TMCnet News
Remote Trigger of Deep Convection by Cold Outflow over the Taiwan Strait in the Mei-Yu Season: A Modeling Study of the 8 June 2007 Case [Monthly Weather Review](Monthly Weather Review Via Acquire Media NewsEdge) ABSTRACT In this study, the heavy-rainfall event over central Taiwan during the mei-yu season on 8 June 2007 is investigated, with an emphasis on the triggering mechanism for the deep convection that produced the rain. Observations indicate that there existed two lines of forcing with convection prior to the rain: one over the northern Taiwan Strait along the mei-yu front and the other over the southern Taiwan Strait. Yet, the convection in question developed over the central strait between these two lines, in an unstable environment with strong westerly vertical wind shear. This motivated the authors to carry out the present study. The Cloud-Resolving Storm Simulation (CReSS) of Nagoya University was used and the event was reproduced at a horizontal grid size of 2 km, including the initiation of new convection over the central strait at the correct location and time. The model results suggest a crucial role played by the series of active, persistent, and propagating storms in the southern strait (along the aforementioned second forcing line). On their back (northern) side, these storms repeatedly produced pulses of cold outflow that traveled toward the north-northeast with positive pressure perturbation. With characteristics of gravity waves, the perturbation propagated faster than the cold air and the associated increase in forward-directed (horizontal) pressure gradient force led to northward acceleration of near-surface flow (by up to 4-5 m s-1 h-1). The stronger southerly flow in turn enhanced downstream convergence, and the deep convection was triggered in the central strait near the arrival of the gravity wave ahead of the cold air. When the convection moved eastward over Taiwan, heavy rainfall resulted. The mechanism presented here for remote triggering of convection over the ocean has not been documented near Taiwan during the mei-yu season. With a better understanding about the behavior of convection, these results can contribute to the improvement of quantitative precipitation forecasts and hazard prevention and reduction. (ProQuest: ... denotes formula omitted.) 1. Introduction During the mei-yu season (May-June) in Taiwan, a subtropical island with steep and complex topography in East Asia (cf. Fig. 1), heavy rainfall often occurs under the influences of the mei-yu front and the prefrontal moist and unstable southwesterly flow (e.g., Kuo and Chen 1990; Chen 1992, Chen and Li 1995; Chen et al. 2008). When a mei-yu front approaches Taiwan, deep convection and organized mesoscale convective systems (MCSs) responsible for the heavy rain can be initiated by frontal uplift, topographic uplift, dynamical forcing (of synoptic or mesoscale) associated with the front, or a combination of these processes (e.g., Chen 1993; Li et al. 1997). The topography of Taiwan, with a length of about 350 km and the highest peak reaching almost 4 km (Fig. 1), often exerts a significant blocking effect on the prevailing low-level flow. The blocking effect can result in flow deceleration and deflection (e.g., Banta 1990; Baines 1995), localized barrier jet (Li and Chen 1998; Chen et al. 2005), and leeside vortex and mesolow (e.g., Smolarkiewicz and Rotunno 1989; Lin et al. 1992; Schar and Smith 1993). It also alters the configuration and location of the low-level forcing to convection (e.g., Durran 1986; Lin 1993; Wang et al. 2005). Besides the role of the terrain on flow dynamics, the island of Taiwan can also exhibit strong thermodynamic effects in early summer, through daytime solar heating (and nighttime cooling) and the development of diurnal circulation (such as land-sea and mountain valley breezes) over complex terrain (e.g., Akaeda et al. 1995; Chen et al. 1999).While these effects alsomodify the stability of the flow (e.g., Georgelin et al. 1996), they provide local forcing for convective development and are responsible for the pronounced diurnal variation in both total rainfall and the heavy-rainfall frequency over Taiwan in the mei-yu season (e.g., Yeh and Chen 1998; Chen et al. 1999). It is well recognized that existing precipitation systems can also provide forcing to trigger new convection through a variety ofmechanisms. These include the uplift at the density currentlike gust front associated with the precipitation-driven, evaporation-cooled outflow (e.g., Purdom 1976; Stobie et al. 1983; Wilson and Schreiber 1986; Koppel et al. 2000), and collision of outflow boundary with a front or dryline (e.g., Bluestein and Weisman 2000; Kingsmill and Crook 2003) or those from multiple storms (e.g., Scofield and Purdom 1986; Lee et al. 2006). In some cases, new convection can be triggered by the advancing gravity wave (often as a bore) well ahead of the cold air source, by a few hundred kilometers (e.g., Carbone et al. 1990; Karyampudi et al. 1995). When interactions among rainfall systems are important, it is obvious that successful forecasts of rainfall for a given area at later times depend heavily on those at earlier times. Because of the complicated and nonlinear interactions among the above-mentioned dynamic and thermodynamic processes with the presence of steep terrain, it is thus often difficult to accurately predict the initiation, development, and dissipation of deep convection and MCSs around Taiwan, and consequently the occurrence of heavy rainfall at the correct location and time (e.g., Chien and Jou 2004). With an ultimate goal to improve the capability of quantitative precipitation forecasting and estimation (QPF/ QPE) during the Asian summer monsoon season, the Southwest Monsoon Experiment/Terrain-influenced Monsoon Rainfall Experiment (SoWMEX/TiMREX) field campaign was carried out in the Taiwan area in May- June 2008, with a pilot experiment inMay-June 2007 (Jou and Lee 2008). Among the various mechanisms mentioned above, the interactions of convective cells or MCSs and the impacts of cold outflows are often at cloud scale and beyond the resolution of conventional data (e.g., Einaudi et al. 1989; Powers 1997; Zhang and Koch 2000). Over nearby oceans such as the Taiwan Strait (cf. Fig. 1), similar processesmay also take place during periods of active convection in the mei-yu season. However, this type of study is rare owing to a lack of observational data in general and those at key locations and time with sufficient resolution in specific. Therefore, how such processes can trigger new convection is not known in details. In this work, we present a modeling study on the heavy-rainfall event on 8 June 2007 during the SoWMEX/TiMREX pilot experiment. It is shown that through enhanced near-surface convergence, new convection over the Taiwan Strait can be remotely triggered by gravity waves ahead of the accelerating cold air outflow from repeated convection farther to the south. The paper is organized as follows. The data and methodology in this study are described in section 2, and section 3 gives an overview on the synoptic environment and the evolution of the present case. In section 4, the numerical model and experiments are described, and the simulation results are presented in section 5. Section 6 provides further discussion, and section 7 gives the concluding summary. 2. Data and methodology The data used in the present study for synoptic discussion include the European Centre for Medium- Range Weather Forecasts (ECMWF) gridded analyses with a horizontal resolution of 1.1258 latitude-longitude at 15 pressure p levels (1000-10 hPa) every 6 h (at 0200, 0800, 1400, and 2000 LST) covering the case period over 7-8 June 2007. The dataset also provides surface variables. At the mean sea level (MSL), subjective manual analysis was performed to provide more detailed information, especially on the location of the surface mei-yu front. At synoptic times, this manual analysis also made use of the twice-daily Quick Scatterometer (QuikSCAT) oceanic winds as an aid for winds over the Taiwan Strait and nearby oceans. To evaluate thermodynamic conditions and vertical shear in the environment, sounding data atMakung over the Taiwan Strait at 12-h intervals (at 0800 and 2000 LST) and froman aircraft dropsonde mission are used (cf. Fig. 1). To document and examine the evolution of the deep convection, both reflectivity composites of vertical maximum-echo indicator (VMI) from the operational radars in Taiwan at 10-min intervals and the infrared cloud imageries from the Multifunctional Transport Satellite (MTSAT) every 1 h are employed. For rainfall distribution over Taiwan, hourly accumulation data from a network of nearly 400 automatic gauges are used. All the above observational data were compared with model outputs to validate simulation results where appropriate. However, not all such comparisons will be shown. 3. Synoptic and case overview a. Synoptic environment In this section, the synoptic environment of the present case is reviewed. The surface weather maps produced through manual analysis, with the aid from ECMWF data and QuikSCAT winds, are presented in Fig. 2. At 2000 LST (1200 UTC) 7 June, the mei-yu front west of 121°E was oriented roughly east-west (E-W) and extended into southern China along about 24.8°N(Fig. 2a). On the other hand, the front east of Taiwan showed a northeast-southwest alignment and had already advanced to 22.5°N along the eastern coast of Taiwan. Such a separation of frontal segments is often observed as a result of the blocking effect by the terrain of northern Taiwan (e.g., Li and Chen 1998). Twelve hours later at 0800 LST (0000 UTC) 8 June (Fig. 2b), the front over the northern Taiwan Strait moved ahead only slightly but became more wavy, while the eastern frontal segment also remained stationary. At this time, southwesterly flow existed at Makung (cf. Fig. 1) to the south of the front over the Taiwan Strait. The ECMWF (1.1258) analyses at 2000 LST 7 June indicate that the mei-yu front was located roughly along 27°N at 850 hPa and mostly north of 28°N at 700 hPa (Figs. 3a,b), thus suggesting a northward tilt of the front with elevation. South of the front, the westerly wind speed increased rapidly with height over the strait, to reach about 15 m s-1 at 700 hPa and exceed the 12.5 m s-1 criterion often used for low-level jets (LLJs; e.g., Chen and Yu 1988). During the next 12 h, low-level winds in the region strengthened considerably and reached almost 20 m s-1 near 22.5°N, 116°E at 700 hPa (Figs. 3c,d). Thus, strong westerly vertical wind shear existed and intensified over the Taiwan Strait during the case period, a condition favorable for sustaining organized convection (e.g., Rotunno et al. 1988; LeMone et al. 1998). Farther aloft at 500 hPa, there was an approaching short-wave trough near 25°N, 113°E and westerly winds of 10-15 m s-1 prevailed over the Taiwan area at 2000LST 7 June (Fig. 4a). Thus, the strong vertical shear was confined only at low levels. At 200 hPa, a westerly jet streak appeared along about 30°N, with its core (about 60 m s-1) near 29°N, 126°E (Fig. 4b). The northern Taiwan Strait and its adjacent area were located under the rear-right quadrant of the jet streak. Over the southern strait, clear directional diffluence also existed within the northwesterly flow. The upper-tropospheric divergence at the south side of the entrance region of a jet streak and diffluence were both favorable for the development of convection (e.g., Uccellini 1990) in our case. b. Thermodynamic and shear conditions The sounding taken at Makung (46734, cf. Fig. 1) at 2000 LST 7 June, prior to the convection south of the front (cf. Fig. 2a) is presented in Fig. 5. The temperature T lapse rate indicates conditional instability below 600 hPa, with high moisture content through deep layers. Using an air parcel mixed over the lowest 50 hPa (e.g., Thompson et al. 2003), the lifting condensation level (LCL) is quite low at 960 hPa, but the level of free convection (LFC) is considerably higher at 837 hPa with a convective inhibition (CIN) of 12.8 J kg-1 (Fig. 5). The convective available potential energy (CAPE) is appreciable at 834.5 J kg-1, but its shape is narrow in the lowest 3 km and indicates only weak buoyancy without external forcing. Consistent with Figs. 2-5, the hodograph shows large westerly vertical wind shear below 700 hPa, especially below 850 hPa (Fig. 6a). The 0-3-km speed difference was about 19 m s-1 and equivalent to a bulk shear (BS) of 6.3 × 10-3 s-1. From 600 to 350 hPa, the westerly winds gradually decreased in speed and turned into west-northwesterly flow at 200 hPa and above. Twelve hours later at 0800 LST 8 June, there was a significant increase in low-level southwesterly flow up to about 700 hPa at Makung, especially at the lowest 1 km (Fig. 6b). This phenomenon will be shown to be linked to the initiation of deep convection over central Taiwan Strait, and will be further elaborated upon in section 5. On the other hand, thermodynamic conditions at Makung only changed marginally at 0800 LST 8 June (not shown), and the CAPE (of a mixed parcel) increased to 904.7 J kg-1 mainly due to a slight cooling in the upper-troposphere brought about by the west-northwesterly flow (cf. Figs. 4b and 6). c. Case overview The heavy-rainfall event over Taiwan on 8 June 2007 during the SoWMEX/TiMREX pilot experiment is reviewed in this subsection. The distribution of the daily total rainfall (Fig. 7) reveals three distinct and separate precipitation centers over northern, central, and southern Taiwan. Among them, the rain over central Taiwan had the highest amount (>250 mm) and its area extended from the coast to the mountainous island interior (cf. Fig. 1). The peak rainfall amount in northern Taiwan was near the coast and also approached 250 mm, while the mountain area in southern Taiwan received 150-200 mm at most (Fig. 7). The evolution of deep convection that led to the heavy rainfall on 8 June 2007 can be examined in detail using radarVMI reflectivity composites every 10 min (cf. Fig. 7 for radar locations). Here, the composites at selected times every 1, 2, or 3 h are shown in Fig. 8, in which horizontal winds and frontal position are also plotted when the ECMWF surface data are available to aid to the discussion. At 0200 LST (Fig. 8a), deep convection had already broken out over the northern Taiwan Strait (cf. Fig. 1) along the mei-yu front with an east-northeast-westsouthwest (ENE-WSW) orientation (labeled as group A). Active convection also appeared at the southern strait over 228-23°N and aligned roughly from the westnorthwest (WNW) to the east-southeast (ESE, marked as group B). Over central Taiwan Strait, meanwhile, convection existed (near 23.6°N, 118.8°E) and propagated eastward with a north-south (N-S) organization (marked as C). Consistent with Fig. 2, the surface front east of Taiwan had advanced farther south, by about 250 km. The same basic configuration of convection persisted through 0400 LST (Figs. 8b,c), but line C weakened and a second line about 120 km behind (marked as D in Fig. 8a) moved into the central strait. These lines (C and D) were ahead of the 500-hPa trough (Fig. 4a) and their alignment and propagation directions were favored under the strong low-level westerly shear (e.g., Rotunno et al. 1988; LeMone et al. 1998). By 0400 LST, new cells (>50 dBZ) were triggered within line D and the line redeveloped actively, especially near 24°N, 119°E (marked by an arrow, Fig. 8c). Afterward, these cells grew stronger andmoved eastward (Figs. 8d-f) and eventually caused the heavy rainfall in centralTaiwan (cf. Fig. 7). FromFig. 8, it is clear that the rainfall over northern Taiwan was caused by the convection along themei-yu front (i.e., groupA), while that in southern Taiwan occurred mostly later in the day after 0800 LST (Fig. 8f), particularly during 1500- 2000 LST (not shown). In Fig. 9, infrared cloud imageries from the Japanese geostationary satellite, the Multifunctional Transport Satellite (MTSAT), at 2-h intervals are presented. The convection over northern and southern Taiwan Strait during 0000-0400 LST was also nicely depicted (Figs. 9a-c). It can be confirmed that the second N-S-oriented convective line seen in Figs. 8a,b (i.e., line D) over the central strait was not active as it moved offshore from China, but redeveloped vigorously south of the front prior to 0400 LST (marked by an arrow, Fig. 9c). In addition, most of the deep convective cells over southern Taiwan Strait had their origin near the coastline of China, especially those farther to the west (not shown). By 0600 LST, the upper-level anvil clouds produced by the convection over northern and central strait had merged together on cloud imageries (Fig. 9d), but the cells remained distinct and separate on radar composites (Fig. 8d). From the case overview above, it is shown that there were two preexisting lines of active convection in the vicinity of Taiwan on 8 June 2007: the first over the northern Taiwan Strait along the mei-yu front (group A) and the second in the southern strait with a WNW-ESE alignment (group B). The cells responsible for the heaviest rainfall, however, were triggered between them in a prefrontal environment (Figs. 8 and 9). It is therefore intriguing to investigate what triggering mechanism was responsible for the active redevelopment of the convection over the central strait near 0400 LST 8 June 2007, when other forcing, apparently stronger, also existed in the region. This question provided the motivation for us to carry out the current study, primarily using a modeling approach. 4. Model and experiments a. Description of the CReSS model The Cloud-Resolving Storm Simulation (CReSS) model used here (version 2.2) is the same as in Wang et al. (2009), and is a nonhydrostatic, fully compressible, cloud model developed at the Hydrospheric Atmospheric Research Center of Nagoya University, Japan (Tsuboki and Sakakibara 2002, 2007). This model employs a height-based terrain-following vertical coordinate, with prognostic equations for 3D momentum (u, y, w), p, potential temperature (u), and mixing ratios of water vapor (qy) and other condensates. To simulate clouds at high resolution, an explicit bulk cold-rain scheme based on Lin et al. (1983), Cotton et al. (1986), Murakami (1990), Ikawa and Saito (1991), and Murakami et al. (1994) are used without any cumulus parameterization. This scheme includes a total of six species (water vapor, cloud water, cloud ice, rain, snow, and graupel) and microphysical processes of nucleation (condensation), sublimation, evaporation, deposition, freezing, melting, falling, conversion, collection, aggregation, and liquid shedding (Tsuboki and Sakakibara 2002). Subgrid-scale turbulent mixing is parameterized using 1.5-order closure with turbulent kinetic energy (TKE) prediction (Tsuboki and Sakakibara 2007), and planetary boundary layer (PBL) processes are parameterized following Mellor and Yamada (1974) and Segami et al. (1989). Momentum and energy fluxes and radiation at the surface are also considered with a substrate model (Kondo 1976, Louis et al. 1981; Segami et al. 1989), but the cloud longwave radiation is neglected. Also, numerical diffusion (fourth order) is used to prevent spurious waves (e.g., Lindzen and Fox-Rabinovitz 1989; Persson and Warner 1991). The CReSS model adopts the Arakawa-C staggered (horizontal) and Lorenz (vertical) grid structure without nesting. For computational efficiency, a time-splitting scheme (Klemp and Wilhelmson 1978) is used, with filtered leapfrog method (Asselin 1972) and the implicit Crank-Nicolson scheme for the integration at large and small time steps, respectively. b. Model data and experiment design To properly simulate the development and evolution of deep convection, two experiments were performed at successively higher resolution but for a smaller domain, as shown in Fig. 1 and Table 1. First, a coarse 10-km run (experiment 1) was carried out with 40 vertical layers, using the 6-hourly ECMWF 1.1258 gridded analyses as initial and lateral boundary conditions (IC/LBC). The starting time of this experiment is 2000 LST 6 June 2007 with an integration length of 72 h, and it is designed to capture the evolution of the environmental atmosphere surrounding Taiwan during the case period. Then, a 50- layer high-resolution 2-km run (experiment 2) was performed using 1-h outputs from the 10-km run as IC/LBC, for 60 h starting at the same time (2000 LST 6 June) with an output frequency of 10 min (Table 1). Although even finer horizontal grid sizes have been used successfully for the CReSS model (e.g., Liu et al. 2004; Wang and Huang 2009), the phenomena of our interests (i.e., convective activities over the Taiwan Strait) are reasonably well reproduced, as will be shown in section 5. For both runs, a dataset of real topography at a horizontal resolution of 30 s [or (1/ 120)°, roughly 900 m] and the observed weekly SST at 1° × 1° resolution (Reynolds et al. 2002) were provided at the lower boundary (Table 1). In the following section, the CReSS model results will be compared with observations, validated, and further used to investigate the mechanism responsible for the redevelopment of deep convection over central Taiwan Strait. For these purposes, only the 2-km results (interpolated from terrain-following coordinate onto constant height levels) are presented. 5. Model results a. Overall results and comparison with observations The 2-km CReSS model simulated surface fields at 0200 LST 8 June 2007 are first compared with the ECMWF1.1258 analyses in Fig. 10. This time is 30 h into the forecast and roughly 2 h before the development of the convection responsible for the heavy rainfall in central Taiwan (cf. Fig. 8). The frontal positions in the model are in good agreement with both the ECMWF data and manual analysis (cf. Figs. 10 and 2). Over the ocean (and low-lying land areas below 50 m) where model p and T are plotted, their general patterns are also similar, except that deep convection over southern Taiwan Strait already appears in the model as observed, thus producing cloud-scale perturbations (Figs. 8a and 10). Likewise, both ECMWF data and model results show southwesterly to westerly surface winds over the South China Sea and the Luzon Strait (south of about 22°N, cf. Fig. 1), but the model produces stronger southerly wind components over the Taiwan Strait north of the convection. The thermodynamic parameters produced by the CReSSmodel run are also compared with observations in Table 2 for two occasions.One is atMakung for 2000 LST and the other is at the location of a dropsonde (cf. Fig. 1) deployed during the SoWMEX/TiMREX pilot experiment for 2310 LST, both on 7 June 2007. In both instances, most model parameters are similar compared to observations using well-mixed air parcels over the lowest 50 hPa, with LCL at 390-470 m, LFC near 0.5-1.5 km, small CIN, and appreciable CAPE (at least 830 J kg-1; Table 2). It is noted that CAPE values were larger and approaching 3000 J kg-1 at the dropsonde site over the southernTaiwan Strait, where deep convection developed prior to 0000 LST 8 June 2007 (cf. Fig. 9a). The comparison of hodographs at Makung suggests that the model westerly winds at 2000 LST 7 June are somewhat too weak at 850-600 hPa but slightly too strong at 500-300 hPa (Figs. 6a and 11a, also Table 2). As the low-level winds from the dropsonde were affected by a nearby storm (not shown), the grid point chosen for comparison from the model output (at 22.48°N, 118.67°E) is not identical, but has a similar location relative to the storm. As shown in Figs. 11b,c, the two wind profiles are quite close, except that the model has a stronger southerly component below about 800 hPa. Overall, themodel simulation is in general agreement with the observation for thermodynamic and wind conditions. Figure 12 presents the column-maximum mixing ratio of precipitating hydrometeors (rain, snow, plus graupel), surface winds, and frontal positions as simulated by the 2-kmCReSSmodel run, and can be compared with Fig. 8. At 0300 LST 8 June (Fig. 12a), deep convective cells about 100-180 km apart are already present over the southern Taiwan Strait with a WNW-ESE orientation. Convection also develops by 0400 LST along the stationary surface mei-yu front over the northern strait (Fig. 12b). Thus, the two preexisting lines are reproduced in the model while the N-S oriented, weakening lines are not (cf. Fig. 8), although widespread stratiform precipitation is simulated over the central strait. Nonetheless, the 2-km run captures the break out of new convection near 24°N, 119°E, roughly 65 km south of the front just before 0400 LST (Fig. 12b), and thus the observed redevelopment in the central strait at the correct time (cf. Fig. 8c). After initiation, this new cell strengthens and travels eastward at 0500 LST (Fig. 12c), merges into and becomes the southern part of the frontal convective line during 0600- 0800 LST (Figs. 12d-f), and eventually makes landfall over central Taiwan after 1000 LST (Figs. 12g,h). Thus, the evolution shown in Fig. 12 agrees reasonably well with the observation in Fig. 8, most notably for the convection ultimately leading to the heavy rainfall in central Taiwan. In Fig. 12, the front over the northern strait also gradually develops a wavy pattern due to the cold air associated with convective downdrafts on the back side. This is not fully reflected in either the manual analysis or the ECMWF data (e.g., at 0800 LST; Figs. 2b and 8e) mainly because of a lack of observation and insufficient data resolution. The observed and model-simulated 6-h rainfall during 0200-0800, 0800-1400, and 1400-2000 LST 8 June 2007 are shown in Fig. 13. The general rainfall patterns for each 6-h periods are quite similar, and the modeled area of heavy rainfall also extends from the coast into the mountainous interior of central Taiwan (Figs. 13e,f) as observed (Figs. 13b,c). However, the heavy rain along the front just misses the northern tip of Taiwan and the amount is not enough over southern Taiwan in the model (Figs. 13e,f). Nevertheless, since the convection responsible for the heavy rainfall is captured (Fig. 12), results of the 2-km run can be further utilized to investigate its initiation process over the central strait. b. Triggering mechanism of convection over the central Taiwan Strait Figure 14 presents the simulated surface conditions at the lowest model level of 50 m leading to the initiation of the convection in central Taiwan Strait. At 0000 LST 8 June (Fig. 14a), the modeled convection already exists over the southern strait along a WNW-ESE orientated wind-shift line, with southwesterly flow of 7-8 m s-1 ahead (to the south) and weak southerly flow behind (to the north). In fact, these storms have existed for at least several hours in both the observation and the model (not shown), and are presumably sustained (and regenerated) through cold-pool dynamics (e.g., Weisman and Klemp 1986; Rotunno et al. 1988). Across the wind-shift line with convergence, the ? difference is typically about 2°-3°C, while pockets of cold air (i.e., cold pools) at the back side (often with ?, 26°C) is associated with divergence (=10-3 s-1 in isolated regions, Fig. 14a). This basic pattern continues through 0400 LST (Figs. 14b,c), at which time the cold pockets can be confirmed to collocate with precipitation (Fig. 12b) and are thus produced by downdrafts through evaporative cooling. Also in agreement with Fig. 12, convergence appears near 24°N, 119°E at 0400 LST and coincides with the initiation of the convection of our interest (Fig. 14c). Prior to this time, it is evident in Fig. 14 that the southwesterly flow just upstream from the area of initiation (to its southwest) gradually intensifies, from about 3 to almost 10 m s-1. At 24°N, 119°E, this enhancement of near-surface wind in themodel is especially significant below about 1.5 km(by as much as 4-5 m s-1) from 0300 to 0400 LST (Fig. 15), and is also reflected in (and supported by) the observation at Makung (Fig. 6). Note that in both Figs. 6 and 15, only near-surface winds intensify at a direction (from the southwest) different from the westerly LLJ at 700 hPa (Fig. 3), so speed increase from downward momentum transport by vertical mixing can be ruled out.At 0500 LST (Fig. 14d), the convective cell becomes fully developed and will later merge into the frontal convective system (cf. Figs. 12c-f). The northward advance of the cold air produced by downdrafts from the convection over the southern Taiwan Strait in the model can be traced through time. As shown in Fig. 16, the surface gust front as identified based on potential temperature perturbation (?') [i.e., at the leading edge of the cold air with strongest horizontal ?' gradient (??')], propagated generally northeastward over the period of 0000-0500 LST 8 June 2007 at a speed of 5-8 m s-1. Here, the perturbation is defined as the departure from the 24-h mean of 0800 LST 7-8 June. Along line AA' from south-southwest (SSW) to northnortheast (NNE) and about 200 km in length, vertical cross sections are constructed for the lower troposphere in Fig. 17 to further examine the structure of the cold air and related features. From 0100 to 0300 LST, the surface- based cold air (marked by ''C'') just behind the gust front (mark by arrows) can be seen to be driven by downdrafts and associated with horizontal divergence, although its strength gradually weakens (in terms of ?') as it travels forward along the section at about 7 m s-1 (Figs. 17a-c,e-g). The downdraft-driven cold air is also accompanied by positive pressure perturbation ( p', up to 0.4 hPa) as the ridge labeled by ''R3,'' although ridges of p' are also found to coincide with cold centers above 1 km as a hydrostatic response (e.g., ''R1'' and ''R2,'' Figs. 17a-c). Using model outputs every 10 min, these p' perturbations can be traced trough time (as labeled) and they propagate at a speed around 20 m s-1, significantly faster than the gust front. For the one associated with ''R3'' and convectively generated (Figs. 17b-d), in particular, the p' gradient (?p') just ahead of the ridge (marked by open triangles) causes an increase in the forward-directed (horizontal) pressure gradient force (PGF). This leads to a strengthening in the near-surface southwesterly wind in the proximity, over a depth of several hundred meters. With significant pressure and wind changes (also cf. Figs. 6 and 15), these p' perturbations exhibit characteristics of gravity waves with speeds comparable to those found in earlier studies (e.g., Bosart and Cussen 1973; Bosart and Seimon 1988; Carbone et al. 1990; Karyampudi et al. 1995). In Figs. 17e-g, surfacebased convergence appears ahead of the gust front in response to its forward movement up to 0300 LST. On the other hand, both the convergence and the associated ascent grow much stronger and deeper near 118.7°E at 0400 LST, and indicate the breakout of deep convection about 70 km ahead of the (weakening) gust front (Fig. 17h) as the gravity wave arrives and the surface wind intensifies more rapidly (cf. Fig. 15). Thus, the convection was triggered remotely, ahead of the cold outflow produced by previous convection to the south. The Hovmoller (distance-time) plot along section AA' (Fig. 18) shows the distributions of equivalent potential temperature ?e at 170 m(the second model level) and the wind speed along the section at 522 m (the fourth level), with locations of the gust front and maximum ?p' (i.e., open triangles in Fig. 17) also marked up to every 10 min. In Fig. 18, multiple cold centers with lower ?e values (typically by about 3 K) appear near 117.9°E (and south of 23°N) roughly every 2-4 h in a pulselike fashion caused by the repeated passages of storms over the southern strait (cf. Fig. 12). Indicated by the slopes of their axes (white dotted lines), these surges of cold air toward the NNE are all at speeds close to the gust front (~7 m s-1, marked by ''x'' and the black dotted line), while the ?e contrast gradually diminishes to beyond recognition before reaching 118.5°E (Figs. 18 and 17). On the other hand, the disturbance in p' associated with the gravity wave (open and closed white dots) travels at the speed ranges of 10-25 m s-1 (with a mean near 19 m s-1), comparable to those found in earlier studies (e.g., Bosart and Seimon 1988; Carbone et al. 1990; Monserrat and Thorpe 1992; Karyampudi et al. 1995; Shige 1999), and is accompanied by an increase in local wind speed since 0200 LST (Fig. 18). Again, as it reaches 118.73°E near 0400 LST, near-surface winds intensify rapidly and deep convection (black triangle) is triggered (also Figs. 12, 15, and 17). c. Additional sensitivity test In the previous subsection, the gravity wave ahead of the cold outflow from the convection over the southern Taiwan Strait and the associated wind speed increase are suggested to be vital for the initiation of new convection at the correct location and time (i.e., near 24°N, 119°E at about 0400 LST 8 June 2007). During the course of this study, other model runs that start at either 12 h earlier (0800 LST 6 June) or later (0800 LST 7 June) were also performed. Although little differences existed in subsynoptic-scale features (such as the LLJ and 500-hPa trough) and convection near the front was simulated, these two runs did not reproduce the convection over the central strait at the correct timing. Thus, they can serve as additional tests. Specifically, if the convection over the southern strait was not captured, the subsequent convective initiation in the central strait at the right time does not occur. Thus, it is confirmed that the earlier and repeated convection to the south was not only vital but also necessary for the convection over the central strait, and thus the correct simulation of rainfall distribution over Taiwan (cf. Figs. 7 and 13), at least in the model. It should be noted that vertical cross sections similar to Fig. 17 but in other directions from the spot of convective initiation in central Taiwan Strait, for instance, toward the west (into China) and north (into the mei-yu front) were also constructed from the 2-km run, but none of them revealed propagating features near the surface that may trigger the deep convection (not shown). For the one toward the west, this result is not surprising since the model did not capture the weakening N-S convective lines (i.e., line D in Fig. 8a; section 5a). Based on model results, their presence is not required, and thus their potential roles were most likely of secondary importance for the convective initiation over the central strait. From the observation, the area of the strongest redevelopment also exhibited an E-W orientation (Fig. 8c, marked by an arrow), as in the model (Fig. 14c), in accordance to the advancing gust front and gravity wave disturbance from the south (Figs. 16-18). 6. Discussion In the previous section, it is found that the convectively generated cold air from the southern Taiwan Strait propagates at about 7 m s-1 toward the NNE. Based on Liu and Moncrieff (1996), the propagation speed of the density current C can be estimated as ... (1) where u0 is the background wind speed, Du is the ? difference across the cold air boundary, g is the gravitational acceleration (9.81 m s-2), and h is the height of the cold current. Britter and Simpson (1978) and Thorpe et al. (1980) also obtained similar formulae. Using values estimated from Figs. 17 and 18 (??/? [asymptotically =] 0.01 at most and h [asymptotically =] 500 m), the first term on the right-hand side of Eq. (1) can be determined to be about 7 m s-1, in agreement with Figs. 16-18 and previous studies in the tropics (e.g., Lac et al. 2002) and midlatitudes (e.g., Fulton et al. 1990; Rauber et al. 2001) under various background wind speed. From Fig. 17, similarly, the ?p' in the horizontal near the surface can be estimated to be about 30 Pa over 50 km on average, and corresponds to an acceleration of 0.5 × 10-3 m s-2 (using r [asymptotically =] 1.2 kg m-3) as a result of the enhanced PGF. Thus, the wind speed induced by the gravity wave can increase by over 5 m s-1 in 3 h, which is comparable to model results (Figs. 14, 17, and 18). In the present case, cold outflows were repeatedly produced by the convective storms in the southern Taiwan Strait (cf. Figs. 8 and 12) at least since 1800 LST 7 June (not shown), and appear as pulses in Figs. 17 and 18. Such a persistent source is suggested to be vital by Simpson (1987) and Koch et al. (1991) to feed the cold air continuously toward the head of density current from rear to front (in a system-relative sense). One phenomenon that bares some similarity to the triggering of convection ahead of an advancing density current (e.g., Simpson et al. 1980) is the effect of sea-breeze front (SBF), which in Florida has been studied extensively (e.g., Pielke 1974; Wakimoto and Atkins 1994; Kingsmill 1995). In several studies, deep convection (often associated with roll clouds) can be initiated at a short distance (within about 20 km) ahead of the advancing SBF (e.g., Nicholls et al. 1991; Fovell and Dailey 2001), or before the collision with an outflow boundary (e.g., Fankhauser et al. 1995). On the other hand, it is possible for new convection to be triggered by a single propagating gravity current (i.e., a bore), produced by earlier convection, at longer distances in favorable environments (Carbone et al. 1990; Karyampudi et al. 1995; Shige 1999). In these studies (also Bosart and Seimon 1988), gravity waves are found to be excited also at the back side of previous convection with comparable speeds of about 14-18 m s-1. However, there is little evidence of a surface-based wave duct (i.e., an inversion or stable layer) in our case (cf. Fig. 5) and the atmosphere, with low static stability and high moisture content (and Table 2), must be prone to energy dispersion and thus unfavorable for the maintenance of gravity waves (e.g., Lindzen and Tung 1976; Uccellini and Koch 1987; Crook 1988; Koch and Siedlarz 1999). Note that the p' disturbances in Fig. 17 weaken and evolve rapidly with time near the surface, in agreement with this assessment. Under such conditions, the persistence of active convection over the southern strait for long hours is presumably also important to excite gravity waves repeatedly (such as R1 and R2 in Fig. 17). Thus, the surface wind speed upstream from 24°N, 119°E can increase gradually for several hours in the model, before the arrival of the particular wave identified in Figs. 17 and 18 near 0400 LST 8 June (also cf. Figs. 14 and 15). Over the surrounding ocean near Taiwan, the detailed mechanism for remote triggering of convection by gravity wave ahead of the gust front during the mei-yu season, as seen in the present study, has not been documented previously. 7. Conclusions In this study, a numerical investigation on the heavy rainfall event over central Taiwan on 8 June 2007 during the SoWMEX/TiMREX pilot experiment is presented. Prior to the rain (overnight of 7-8 June), objective and manual analyzed weather maps and radar and satellite observations indicate two preexisting lines of forcing roughly 250 km apart: one along the ENE-WSWoriented surface mei-yu front in the northern Taiwan Strait and the other associated with a series of deep convection over southern strait with a WNW-ESE alignment. The convection responsible for the heavy rainfall, however, developed between these two lines in central Taiwan Strait near 0400 LST 8 June. Thus, we were motivated to investigate on the detailed mechanism leading to its initiation. By using the cloud-resolving CReSS model (Tsuboki and Sakakibara 2007) at a horizontal grid size of 2 km, the present heavy rainfall event is reasonably well reproduced. Both the preexisting lines of convection and the initiation of new cells over the central strait just prior to 0400 LST, as well as their subsequent evolution, are successfully captured. With general agreement with observations, the model results are further used to investigate the triggering mechanism for the convection in central Taiwan Strait. The major findings, including both those from modeling and observation, are summarized below. 1) The favorable synoptic conditions supporting the MCSs and deep convection in the present case include the approaching of the mei-yu front, a strengthening LLJ from the SSW to the west at 850 and 700 hPa to provide westerly vertical wind shear, and a diffluent flow pattern in the upper troposphere. 2) The thermodynamic conditions over the central Taiwan Strait prior to the convection indicate conditional instability, with appreciableCAPE of at least 830 m2 s-2, small CIN, and a height of LFC near 1.5 km (for mixed air parcels). Thus, deep convection can be achieved through sufficient forcing, as indeed observed in the present case. 3) The convection over the southern Taiwan Strait during the case period originated near the coast of China. This series of storms, while moving eastward, repeatedly produced cold outflow that propagated toward the NNE at about 7 m s-1 as well as positive pressure perturbations (p') that traveled at much higher speed of about 19 m s-1. With characteristics of gravity wave, the p' disturbance caused enhanced (horizontal) PGF and an intensification of southerly winds (by up to 4-5 m s-1 h-1). Eventually, the deep convection was triggered over central Taiwan Strait, roughly 70 km ahead of the gust front, near 0400 LST 8 June 2007 upon the arrival of the gravity wave disturbance. Consequently, heavy rainfall is resulted. 4) The persistent convection for long hours over the southern Taiwan Strait is helpful for the maintenance of the gust front (density current) and presumably important for the repeated generation of gravity waves, since little evidence exists for a low-level wave duct and the conditions are prone to energy dispersion in our case. On the other hand, the propagation speed of the gravity wave is comparable to those found in earlier studies. 5) In sensitivity tests starting at a different initial time, if there is no storm over the southern Taiwan Strait, no new convection is triggered in the central strait at the correct time consequently. This provides additional support that the convection responsible for the heavy rainfall was indeed linked to existing storms over the southern Taiwan Strait. In this study, the detailed mechanism for a remote trigger of new convection over the Taiwan Strait, by gravity waves associated with cold outflows from existing storms to the south, for a heavy rainfall case during the mei-yu season is illustrated, mainly through high-resolution numerical simulation. It is believed that similar processes are not uncommon during periods of active convection over the ocean surrounding Taiwan, but was not uncovered before as a result of a general lack of observation. Likewise, the evolution of convection during this event can be reproduced using cloud-resolving model under the influence of the mei-yu front and high topography, but operational models in general do not have the resolution needed to capture such events. Nevertheless, with better understanding about how deep convection is triggered and how MCSs behave, the present study can contribute toward the improvement of QPFs at short ranges and, ultimately, the reduction of losses to the society caused by heavy rainfall. Acknowledgments. The comments and suggestions from the two reviewers and Dr. Mark I.-M. Wang are greatly appreciated. The radar composite data were provided by the National Science and Technology Center forDisaster Reduction (NCDR) of Taiwan, and this study was supported by the National Science Council of Taiwan under Grants NSC-99-2111-M-003- 004-MY3, NSC-100-2625-M-003-002, NSC-100-2119- M-002-010, and NSC-100-2625-M-002-001. REFERENCES Akaeda, K., J. Reisner, and D. Parsons, 1995: The role of mesoscale and topographically induced circulations initiating a flash flood observed during the TAMEX project. Mon. Wea. Rev., 123, 1720-1739. Asselin, R., 1972: Frequency filter for time integrations. Mon. Wea. Rev., 100, 487-490. Baines, P.G.,1995: Topographic Effects in Stratified Flows.Cambridge University Press, 482 pp. Banta, R. M., 1990: The role of mountain flows in making clouds. Atmospheric Processes over Complex Terrain, Meteor. Monogr., No. 45, Amer. Meteor. Soc., 229-284. Bluestein, H. B., and M. L. Weisman, 2000: The interaction of numerically simulated supercells initiated along lines. Mon. Wea. Rev., 128, 3128-3149. Bosart, L. F., and J. P. Cussen Jr., 1973: Gravity wave phenomena accompanying East Coast cyclogenesis. Mon. Wea. Rev., 101, 446-454. _____, and A. Seimon, 1988: A case study of an unusually intense atmospheric gravity wave. Mon. Wea. Rev., 116, 1857-1886. Britter, R. E., and J. E. Simpson, 1978: Experiments on the dynamics of a gravity current head. J. Fluid Mech., 88, 223-240. Carbone, R. E., J. W. Conway, N. A. Crook, and M. W. Moncrieff, 1990: The generation and propagation of a nocturnal squall line. Part I: Observations and implications for mesoscale predictability. Mon. Wea. Rev., 118, 25-49. Chen, G. T.-J., 1992: Mesoscale features observed in Taiwan Mei- Yu season. J. Meteor. Soc. Japan, 70, 497-516. _____, and C.-C. Yu, 1988: Study of low-level jet and extremely heavy rainfall over northern Taiwan in the mei-yu season. Mon. Wea. Rev., 116, 884-891. _____, C.-C. Wang, and D. T.-W. Lin, 2005: Characteristics of lowlevel jets over Northern Taiwan in mei-yu season and their relationship to heavy rain events. Mon. Wea. Rev., 133, 20-43. _____, _____, and S.-W. Chang, 2008: A diagnostic case study of meiyu frontogenesis and development of wavelike frontal disturbances in the subtropical environment. Mon. Wea. Rev., 136, 41-61. Chen, T.-C., M.-C. Yen, J.-C. Hsieh, and R. W. Arritt, 1999: Diurnal and seasonal variations of the rainfall measured by the automatic rainfall and meteorological telemetry system in Taiwan. Bull. Amer. Meteor. Soc., 80, 2299-2312. Chen, Y.-L., 1993: Some synoptic-scale aspects of the surface fronts over southern China during TAMEX. Mon. Wea. Rev., 121, 50-64. _____, and J. Li, 1995: Large-scale conditions favorable for the development of heavy rainfall during TAMEX. Mon. Wea. Rev., 123, 2978-3002. Chien, F.-C., and B. J.-D. Jou, 2004: MM5 emsemble mean precipitation in the Taiwan area for three early summer convective (mei-yu) seasons. Wea. Forecasting, 19, 735-750. Cotton, W. R., G. J. Tripoli, R. M. Rauber, and E. A. Mulvihill, 1986: Numerical simulation of the effects of varying ice crystal nucleation rates and aggregation processes on orographic snowfall. J. Climate Appl. Meteor., 25, 1658-1680. Crook, N. A., 1988: Trapping of low-level internal gravity waves. J. Atmos. Sci., 45, 1533-1541. Durran, D. R., 1986: Mountain waves. Mesoscale Meteorology and Forecasting, P. S. Ray, Ed., Amer. Meteor. Soc., 472-492. Einaudi, F., A. J. Bedard, and J. J. Finnigan, 1989: A climatology of gravity waves and other coherent disturbances at the Boulder Atmospheric Observatory during March-April 1984. J. Atmos. Sci., 46, 303-329. Fankhauser, J.C., N. A. Crook, J. Tuttle, L. J. Miller, andC.G.Wade, 1995: Initiation of deep convection along boundary layer convergence lines in a semitropical environment. Mon. Wea. Rev., 123, 291-314. Fovell, R. G., and P. S. Dailey, 2001: Numerical simulation of the interaction between the sea-breeze front and horizontal convective rolls. Part II: Alongshore ambient flow. Mon. Wea. Rev., 129, 2057-2072. Fulton, R., D. S. Zrnic, and R. J. Doviak, 1990: Initiation of a solitary wave family in the demise of a nocturnal thunderstorm density current. J. Atmos. Sci., 47, 319-337. Georgelin, M., E. Richard, and M. Petididier, 1996: The impact of diurnal cycle on a low-Froude number flow observed during the PYREX experiment. Mon. Wea. Rev., 124, 1119- 1131. Ikawa, M., and K. Saito, 1991: Description of a nonhydrostatic model developed at the Forecast Research Department of the MRI. MRI Tech. Rep. 28, 238 pp. Jou, B. J.-D., and W.-C. Lee, 2008: An overview of SoWMEX/ TiMREX operation. Proc. First SoWMEX/TiMREX Science Workshop, Taipei, Taiwan, Taiwan Weather Research Program Project Office, 12-13. Karyampudi, V. M., S. E. Koch, C. Chen, J. W. Rottman, and M. L. Kaplan, 1995: The influence of the Rocky Mountains on the 13-14 April 1986 severe weather outbreak. Part II: Evolution of a prefrontal bore and its role in triggering a squall line. Mon. Wea. Rev., 123, 1423-1446. Kingsmill, D. E., 1995: Convection initiation associated with a seabreeze front, a gust front, and their collision. Mon. Wea. Rev., 123, 2913-2933. _____, and N. A. Crook, 2003: An observational study of atmospheric bore formation from colliding density currents. Mon. Wea. Rev., 131, 2985-3002. Klemp, J. B., and R. B. Wilhelmson, 1978: The simulation of threedimensional convective storm dynamics. J. Atmos. Sci., 35, 1070-1096. Koch, S. E., and L. M. Siedlarz, 1999: Mesoscale gravity waves and their environment in the central United States during STORM-FEST. Mon. Wea. Rev., 127, 2854-2879. _____, P. B. Dorian, R. Ferrare, S. H. Melfi, W. C. Skillman, and D. Whiteman, 1991: Structure of an internal bore and dissipating gravity current as revealed by Raman lidar. Mon. Wea. Rev., 119, 857-887. Kondo, J., 1976: Heat balance of the China Sea during the air mass transformation experiment. J. Meteor. Soc. Japan, 54, 382-398. Koppel, L. L., L. F. Bosart, and D. Keyser, 2000: A 25-yr climatology of large-amplitude hourly surface pressure changes over the conterminous United States. Mon. Wea. Rev., 128, 51-68. Kuo, Y.-H., and G. T.-J. Chen, 1990: The Taiwan Area Mesoscale Experiment (TAMEX): An overview. Bull. Amer. Meteor. Soc., 71, 488-503. Lac, C., J.-P. Lafore, and J.-L. Redelsperger, 2002: Role of gravity waves in triggering deep convection during TOGA COARE. J. Atmos. Sci., 59, 1293-1316. Lee, B. D., B. F. Jewett, and R. B. Wilhelmson, 2006: The 19 April 1996 Illinois tornado outbreak. Part I: Cell evolution and supercell isolation. Wea. Forecasting, 21, 433-448. LeMone, M. A., E. J. Zipser, and S. B. Trier, 1998: The role of environmental shear and thermodynamic conditions in determining the structure and evolution of mesoscale convective systems during TOGA COARE. J. Atmos. Sci., 55, 3493- 3518. Li, J., and Y.-L. Chen, 1998: Barrier jets during TAMEX. Mon. Wea. Rev., 126, 959-971. _____, _____, and W.-C. Lee, 1997: Analysis of a heavy rainfall event during TAMEX. Mon. Wea. Rev., 125, 1060-1081. Lin, Y.-L., 1993: Orographic effects on airflow and mesoscale weather systems over Taiwan. Terr. Atmos. Oceanic Sci., 4, 381-420. _____, R. D. Farley, and H. D. Orville, 1983: Bulk parameterization of the snow field in a cloud model. J. Climate Appl. Meteor., 22, 1065-1092. _____, N.-H. Lin, and R. P. Weglarz, 1992: Numerical modeling studies of lee mesolows, mesovortices and mesocyclones with application to the formation of Taiwan mesolows. Meteor. Atmos. Phys., 49, 43-67. Lindzen, R. S., and K. K. Tung, 1976: Banded convective activity and ducted gravity waves. Mon. Wea. Rev., 104, 1602-1617. _____, and M. Fox-Rabinovitz, 1989: Consistent vertical and horizontal resolution. Mon. Wea. Rev., 117, 2575-2583. Liu, A. Q., G. W. K. Moore, K. Tsuboki, and I. A. Renfrew, 2004:A high-resolution simulation of convective roll clouds during a cold-air outbreak. Geophys. Res. Lett., 31, L03101, doi:10.1029/ 2003GL018530. Liu, C., and M. W. Moncrieff, 1996: A numerical study of the effects of ambient flow and shear on density currents. Mon. Wea. Rev., 124, 2282-2303. Louis, J. F., M. Tiedtke, and J. F. Geleyn, 1981: A short history of the operational PBL parameterization at ECMWF. Proc. Workshop on Planetary Boundary Layer Parameterization, Reading, United Kingdom, ECMWF, 59-79. Mellor, G. L., and T. Yamada, 1974: A hierarchy of turbulent closure models for planetary boundary layers. J. Atmos. Sci., 31, 1791-1806. Monserrat, S., and A. J. Thorpe, 1992: Gravity-wave observations using an array of microbarographs in the Balearic Islands. Quart. J. Roy. Meteor. Soc., 118, 259-282. Murakami, M., 1990: Numerical modeling of dynamical and microphysical evolution of an isolated convective cloud-The 19 July 1981 CCOPE cloud. J. Meteor. Soc. Japan, 68, 107- 128. _____, T. L. Clark, and W. D. Hall, 1994: Numerical simulations of convective snow clouds over the Sea of Japan: Twodimensional simulation of mixed layer development and convective snow cloud formation. J.Meteor. Soc. Japan, 72, 43-62. Nicholls, M. E., R. A. Pielke, and W. R. Cotton, 1991: A twodimensional numerical investigation of the interaction between sea breezes and deep convection over the Florida peninsula. Mon. Wea. Rev., 119, 298-323. Persson, P. O. G., and T. T. Warner, 1991: Model generation of spurious gravity waves due to inconsistency of the vertical and horizontal resolution. Mon. Wea. Rev., 119, 917-935. Pielke, R. A., 1974: A three-dimensional numerical model of the sea breezes over south Florida. Mon. Wea. Rev., 102, 115- 139. Powers, J. G., 1997: Numerical model simulations of a mesoscale gravity wave event: Sensitivity tests and spectral analysis. Mon. Wea. Rev., 125, 1838-1869. Purdom, J. F. W., 1976: Some uses of high resolution GOES imagery in the mesoscale forecasting of convection and its behavior. Mon. Wea. Rev., 104, 1474-1483. Rauber, R. M., M. Yang, M. K. Ramamurthy, and B. F. Jewett, 2001: Origin, evolution, and finescale structure of the St. Valentine's Day mesoscale gravity wave observed during STORM-FEST. Part I: Origin and evolution. Mon. Wea. Rev., 129, 198-217. Reynolds, R. W., N. A. Rayner, T. M. Smith, D. C. Stokes, and W. Wang, 2002: An improved in situ and satellite SST analysis for climate. J. Climate, 15, 1609-1625. Rotunno, R., J. B. Klemp, and M. L. Weisman, 1988: A theory for strong, long-lived squall lines. J. Atmos. Sci., 45, 463-485. Schar, C., and R. B. Smith, 1993: Shallow-water flow past isolated topography. Part I: Vorticity production and wake formation. J. Atmos. Sci., 50, 1373-1400. Scofield, R. A., and J. F. W. Purdom, 1986: The use of satellite data for mesoscale analysis and forecasting applications. Mesoscale Meteorology and Forecasting, P. S. Ray, Ed., Amer. Meteor. Soc., 118-150. Segami, A., K. Kurihara, H. Nakamura,M. Ueno, I. Takano, and Y. Tatsumi, 1989: Operational mesoscale weather prediction with Japan Spectral Model. J. Meteor. Soc. Japan, 67, 907-924. Shige, S., 1999: Disturbances of 1-2 hour periods observed in the tropical lower troposphere during the TOGA-COARE IOP. J. Meteor. Soc. Japan, 77, 1123-1136. Simpson, J. E., 1987: Gravity Currents: In the Environment and the Laboratory. John Wiley and Sons, 244 pp. _____, N. E. Westcott, R. J. Clerman, and R. A. Pielke, 1980: On cumulus mergers. Arch. Meteor. Geophys. Bioklimatol. Ser. A, 29, 1-40. Smolarkiewicz, P. K., and R. Rotunno, 1989: Low Froude number flow past three-dimensional obstacles. Part I: Baroclinically generated lee vortices. J. Atmos. Sci., 46, 1154-1164. Stobie, J. G., F. Einaudi, and L. W. Uccellini, 1983: A case study of gravity wave-convective storm interaction: 9 May 1979. J. Atmos. Sci., 40, 2804-2830. Thompson, R. L., R. Edwards, J. A. Hart, K. L. Elmore, and P. Markowski, 2003: Close proximity soundings within supercell environments obtained from the Rapid Update Cycle. Wea. Forecasting, 18, 1243-1261. Thorpe, A. J., M. J. Miller, and M. W. Moncrieff, 1980: Dynamical models of two-dimensional downdraughts. Quart. J. Roy. Meteor. Soc., 106, 463-484. Tsuboki, K., and A. Sakakibara, 2002: Large-scale parallel computing of cloud resolving storm simulator. High Performance Computing, H. P. Zima et al., Eds., Springer, 243-259. _____, and _____, 2007: Numerical prediction of high-impact weather systems. Textbook for the 17th IHP training course in 2007, HyARC, Nagoya University and UNESCO, 273 pp. Uccellini, L. W., 1990: Processes contributing to the rapid development of extratropical cyclones. Extratropical Cyclones: The Erik Palmén Memorial Volume, C. W. Newton and E. O. Holopainen, Eds., Amer. Meteor. Soc., 81-105. _____, and S. E. Koch, 1987: The synoptic setting and possible energy sources for mesoscale gravity disturbances. Mon. Wea. Rev., 115, 721-729. Wakimoto, R.M., and N. T. Atkins, 1994: Observations of the seabreeze front during CaPE. Part I: Single-Doppler, satellite, and cloud photogrammetry analysis. Mon. Wea. Rev., 122, 1092-1114. Wang, C.-C., and W.-M. Huang, 2009: High-resolution simulation of a nocturnal narrow convective line off the southeastern coast of Taiwan in the Mei-yu season. Geophys. Res. Lett., 36, L06815, doi:10.1029/2008GL037147. _____,G. T.-J. Chen, T.-C. Chen, and K. Tsuboki, 2005:Anumerical study on the effects of Taiwan topography on a convective line during the mei-yu season. Mon. Wea. Rev., 133, 3217-3242. _____, _____, S.-C. Yang, and K. Tsuboki, 2009: Wintertime supercell thunderstorms in a subtropical environment: Numerical simulation. Mon. Wea. Rev., 137, 2175-2202. Weisman, M. L., and J. B. Klemp, 1986: Characteristics of isolated convective storms. Mesoscale Meteorology and Forecasting, P. S. Ray, Ed., Amer. Meteor. Soc., 331-358. Wilson, J. W., and W. E. Schreiber, 1986: Initiation of convective storms at radar-observed boundary-layer convergence lines. Mon. Wea. Rev., 114, 2516-2536. Yeh, H.-C., and Y.-L. Chen, 1998: Characteristics of rainfall distributions over Taiwan during the Taiwan Area Mesoscale Experiment (TAMEX). J. Appl. Meteor., 37, 1457-1469. Zhang, F., and S. E. Koch, 2000: Numerical simulation of a gravity wave event over CCOPE. Part II: Waves generated by an orographic density current. Mon. Wea. Rev., 128, 2777-2796. CHUNG-CHIEH WANG Department of Earth Sciences, National Taiwan Normal University, Taipei, Taiwan GEORGE TAI-JEN CHEN AND SHIN-YI HUANG Department of Atmospheric Sciences, National Taiwan University, Taipei, Taiwan (Manuscript received 31 August 2010, in final form 23 March 2011) Corresponding author address: Prof. George Tai-Jen Chen, Department of Atmospheric Sciences, National Taiwan University, No. 61, Ln. 144, Sec. 4, Keelung Rd., Taipei 10772, Taiwan. E-mail: [email protected] (c) 2011 American Meteorological Society |