## 2010 Vol. 27, No. 2

Display Method:
2010, 27(2): 195-210. doi: 10.1007/s00376-009-8212-6
Abstract:
Rainfall prediction remains one of the most challenging problems in weather forecasting. In order to improve high-resolution quantitative precipitation forecasts (QPF), a new procedure for assimilating rainfall rate derived from radar composite reflectivity has been proposed and tested in a numerical simulation of the Chicago floods of 17--18 July 1996. The methodology is based on the one-dimensional variation scheme (1DVAR) assimilation approach introduced by Fillion and Errico but applied here using the Kain-Fritsch convective parameterization scheme (KF CPS). The novel feature of this work is the continuous assimilation of radar estimated rain rate over a three hour period, rather than a single assimilation at the initial (analysis) time. Most of the characteristics of this precipitation event, including the propagation, regeneration of mesoscale convective systems, the frontal boundary across the Midwest and the evolution of the low-level jet are better captured in the simulation as the radar-estimated precipitation rate is assimilated. The results indicate that precipitation assimilation during the early stage can improve the simulated mesoscale feature of the convection system and shorten the spin-up time significantly. Comparison of precipitation forecasts between the experiments with and without the 1DVAR indicates that the 1DVAR scheme has a positive impact on the QPF up to 36 hours in terms of the bias and bias equalized threat scores.
2010, 27(2): 211-229. doi: 10.1007/s00376-009-8179-3
Abstract:
Using monthly mean sea ice velocity data obtained from the International Arctic Buoy Programme (IABP) for the period of 1979--1998 and the monthly mean NCEP/NCAR re-analysis dataset (1960--2002), we investigated the spatiotemporal evolution of the leading sea ice motion mode (based on a complex correlation matrix constructed of normalized sea ice motion velocity) and their association with sea level pressure (SLP) and the predominant modes of surface wind field variability. The results indicate that the leading winter sea ice motion modes spatial evolution is characterized by two alternating and distinct sea ice modes, or their linear combination. One mode (M1) shows a nearly closed cyclonic or anti-cyclonic circulation anomaly in the Arctic Basin and its marginal seas, resembling to a large extent the response of sea ice motion to the Arctic Oscillation (AO), as many previous studies have revealed. The other mode (M2) displays a coherent cyclonic or anti-cyclonic circulation anomaly with its center close to the Laptev Sea, which has not been identified in previous observational studies. In fact, M1 and M2 respectively reflect the responses of sea ice motion to two predominant modes of winter surface wind variability north of 70N, which well correspond, with slight differences, to the first two modes of EOF analysis of winter monthly mean SLP north of 70N. These slight differences in SLP anomalies lead to a difference of M2 from the response of sea ice motion to the dipole anomaly. Although the AO significantly influences sea ice motion, it is not crucial for the existence of M1. The new sea ice motion mode (M2) has the largest variance and clearly differs from the response of winter monthly mean sea ice motion to the dipole anomaly in SLP fields, and corresponding SLP anomalies also show differences compared to the dipole anomaly. This study indicates that in the Arctic Basin and its marginal seas, slight differences in SLP anomaly patterns can force distinctly different sea ice motion anomalies.
2010, 27(2): 230-242. doi: 10.1007/s00376-009-8170-z
Abstract:
Direct climate responses to dust shortwave and longwave radiative forcing (RF) are studied using the NCAR Community Atmosphere Model Version 3 (CAM3). The simulated RF at the top of the atmosphere (TOA) is -0.45 W m-2 in the solar spectrum and +0.09 W m-2 in the thermal spectrum on a global average. The magnitude of surface RF is larger than the TOA forcing, with global mean shortwave forcing of -1.76 W m-2 and longwave forcing of +0.31 W m-2. As a result, dust aerosol causes the absorption of 1.1 W m-2 in the atmosphere. The RF of dust aerosol is predicted to lead to a surface cooling of 0.5 K over the Sahara Desert and Arabian Peninsula. In the meantime, the upper troposphere is predicted to become warmer because of the absorption by dust. These changes in temperature lead to a more stable atmosphere, which results in increases in surface humidity. The upward sensible and latent heat fluxes at the surface are reduced, largely balancing the surface energy loss caused by the backscattering and absorption of dust aerosol. Precipitation is predicted to decrease moderately on a global scale.
2010, 27(2): 243-252. doi: 10.1007/s00376-009-8183-7
Abstract:
In this study, the barotropic stability of vortex Rossby waves (VRWs) in 2D inviscid tropical cyclone (TC)-like vortices is explored in the context of rotational dynamics on an f-plane. Two necessary instable conditions are discovered: (a) there must be at least one zero point of basic vorticity gradient in the radial scope; and (b) the relative propagation velocity of perturbations must be negative to the basic vorticity gradient, which reflects the restriction relationship of instable energy. The maximum growth rate of instable waves depends on the peak radial gradient of the mean vorticity and the tangential wavenumber (WN). The vortex-semicircle theorem is also derived to provide bounds on the growth rates and phase speeds of VRWs. The typical basic states and different WN perturbations in a tropical cyclone (TC) are obtained from a high resolution simulation. It is shown that the first necessary condition for vortex barotropic instability can be easily met at the radius of maximum vorticity (RMV). The wave energy tends to decay (grow) inside (outside) the RMV due mainly to the negative (positive) sign of the radial gradient of the mean absolute vorticity. This finding appears to help explain the developemnt of strong vortices in the eyewall of TCs.
2010, 27(2): 253-264. doi: 10.1007/s00376-009-8079-6
Abstract:
In this study, long-term (1777--1997) precipitation data for Seoul, Korea, wetness indices from eastern China, and modern observations are used to identify the interdecadal variability in East Asian summer monsoon precipitation over the last 220 years. In the East Asian monsoon region, two long-term timescales of dry--wet transitions for the interdecadal variability and quasi-40- and quasi-60-year timescales are dominant in the 220-year precipitation data of Seoul, as well as in the wetness indices over China. The wet and dry spells between Seoul (southern China) and northern China are out-of-phase (out-of-phase) at the quasi-60-year timescale, and in-phase (out-of-phase by approximately 90o before 1900 and in-phase after 1900) at the quasi-40-year timescale. In particular, during the last century, the dominant long-term timescales over East Asia tend to decrease from the quasi-60-year to the quasi-40-year with increasing time. The dominant quasi-40-year and quasi-60-year timescales of the Seoul precipitation in Korea are strongly correlated with these timescales of the northern Pacific Ocean.
2010, 27(2): 265-273. doi: 10.1007/s00376-009-8141-4
Abstract:
A distinct cold tongue has recently been noticed in the South China Sea during the winter monsoon, with the cold tongue temperature minimum occurring in the January or February. This cold tongue shows significant links with the Maritime Continents rainfall during the winter period. The cold tongue and its interaction with the Maritime Continents weather were studied using Reynolds SST data, wind fields from the NCEP--NCAR reanalysis dataset and the quikSCAT dataset. In addition, rainfall from the GOES Precipitation Index (GPI) for the periods 2000 to 2008 was also used. The propagation of the cold tongue towards the south is explained using wind dynamics and the western boundary current. During the period of strong cold tongue, the surface wind is strong and the western boundary current advects the cold tongue to the south. During the period of strong winds the zonal gradient of SST is high [0.5oC (25 km)-1]. The cold tongue plays an important role in regulating the climate over the Maritime Continent. It creates a zonal/meridional SST gradient and this gradient ultimately leads in the formation of convection. Hence, two maximum precipitation zones are found in the Maritime Continent, with a zone of relatively lower precipitation between, which coincides with the cold tongues regions. It was found that the precipitation zones have strong links with the intensity of the cold tongue. During stronger cold tongue periods the precipitation on either side of the cold tongue is considerably greater than during weaker cold tongue periods. The features of convection on the eastern and western sides of the cold tongue behave differently. On the eastern side convection is preceded by one day with SST gradient, while on the western side it is four days.
2010, 27(2): 274-284. doi: 10.1007/s00376-009-8071-1
[Abstract](648) [PDF 374KB](1431)
Abstract:
General circulation models (GCMs) are often used in assessing the impact of climate change at global and continental scales. However, the climatic factors simulated by GCMs are inconsistent at comparatively smaller scales, such as individual river basins. In this study, a statistical downscaling approach based on the Smooth Support Vector Machine (SSVM) method was constructed to predict daily precipitation of the changed climate in the Hanjiang Basin. NCEP/NCAR reanalysis data were used to establish the statistical relationship between the larger scale climate predictors and observed precipitation. The relationship obtained was used to project future precipitation from two GCMs (CGCM2 and HadCM3) for the A2 emission scenario. The results obtained using SSVM were compared with those from an artificial neural network (ANN). The comparisons showed that SSVM is suitable for conducting climate impact studies as a statistical downscaling tool in this region. The temporal trends projected by SSVM based on the A2 emission scenario for CGCM2 and HadCM3 were for rainfall to decrease during the period 2011--2040 in the upper basin and to increase after 2071 in the whole of Hanjiang Basin.
2010, 27(2): 285-292. doi: 10.1007/s00376-009-8173-9
Abstract:
The authors examine the effects of external forcing agents such as greenhouse gases (GHGs) and aerosols, as well as solar variability and ozone, on global land monsoon precipitation by using a coupled climate model HadGEM1, which was developed by the Met Office Hadley Centre for Climate Research. The results indicate that HadGEM1 performs well in simulating the observed decreasing trend of global land monsoon precipitation over the past 50 years. This trend mainly occurred in the Northern Hemisphere and is significantly different from the trend of natural variability due to ocean-atmosphere-land interactions. The coherence between the simulation and the observations indicates that the specified external forcing agents, including GHGs and aerosols as well as solar variability and ozone, are important factors that have affected the decreasing trend of global land monsoon precipitation in the past 50 years.
2010, 27(2): 293-302. doi: 10.1007/s00376-009-8112-9
Abstract:
In order to improve the efficiency of the Ocean Variational Assimilation System (OVALS), which has been widely used in various applications, an improved OVALS (OVALS2) is developed based on the recursive filter (RF) algorithm. The first advantage of OVALS2 is that memory storage can be substantially reduced in practice because it implicitly computes the background error covariance matrix; the second advantage is that there is no inversion of the background error covariance by preconditioning the control variable. For comparing the effectiveness between OVALS2 and OVALS, a set of experiments was implemented by assimilating expendable bathythermograph (XBT) and ARGO data into the Tropical Pacific circulation model. The results show that the efficiency of OVALS2 is much higher than that of OVALS. The computational time and the computer storage in the assimilation process were reduced by 83% and 77%, respectively. Additionally, the corresponding results produced by the RF are almost as good as those obtained by OVALS. These results prove that OVALS2 is suitable for operational numerical oceanic forecasting.
2010, 27(2): 303-314. doi: 10.1007/s00376-009-8132-5
Abstract:
The influence of soil moisture on Asian monsoon simulation/prediction was less studied, partly due to a lack of available and reliable soil moisture datasets. In this study, we firstly compare several soil moisture datasets over the Tibetan Plateau, and find that the remote sensing products from the Advanced Microwave Scanning Radiometer for Earth Observing System (AMSR-E) can capture realistic temporal variations of soil moisture better than the two reanalyses (NCEP and ECMWF) during the pre-monsoon seasons. Using the AMSR-E soil moisture product, we investigate the impacts of soil moisture over the Tibetan Plateau on Asian summer monsoon onset based on a Spectral Atmospheric Model developed at IAP/LASG (SAMIL). Comparison between results with and without the assimilation of remotely sensed soil moisture data demonstrates that with soil moisture assimilated into SAMIL, the land-sea thermal contrast during pre-monsoon seasons is more realistic. Accordingly, the simulation of summer monsoon onset dates over both the Bay of Bengal and South China Sea regions are more accurate with AMSR-E soil moisture assimilated. This study reveals that the application of the soil moisture remote sensing products in a numerical model could potentially improve prediction of the Asian summer monsoon onset.
2010, 27(2): 315-327. doi: 10.1007/s00376-009-8219-z
Abstract:
Persistent heavy rainfall events in South China can be divided into pre- and post-monsoon-onset events according to the onset of the South China Sea Summer Monsoon. In this study, daily rainfall data from 174 stations in South China and daily NCEP/NCAR reanalysis data are used to investigate pre-monsoon-onset events. The synoptic characteristics of pre-monsoon-onset heavy rainfall events are examined in detail. It is found that 21 heavy rainfall cases happened in the pre-monsoon period between 1961 and 2005. Among them, more than 60% of the events happened under a saddle pattern circulation. Using a case study, the role of the saddle field is investigated and slantwise vorticity development (SVD) theory is applied to diagnose the mechanisms for heavy rainfall development. It is found that a low-level saddle field and low-level jets result in the accumulation of warm moist air in the lower troposphere over South China and provide the necessary unstable conditions for heavy rainfall development. The existence of a saddle field plays an important role in maintaining these unstable conditions. The slantwise movement of the isentropic surface over South China can increase local vorticity and lead to strong vertical motion, which then triggers heavy rainfall.
2010, 27(2): 328-336. doi: 10.1007/s00376-009-9024-4
Abstract:
The regional heat flux exchange between heterogeneous landscapes and the nearby surface layer (SL) is a key issue in the study of land-atmosphere interactions over arid areas such as the Heihe River basin in northwestern China and in high elevation areas such as the Tibetan Plateau. Based on analysis of the land surface heterogeneity and its effects on the overlying air flow, the use of SL observations, atmospheric boundary layer (ABL) observations, and satellite remote sensing (RS) measurements along with three parameterization methodologies (here, termed as the RS, tile, and blending approaches) have been proposed to estimate the surface heat flux densities over heterogeneous landscapes. The tile and blending approaches have also been implemented during HEIhe basin Field Experiment (HEIFE), the Coordinated Enhanced Observing Period (CEOP) Asia-Australia Monsoon Project on the Tibetan Plateau (CAMP/Tibet), the Arid Environment Comprehensive Monitoring Plan95 (AECMP95), and the DunHuang Experiment (DHEX). The results showed that these two proposed parameterization methodologies can be accurately used over heterogeneous land surfaces.
2010, 27(2): 337-355. doi: 10.1007/s00376-009-8209-1
Abstract:
A new parameterization scheme of sea surface momentum roughness length for all wind regimes, including high winds, under tropical cyclone (TC) conditions is constructed based on measurements from Global Positioning System (GPS) dropsonde. It reproduces the observed regime transition, namely, an increase of the drag coefficient with an increase in wind speed up to 40 m s-1, followed by a decrease with a further increase in wind speed. The effect of this parameterization on the structure and intensity of TCs is evaluated using a newly developed numerical model, TCM4. The results show that the final intensity is increased by 10.5% (8.9%) in the maximum surface wind speed and by 8.1 hPa (5.9 hPa) in the minimum sea surface pressure drop with (without) dissipative heating. This intensity increase is found to be due mainly to the reduced frictional dissipation in the surface layer and little to do with either the surface enthalpy flux or latent heat release in the eyewall convection. The effect of the new parameterization on the storm structure is found to be insignificant and occurs only in the inner core region with the increase in tangential winds in the eyewall and the increase in temperature anomalies in the eye. This is because the difference in drag coefficient appears only in a small area under the eyewall. Implications of the results are briefly discussed.
2010, 27(2): 356-370. doi: 10.1007/s00376-009-8091-x
Abstract:
An observational analysis of satellite blackbody temperature (TBB) data and radar images suggests that the mesoscale vortex generation and merging process appeared to be essential for a tropical-depression-related heavy rain event in Shanghai, China. A numerical simulation reproduced the observed mesoscale vortex generation and merging process and the corresponding rain pattern, and then the model outputs were used to study the related dynamics through diagnosing the potential vorticity (PV) equation. The tropical depression (TD) was found to weaken first at lower levels and then at upper levels due to negative horizontal PV advection and diabatic heating effects. The meso-vortices developed gradually, also from the lower to the upper levels, as a result of positive horizontal PV advection and diabatic heating effects in the downshear left quadrant of the TD. One of these newly-generated vortices, V1, replaced the TD ultimately, while the other two, V2 and V3, merged due to the horizontal PV advection process. Together with the redevelopment of V1, the merging of V2 and V3 triggered the very heavy rain in Shanghai.
2010, 27(2): 371-381. doi: 10.1007/s00376-009-9047-x
Abstract:
In this paper, the role of westerly winds at southern high latitudes in global climate is investigated in a fully coupled ocean-atmosphere general circulation model. In the model, the wind stress south of 40oS is turned off with ocean and atmosphere fully coupled both locally and elsewhere. The coupled model explicitly demonstrates that a shutdown of southern high latitude wind stress induces a general cooling over the Antarctic Circumpolar Current (ACC) region, with surface Ekman flow and vertical mixing playing competitive roles. This cooling leads to an equatorward expansion of sea ice and triggers an equivalent barotropic response in the atmosphere to accelerate westerly anomalies. The shutdown of southern high latitude wind stress also significantly reduces global meridional overturning circulation (MOC). The Antarctic MOC (AnMOC) nearly disappears while the Atlantic MOC (AMOC) is weakened by 50%, suggesting a strong control of the southern high latitude winds over the thermohaline circulation (THC). In spite of a substantial weakening of the AMOC, the interhemispheric SST seesaw appears to be not significant due to an equatorward extension of the southern extratropical cooling through coupled wind-evaporation-SST (WES) feedback. In addition, it is found that the weakening of Atlantic MOC by as much as 50% is capable of cooling the time mean subpolar Atlantic temperature by only about 1oC.
2010, 27(2): 382-392. doi: 10.1007/s00376-009-8164-x
Abstract:
The AREMv2.3 mesoscale numerical model is used to explore storm processes in South China during the pre-rainy season in 2006 by imposing perturbations on the initial fields of physical variables (temperature, humidity, and wind fields). Sensitivity experiments are performed to examine the impacts of initial uncertainties on precipitation, on the error growth, and on the predictability of mesoscale precipitation in South China. The primary conclusion is that inherent initial condition uncertainties can significantly limit the predictability of storm. The 24-h accumulated precipitation is most sensitive to temperature perturbations. Larger-amplitude initial uncertainties generally lead to larger perturbation energies than those with smaller amplitude, but these kinds of differences decrease with time monotonically so the mechanism for the growth of perturbation energy is nonlinear. The power spectra of precipitation differences indicate that predictability increases with accumulated time. This also indicates the difficulties faced for short-term, small-scale precipitation forecasting.
2010, 27(2): 393-402. doi: 10.1007/s00376-009-8195-3
Abstract:
A fully coupled ocean-atmosphere model is applied to highlight the mechanism of the long-term variability (including decadal and longer time scales) in the Pacific Ocean. We are interested in the effect of ocean-atmosphere coupling of different regions during these processes. The control run successfully simulates the Pacific long-term variability, whose leading modes are the Pacific (inter) Decadal Oscillation (PDO) and the North Pacific mode (NPM). Furthermore, three numerical experiments are conducted, shutting down the ocean-atmosphere coupling in the North Pacific, the tropical Pacific, and the South Pacific, respectively. The results show that regional ocean-atmosphere coupling is not only important to the strength of local long-term SST variability but also has an influence on the variability further afield. In both the tropical Pacific and North Pacific, this local effect is the main control, which is much more obvious in the tropical regions. The existence of the PDO is extremely dependent on the coupling in the tropical Pacific. However, extratropical coupling, in particular that in the North Pacific, is also important to form its spatial pattern and strengthen the variability in some tropical areas. For the NPM, its existence is primarily determined by the coupling in the North Pacific.
2010, 27(2): 403-420. doi: 10.1007/s00376-009-9005-7
Abstract:
A number of AGCM simulations were performed by including various land--sea distributions (LSDs), such as meridional LSDs, zonal LSDs, tropical large-scale LSDs, and subcontinental-scale LSDs, to identify their effects on the Asian monsoon. In seven meridional LSD experiments with the continent/ocean located to the north/south of a certain latitude, the LSDs remain identical except the southern coastline is varied from 40o to 4oN in intervals of 5.6o. In the experiments with the coastline located to the north of 21oN, no monsoon can be found in the subtropical zone. In contrast, a summer monsoon is simulated when the continent extends to the south of 21oN. Meanwhile, the earlier onset and stronger intensity of the tropical summer monsoon are simulated with the southward extension of the tropical continent. The effects of zonal LSDs were investigated by including the Pacific and Atlantic Ocean into the model based on the meridional LSD run with the coastline located at 21oN. The results indicate that the presence of a mid-latitude zonal LSD induces a strong zonal pressure gradient between the continent and ocean, which in turn results in the formation of an East Asian subtropical monsoon. The comparison of simulations with and without the Indian Peninsula and Indo-China Peninsula reveals that the presence of two peninsulas remarkably strengthens the southwesterly winds over South Asia due to the tropical asymmetric heating between the tropical land and sea. The tropical zonal LSD plays a crucial role in the formation of cumulus convection.
2010, 27(2): 421-426. doi: 10.1007/s00376-009-8216-2
[Abstract](710) [PDF 286KB](1009)
Abstract:
The relationship between the radiance ratio (radiance at wavelength 450 nm to 650 nm) and aerosol optical depth (AOD) is analyzed in this paper by numerical simulation and a LUT' (look-up table) approach is then presented for the retrieval of AOD from the radiance ratio. In this LUT approach, the AOD retrieval error depends mainly on the assumption of aerosol types. From the preliminary simulation, a typical error of 15%--20% in AOD obtained from the radiance ratio is estimated, due to neglecting changes in the ground albedo and background atmosphere. At its worst, the AOD error reached a maximum of around 50%, which will be refined in the future. In the latter part of the paper, comparisons are made between AOD from the imager and from the CE-318 sun photometer, both located at Xianghe observatory in Hebei Province (39.75oN, 116.96oE). This field experiment shows that AOD from the imager is highly correlated with AOD from the sun photometer, with a correlation coefficient of 0.95 and an average retrieval error of around 7%. A contrast experiment confirms the feasibility of retrieving AOD from all-sky images, but more analysis and future research are required to improve the accuracy.
2010, 27(2): 427-442. doi: 10.1007/s00376--009-9006--6
Abstract:
Using the regional terrestrial Net Primary Production (NPP) from different observations and models over China, we validated the NPP simulations and explored the relationship between NPP and climate variation at interannual and decadal scales in the Modified Sheffield Dynamic Global Vegetation Model (M-SDGVM) during 1981--2000. M-SDGVM shows agreement with the NPP data from 743 sites under the Global Primary Production Data Initiative (GPPDI). The spatial and the zonal averaged NPP of M-SDGVM agree well with different historic datasets and are closest to the IGBP NPP. Compared to the 1980s, NPP in the 1990s increases in most of China with a high degree of spatial heterogeneity. The multi-year mean NPP of forest types is reasonably modeled (above 500 g C m-2 yr-1 while that of C3 path of photosynthesis (C3 grasslands is underestimated. The NPP of 7 M-SDGVM main plant functional types (PFTs) increases and the increment of the broad-leaved deciduous forest is the most obvious (5.05 g C m-2 yr-1. During the studied period, the annual NPP of M-SDGVM over China increases, with significant fluctuations, at an average rate of 0.0164 Gt C yr-1. Regulated by annual temperature and precipitation, the interannual variation of the total NPP shows more significant correlation with temperature (relativity and probability are R=0.61, P=0.00403) than precipitation (R=0.40, P=0.08352). CO2 fertilization may play a key role in the increase of terrestrial ecosystem NPP over continental China, and CO2 stimulation increases with CO2 concentrations, and also with the climate variability of the 1980s and 1990s.
2010, 27(2): 443-454. doi: 10.1007/s00376-009-9106-3
[Abstract](643) [PDF 883KB](1312)
Abstract:
This study examined the impact of tropical cyclone (TC) landfalls on the spatiotemporal variations in the rainfall over South China for the period 1957--2005. The target region was selected to show the noteworthy contribution of TC landfalls to the total rainfall during the typhoon season (July--October). Two prevailing spatial variations in the rainfall were obtained from an EOF analysis. The first EOF mode displays single-sign variability over South China with an explained variance of 23.4%. The associated time series of this mode fluctuates on a decadal timescale and was found to be correlated with TC genesis in the South China Sea. The second EOF mode shows a seesaw pattern between Hainan Island/Guangdong Province and the remaining regions with an explained variance of 11.4%. This seesaw pattern results from an anti-correlation in seasonal TC landfalls between the two regions, which was found in previous studies. This is related to the strengthening (weakening) of the upper tropospheric jets and the corresponding development of a massive anticyclonic (cyclonic) circulation over East Asia. The EOF analysis was also conducted using just the data for rainfall caused by landfalling TCs. This revealed that the first EOF mode using just the TC-induced rainfall is nearly identical to the second mode from the total rainfall. The obvious seesaw pattern of the first mode when employing just the TC-induced rainfall in the EOF analysis implies that this pattern has larger temporal variability than the single-signed pattern (i.e., the first EOF mode using the total rainfall) in terms of TC landfalls. This study suggests that TC landfalls over South China and the accompanying rainfall significantly modulate the spatial variation of the typhoon season rainfall there.