impacts of Indian Ocean dipole mode events on global climate. Climate Res, 25 , 151–169.
Saji, N. H., B. N. Goswami, P. N. Vinayachandran, and T. Yamagata, 1999: A dipole mode in the tropical Indian Ocean. Nature, 401 , 360–363.
Sen, Z., 1983: Hydrology of Saudi Arabia. Proc. Symp. on Water Resources in Saudi Arabia: Management, Treatment, and Utilization, Riyadh, Saudi Arabia, Ministry of Agriculture and Water, 68–94.
Shukla, J., and D. Paolino, 1983: The Southern Oscillation and long-range forecasting of summer monsoon rainfall over India. Mon. Wea. Rev, 111 , 1830–1837.
Walker, G. T., 1923: Correlation in seasonal variations of weather. VIII: A preliminary study of world weather. Mem. Indian Meteor. Dept, 24 , 75–131.
Wang, B., R. Wu, and K-M. Lau, 2001: Interannual variability of the Asian summer monsoon: Contrasts between the Indian and western North Pacific–east Asian monsoons. J. Climate, 14 , 4073–4089.
Webster, P. J., V. O. Magana, 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 , 14451–14510.
Webster, P. J., A. Moore, J. Loschnigg, and M. Leban, 1999: Coupled ocean–atmosphere dynamics in the Indian Ocean during 1997–98. Nature, 40 , 356–360.
Willmott, C. J., S. M. Robeson, and M. J. Janis, 1996: Comparison of approaches for estimating time-averaged precipitation using data from the USA. Int. J. Climatol, 16 , 1103–1115.
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. Amer. Meteor. Soc, 78 , 2539–2558.
Xie, S-P., H. Annamalai, F. A. Schott, and J. P. McCreary, 2002: Structure and mechanism of south Indian Ocean climate variability. J. Climate, 15 , 864–878.
Yamagata, T., S. K. Behera, J-J. Luo, S. Masson, M. Jury, and S. A. Rao, 2004: Coupled ocean atmosphere variability in the tropical Indian Ocean. Ocean–Atmosphere Interaction and Climate Variability, Geophys. Monogr., Amer. Geophys. Union, Vol. 147, 189–211.
APPENDIX
Moisture Flux Calculation
Moisture-flux quantification
(a) Diagram showing the adjacent seas of Saudi Arabia along with topography; (b) domains for moisture-budget analysis
Citation: Monthly Weather Review 134, 2; 10.1175/MWR3085.1
(a) Monthly and (b) seasonal average climatological precipitation (mm month−1) from Xie–Arkin over Saudi Arabia
Citation: Monthly Weather Review 134, 2; 10.1175/MWR3085.1
Equatorial averaged (5°S–5°N) SST used for CTL experiment and SST anomalies used for IOD_EXP1, IOD_EXP2, and ENSO_EXP sensitivity experiments
Citation: Monthly Weather Review 134, 2; 10.1175/MWR3085.1
(right) Divergent component and (left) rotational component of integrated moisture-flux vector (kg m−1 s−1) along with precipitable water in the lower troposphere for NDJ, FMA, and JJA. The colored regions in the right column are the divergent (source) and convergent (sink) component of moisture flux (unit: 10−4 kg m−2 s−1)
Citation: Monthly Weather Review 134, 2; 10.1175/MWR3085.1
Same as Fig. 4 but for middle troposphere
Citation: Monthly Weather Review 134, 2; 10.1175/MWR3085.1
Same as Fig. 4 but for upper troposphere
Citation: Monthly Weather Review 134, 2; 10.1175/MWR3085.1
(a) Seasonal variation of net moisture flux (106 kg s−1) and (b) average monthly rainfall over southwestern Saudi Arabia (Xie–Arkin and Willmott) and the Abha region (rain gauge stations from MEPA)
Citation: Monthly Weather Review 134, 2; 10.1175/MWR3085.1
Seasonality of moisture transport vector across the walls for different tropospheric layers for the (left) ABCD domain and (right) EFGH domain. Quantities within the circle represent net tropospheric moisture transport to the region
Citation: Monthly Weather Review 134, 2; 10.1175/MWR3085.1
Interannual variations of (a) DMI and ENSO, and moisture-flux indices through (b) AB and BC, (c) CD and DA, (d) EF and FG, and (e) GH and HE (using 5-month running mean)
Citation: Monthly Weather Review 134, 2; 10.1175/MWR3085.1
Partial-lag correlation of DMI and ENSO with moisture flux (positive lag means DMI or ENSO leads moisture flux) across the Red Sea sidewall
Citation: Monthly Weather Review 134, 2; 10.1175/MWR3085.1
Vertically integrated moisture-flux-vector (kg m−1 s−1) anomalies (NCEP–NCAR) and shaded rainfall anomalies (mm day−1) from Xie–Arkin for three particular years for NDJ
Citation: Monthly Weather Review 134, 2; 10.1175/MWR3085.1
Model-simulated lower-tropospheric moisture-flux (right) divergent component and (left) rotational component (kg m−1 s−1) anomalies of IOD_EXP1, IOD_EXP2, and ENSO_EXP for September–October. Shaded regions are the divergent (source) and convergent (sink) components of moisture flux (10−4 kg m−2 s−1)
Citation: Monthly Weather Review 134, 2; 10.1175/MWR3085.1
Vertically integrated moisture-flux (kg m−1 s−1) anomalies and shaded rainfall (mm day−1) anomalies of IOD_EXP1, IOD_EXP2, and ENSO_EXP simulations for (left) September–October and (right) NDJ
Citation: Monthly Weather Review 134, 2; 10.1175/MWR3085.1