1. Introduction
One of the primary findings of the Taiwan Area Mesoscale Experiment (TAMEX) conducted jointly by scientists from Taiwan and the United States during May–June 1987 (Kuo and Chen 1990) was that during the early summer monsoon rainy season over Taiwan, precipitation and airflow are significantly modulated by the diurnal heating cycle (Johnson and Bresch 1991; Yeh and Chen 1998; and others). Along the western and southwestern windward coasts, the hourly rainfall frequencies have a weak early morning maximum under the southwesterly monsoon flow (Kerns et al. 2010). Chen and Li (1995) show that under the southwesterly monsoon flow during TAMEX, island blocking (Li and Chen 1998) exhibits diurnal variations that are most significant before sunrise. Linear convective lines along the land-breeze front off the northwestern coast were also observed from radar data recorded in the early morning during the TAMEX intensive observing period (IOP) 13 (Li et al. 1997).
Most of the heavy rainfall events over southwestern Taiwan originated from preexisting convective showers or mesoscale convective systems (MCSs) that drifted inland and interacted with the terrain and local winds (Chen et al. 2005, 2007, 2008). The weak early morning rainfall maximum along the western and southwestern windward coasts occurs only under the southwesterly monsoon flow during the warm season and is likely caused by the convergence between the offshore flow and incoming, decelerating southwesterly flow when the land surface is the coldest (Kerns et al. 2010). However, occurrences of coastal rainfall over southwestern Taiwan are not regular daily occurrences (Kerns et al. 2010).
Both the hourly rainfall frequencies during TAMEX (Yeh and Chen 1998) and the climatologically heavy rainfall occurrences during the early summer monsoon season (Chen et al. 2007) have a profound afternoon maximum on the southwestern slopes. Johnson and Bresch (1991) showed that under weak large-scale forcing with relatively calm winds over Taiwan, the convective showers occur at the 100–500-m level of mountain slopes due to sea breezes advancing inland and being lifted at the foothills of the mountains. Under disturbed conditions, with persistent orographic lifting and anabatic winds, more rainfall occurs on the mountain interior than the lower slopes in the afternoon hours, corresponding to orographic lifting by the steep terrain (Kerns et al. 2010).
Chen (2000) suggested that under favorable large-scale conditions, diurnal and local effects are important for the timing and location of heavy rainfall occurrences. To understand the physical processes leading to the development of localized heavy rainfall, detailed case studies are required. TAMEX (1987) focused on northwestern Taiwan and deployed three C-band Doppler radars over the central and northwestern Taiwan coast (Kuo and Chen 1990). The project also used 85 hourly rain gauges, which were not evenly distributed, and only three surface stations were available along the southwestern coast. After TAMEX, the Central Weather Bureau (CWB) in Taiwan installed the routine Automatic Rainfall and Meteorological Telemetry System, a dense, automatic, hourly-recording, rain gauge and weather station network (Kerns et al. 2010), which was completed in August 1997. An operational Doppler radar network around the island has also been completed (Fig. 1e).
Rainfall accumulation (mm) during (a) 0000–0800 LT 31 May, (b) 0800–1800 LT 31 May, (c) 0000–0800 LT 16 Jun, and (d) 0800–1800 LT 16 Jun. (e) Terrain height for Taiwan (m). Banciao (25°N, 121.43°E), Penghu (23.56°N, 119.63°E), Tainan (Yongkang; 23.04°N, 120.23°E), Kaohisung (22.57°N, 120.31°E), Hengchun (22.01°N, 120.74°E), and South Ship (21.46°N, 118.36°E) are marked as B, P, T, K, H, and S, respectively. NCAR S-Pol radar is located at SPOL. Red circles mark rawinsonde sites. Purple plus signs mark Doppler radar sites.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Rainfall accumulation (mm) during (a) 0000–0800 LT 31 May, (b) 0800–1800 LT 31 May, (c) 0000–0800 LT 16 Jun, and (d) 0800–1800 LT 16 Jun. (e) Terrain height for Taiwan (m). Banciao (25°N, 121.43°E), Penghu (23.56°N, 119.63°E), Tainan (Yongkang; 23.04°N, 120.23°E), Kaohisung (22.57°N, 120.31°E), Hengchun (22.01°N, 120.74°E), and South Ship (21.46°N, 118.36°E) are marked as B, P, T, K, H, and S, respectively. NCAR S-Pol radar is located at SPOL. Red circles mark rawinsonde sites. Purple plus signs mark Doppler radar sites.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Rainfall accumulation (mm) during (a) 0000–0800 LT 31 May, (b) 0800–1800 LT 31 May, (c) 0000–0800 LT 16 Jun, and (d) 0800–1800 LT 16 Jun. (e) Terrain height for Taiwan (m). Banciao (25°N, 121.43°E), Penghu (23.56°N, 119.63°E), Tainan (Yongkang; 23.04°N, 120.23°E), Kaohisung (22.57°N, 120.31°E), Hengchun (22.01°N, 120.74°E), and South Ship (21.46°N, 118.36°E) are marked as B, P, T, K, H, and S, respectively. NCAR S-Pol radar is located at SPOL. Red circles mark rawinsonde sites. Purple plus signs mark Doppler radar sites.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
During 15 May–30 June 2008, the joint U.S.–Taiwan Terrain-Influenced Monsoon Rainfall Experiment (TiMREX) was conducted to study multiple-scale physical processes leading to the development of localized heavy rainfall during the early summer monsoon season, especially over the south–southwestern part of Taiwan. Davis and Lee (2012) suggested that during 13–18 June 2008, the warm, moist low-level jet (LLJ) lifted by quasi-steady shallow frontal boundaries was important for convection initiation and intensification. Xu et al. (2012) viewed the initiation and maintenance of the long-lived heavy-precipitation MCSs upstream of southwestern Taiwan on 16 June through the back-building/quasi-stationary process. The convection developed continuously near the boundary of 1) the warm, moist unstable air mass associated with a LLJ over the upstream ocean and 2) a remnant cold pool generated by prior precipitation and orographic effects over southwestern Taiwan and the adjacent oceans. For this heavy rainfall event, the large-scale temperature gradients associated with the mei-yu front (Chen et al. 1989) were over southern China (Xu et al. 2012).
In this study, we use high-resolution data collected during TiMREX and model results from a mesoscale model to diagnose the mechanisms of two contrasting localized heavy rainfall events along the windward mountain range (31 May during IOP 3) and over the southwestern coast of Taiwan (16 June during IOP 8) (Fig. 1). In particular, we wish to address the following questions. How do the terrain, diurnally driven local winds, and rain evaporative cooling from preexisting convection affect the MCSs embedded in the southwesterly monsoon flow as they drift inland? Why is the localized heavy rainfall concentrated over the southwestern coast of Taiwan with decreasing rainfall inland for the 16 June case, in contrast to the 31 May case? The primary goal of this study is to investigate the main reasons that account for the differences in the timing and location of heavy rainfall for these two cases using a combination of observations and numerical simulations. We first looked at the large-scale settings for the development of these two events. Next we studied the physical processes, including diurnal and local effects, leading to the occurrences of localized heavy rainfall. Additionally, for the 16 June case, we would also like to study the mechanisms for the development of localized afternoon heavy rainfall associated with thunderstorm activity over northern Taiwan (Fig. 1d).
2. Data and methodology
a. Intensive observing periods during TiMREX
Table 1 lists significant weather events during nine TiMREX IOPs. During IOP 2 and IOP 9, the prefrontal southwesterly flow interacted with the terrain and initiated convective rainfall over central western and southwestern Taiwan. MCSs associated with a surface front developed over Taiwan during IOP 1, IOP 4, and IOP 7. A subtropical oceanic mesoscale convective vortex approached southwestern Taiwan during IOP 6 (Lai et al. 2011). After a frontal convective line north of Taiwan passed through northwestern Taiwan, a frontal cyclone moved across CMR during IOP 3. The offshore warm, moist, southwesterly flow was lifted by the shallow frontal boundary (Davis and Lee 2012) or the cool pool (Xu et al. 2012) during IOP 5 and IOP 8.
Significant weather events during TiMREX IOPs.
b. Local observations, analyses, and satellite observations
The Atmospheric Sounding Processing Environment software system was used for quality control (QC) of dropsonde data (Davis and Lee 2012). Humidity, wind, temperature, and geopotential height from nine rawinsonde stations and the South Ship were corrected through a four-stage QC procedure (Ciesielski et al. 2009, 2010). The rawinsonde data were finally interpolated onto uniform 5-hPa intervals with QC flags assigned to each variable. Bad data were identified and set to missing values through the application of both objective QC tests (Loehrer et al. 1996) and visual inspection (Ciesielski et al. 2010). Soundings from Penghu (Station P in Fig. 1e) in the central Taiwan Strait depict airflow and thermodynamic structure at low levels during the passage of a surface front during IOP 3. Time series of surface air temperature over southwestern Taiwan and the South Ship (Stations T, K, H, and S in Fig. 1e) are used to delineate the low-level thermodynamic structure during IOP 8. The temperature at the lowest level of shipboard soundings launched every 6 h is used as the surface air temperature at the South Ship. During 14–16 June, the location of the South Ship was not stationary (Davis and Lee 2012). The surface wind and temperature data from 25 surface weather stations and offshore dropsondes are used to study the island-induced airflow. Time series of 3-h rawinsonde winds, virtual potential temperature, and equivalent potential temperature at Tainan (Station T in Fig. 1e) show the diurnal variations of winds and thermodynamic structure.
Radar reflectivities from four operational Doppler radar stations (Fig. 1e) are used to depict the evolution of convective activity during IOP 3 and IOP 8. The National Center for Atmospheric Research (NCAR)’s S-band polarimetric (S-Pol) radar was deployed at the southwestern Taiwan coast (Tai et al. 2011; Xu et al. 2012) (Fig. 1e). Rainfall accumulation maps were generated from 429 rainfall stations, which include conventional weather stations and the Automatic Rainfall and Meteorological Telemetry System (Kerns et al. 2010). Lightning data were collected from Tai-Power Company of Taiwan using the Lightning Location System, which can detect cloud-to-ground lightning events. The system consists of one Advanced Position Analyzer and six Direction Finders (DF) covering the entire area of Taiwan. Its sensors are the same as those used by the National Lightning Detection Network in the United States. A DF automatically detects more than 90% of all cloud-to-ground lightning with a maximum detectable distance of 200- and 5-km accuracy (Liou and Kar 2010).
c. Regional weather patterns
The World Climate Research Programme and World Weather Research Programme/The Observing System Research and Predictability Experiment conducted a framework of coordinated observing, modeling, and forecasting of organized tropical convection, which is known as the Year of Tropical Convection (YOTC; Moncrieff et al. 2012). In this study, we use the YOTC analysis of winds, geopotential heights, moisture, and thermodynamic fields with a 0.5° × 0.5° horizontal grid spacing interpolated from the original 25-km grids (http://apps.ecmwf.int/datasets/data/yotc_od/) to describe the evolution of weather patterns.
d. Numerical modeling
The Advanced Research Weather Research and Forecasting Model (WRF-ARW, hereafter WRF; Skamarock et al. 2008) with a 27-km grid spacing is used to simulate the evolution of subsynoptic weather patterns and for the computations of quasigeostrophic (QG) forcing in comparison with the simulated vertical motion from the same domain. The simulated subsynoptic weather patterns are compared with the YOTC analysis. Model results simulated in domain 2 (9-km grid spacing) (Fig. 2) from the control run (CTRL) are compared with the results without the latent heat (LH) release run (WOLH) to assess the impact of latent heat release on the deepening of the frontal cyclone (IOP 3) and a prefrontal weak trough (IOP 8). The convection-allowing (CAR) WRF-ARW with a 3-km grid (domain 3) is used as a diagnostic tool to assess the impact of terrain and land surface properties on local circulations and rainfall. The model results are compared with the high-resolution TiMREX data. Clark et al. (2009) show that simulations using CAR models have a better depiction of location and timing of precipitation than simulations using parameterized-convection resolution (PCR) models because smaller scales are better resolved by CAR than PCR. The model uses the terrain-following hydrostatic-pressure vertical coordinate (Laprise 1992). A two-way nesting procedure is used.
The three domains for the WRF simulation with grid spacing of 27, 9, and 3 km, respectively.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
The three domains for the WRF simulation with grid spacing of 27, 9, and 3 km, respectively.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
The three domains for the WRF simulation with grid spacing of 27, 9, and 3 km, respectively.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
There are 45 sigma levels1 from the surface to the 30-hPa level. The Rapid Radiative Transfer Model (Mlawer et al. 1997), Goddard shortwave (Chou and Suarez 1994) schemes, Noah land surface model (Chen and Dudhia 2001), and Yonsei University planetary boundary layer scheme (Hong et al. 2006) are used. The precipitation process in the model is represented by the grid-resolvable Goddard microphysics (Tao and Simpson 1993) and a modified Kain–Fritsch cumulus parameterization scheme (Kain 2004). The cumulus parameterization is not applied in the 3-km domain (domain 3). The precipitation process and the parameterization scheme used in this study are the same as those used by Hsiao et al. (2012). The National Centers for Environmental Prediction Final Analysis (NCEP FNL) data, using a 1° grid spacing, provide the initial and boundary conditions for the model simulations. A 0.5° daily, real-time, global, sea surface temperature analysis developed at the NCEP Marine Modeling and Analysis Branch (Gemmill et al. 2007) is used for the lower boundary conditions over the ocean. The CWB provided the updated land-use data used in the model (J.-S. Hong 2011, personal communication).
For the simulations of the evolution of weather patterns (domain 1) and the study of the effects of latent heat release (domain 2), the model is initialized at 0000 UTC (0800 LT) 30 May for the IOP 3 case, and 0000 UTC (0800 LT) 15 June for the IOP 8 case. For the simulations of regional flow over Taiwan and to assess the effects of local circulation and terrain on rainfall, the CAR model is nested in the domain 2 model. For the IOP 3 case, with significant variations of the large-scale flow during the heavy rainfall episode, the model is initialized at 0000 UTC (0800 LT) 30 May to study the passage of a shallow frontal system the following night, and 0000 UTC (0800 LT) 31 May to study the passage of frontal cyclone above the shallow front during the daytime. For the IOP 8 case, which is under the prefrontal southwesterly–westerly flow regime with remnant cold pool from antecedent rains, the convection-allowing model is initialized at 0000 UTC 15 June. The terrain height is removed in the without terrain (WOT) runs to assess the impact of terrain on local circulations and rainfall patterns.
3. Intensive observation period 3
a. Favorable conditions for the development of heavy precipitation
1) Evolution of subsynoptic weather patterns
From 1200 UTC 29 May to 0600 UTC 30 May 2008, an upper-level trough, originated in the lee side of the Yun-Gue Plateau over southern China and moved eastward over the moisture tongue over southeastern China (Figs. 3b,c). To the north, a midlatitude trough propagated eastward over northern China and reached the eastern China coast (Fig. 3b). A polar front associated with the midlatitude trough was located off the eastern China coast (Figs. 3a,b). A broad region with a northerly flow over the China plain between the polar front and the migratory high that moved along the northeastern periphery of the Tibetan Plateau is evident (Fig. 3a). The northerly flow from the north converged with the warm, moist, southwesterly monsoon flow from the South China Sea with a well-defined frontal boundary over southern China.
(a) The horizontal distribution of 900-hPa geopotential height (gpm), and winds (m s−1, full barb represents 10 m s−1) at 1200 UTC 29 May 2008 and (b) 500-hPa geopotential height (gpm) at 0600 UTC 30 May; (c) total precipitable water (mm) at 0600 UTC 30 May; (d) 700-hPa QG omega (dPa s−1) and geopotential height (gpm) at 0600 UTC 30 May from WRF (domain 1) results; and (e) 900-hPa geopotential height (gpm), and winds (m s−1) at 1200 UTC 30 May 2008 from the YOTC analysis. (f) As in (e), but from WRF (domain 1) results. The 500-hPa trough axis is marked by a dashed line; L marks the frontal cyclone.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
(a) The horizontal distribution of 900-hPa geopotential height (gpm), and winds (m s−1, full barb represents 10 m s−1) at 1200 UTC 29 May 2008 and (b) 500-hPa geopotential height (gpm) at 0600 UTC 30 May; (c) total precipitable water (mm) at 0600 UTC 30 May; (d) 700-hPa QG omega (dPa s−1) and geopotential height (gpm) at 0600 UTC 30 May from WRF (domain 1) results; and (e) 900-hPa geopotential height (gpm), and winds (m s−1) at 1200 UTC 30 May 2008 from the YOTC analysis. (f) As in (e), but from WRF (domain 1) results. The 500-hPa trough axis is marked by a dashed line; L marks the frontal cyclone.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
(a) The horizontal distribution of 900-hPa geopotential height (gpm), and winds (m s−1, full barb represents 10 m s−1) at 1200 UTC 29 May 2008 and (b) 500-hPa geopotential height (gpm) at 0600 UTC 30 May; (c) total precipitable water (mm) at 0600 UTC 30 May; (d) 700-hPa QG omega (dPa s−1) and geopotential height (gpm) at 0600 UTC 30 May from WRF (domain 1) results; and (e) 900-hPa geopotential height (gpm), and winds (m s−1) at 1200 UTC 30 May 2008 from the YOTC analysis. (f) As in (e), but from WRF (domain 1) results. The 500-hPa trough axis is marked by a dashed line; L marks the frontal cyclone.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
The vertical cross section along the frontal zone at 0600 UTC 30 May (Fig. 4) attests that this frontal system exhibited baroclinic characteristics. The postfrontal northeasterly flow converged with the southwesterly monsoon flow over the frontal zone (~25°–27°N; Figs. 4a,b) with a marked vertical tilt and large horizontal thermal gradients at low levels (Figs. 4c,d). From 0600 to 1200 UTC 30 May, as the 500-hPa lee trough over southern China moved toward the moisture tongue along the southeastern China coast (Figs. 3b,c), the low-level frontal cyclone became well developed with larger geopotential height gradients (Fig. 3e). At this time, the geopotential height over southeastern China increased as a result of the southeastward advance of the migratory high behind the midlatitude polar front.
Latitude–height cross section along 119°E (solid line in Fig. 3d) of (a) zonal winds (m s−1), (b) meridional wind (m s−1), (c) potential temperature (K), and (d) equivalent potential temperature (K) at 0600 UTC (1400 LT) 30 May from the YOTC analysis.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Latitude–height cross section along 119°E (solid line in Fig. 3d) of (a) zonal winds (m s−1), (b) meridional wind (m s−1), (c) potential temperature (K), and (d) equivalent potential temperature (K) at 0600 UTC (1400 LT) 30 May from the YOTC analysis.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Latitude–height cross section along 119°E (solid line in Fig. 3d) of (a) zonal winds (m s−1), (b) meridional wind (m s−1), (c) potential temperature (K), and (d) equivalent potential temperature (K) at 0600 UTC (1400 LT) 30 May from the YOTC analysis.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
The evolution of subsynoptic weather patterns is rather typical as compared to those found in the literature (Chen et al. 1989; Chen and Hui 1990, 1992; Trier et al. 1990; and others). The frontal systems over southern China during the early summer rainy season frequently possess baroclinic characteristics with a marked vertical tilt (e.g., Chen et al. 1989, their Fig. 9) with quasigeostrophic frontogenesis along the frontal boundary (Chen 1993, his Fig. 13b). Diagnosis of the energy budget associated with a deep frontal cyclone that occurred during 1–2 June 1987 over southern China (Chen et al. 1994) shows that the deepening of the frontal cyclone and the accompanying strengthening of the subsynoptic low-level jet is related to the moist baroclinic process across the frontal zone (Chen et al. 1997; Chen and Chen 2002). This case is in contrast to a dry event presented by Chen and Hui (1992). For that case, under relatively dry conditions because of the westward extension of the western Pacific subtropical high (WPSH), the cyclone that originated over southwestern China dissipated as it moved toward the southeastern China coast.
2) Mesoscale WRF modeling and diagnosis
At 0600 UTC 30 May, the simulated 700-hPa trough from the 27-km WRF (domain 1) (Fig. 3d) is along the southeastern China coast. QG frontogenesis is diagnosed ahead of the trough axis (not shown). The maximum vertical motion forced by the adiabatic baroclinic process is on the order of −8 dPa s−1 (Fig. 3d). A well defined low-level frontal cyclone is simulated along the southeastern China coast (Fig. 3f) as the 700-hPa trough (Fig. 3d) moves over the moist tongue (Fig. 3c) in agreement with the YOTC analysis (Fig. 3e). The vertical cross section of θ and θe across the front from the WRF (domain 1) simulation (Figs. 5c,d) agrees well with the YOTC analysis (Figs. 4c,d). The simulated prefrontal southwesterly flow and postfrontal northeasterly flow (Figs. 5a,b) are also consistent with the YOTC analysis (Figs. 4a,b). The maximum vertical motion diagnosed from the adiabatic QG forcing (−14 dPa s−1 at the 900-hPa level; Fig. 5f; domain 1) is much smaller than the simulated maximum simulated vertical motion (−127 dPa s−1 at the 600-hPa level) from the same domain because of the feedback effects from latent heat release (Fig. 5e). Our model results from the 9-km WRF (domain 2) show that the latent heat release associated with the convective activity ahead of the trough generates significant potential vorticity (PV) at low levels that leads to a more intense frontal cyclone over the southeast coast of China (Figs. 6a,b).
Model results for the latitude–height cross section along 119°E (Fig. 3d) from the domain 1 model with a 27-km grid valid at 0600 UTC (1400 LT) 30 May: (a) zonal winds (m s−1), (b) meridional wind (m s−1), (c) potential temperature (K), (d) equivalent potential temperature (K), (e) vertical velocity ω (dPa s−1; contoured every 30 dPa s−1) and horizontal wind field (u, υ), and (f) QG omega (dPa s−1; contoured every 2 dPa s−1). WRF is initialized at 0000 UTC 30 May.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Model results for the latitude–height cross section along 119°E (Fig. 3d) from the domain 1 model with a 27-km grid valid at 0600 UTC (1400 LT) 30 May: (a) zonal winds (m s−1), (b) meridional wind (m s−1), (c) potential temperature (K), (d) equivalent potential temperature (K), (e) vertical velocity ω (dPa s−1; contoured every 30 dPa s−1) and horizontal wind field (u, υ), and (f) QG omega (dPa s−1; contoured every 2 dPa s−1). WRF is initialized at 0000 UTC 30 May.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Model results for the latitude–height cross section along 119°E (Fig. 3d) from the domain 1 model with a 27-km grid valid at 0600 UTC (1400 LT) 30 May: (a) zonal winds (m s−1), (b) meridional wind (m s−1), (c) potential temperature (K), (d) equivalent potential temperature (K), (e) vertical velocity ω (dPa s−1; contoured every 30 dPa s−1) and horizontal wind field (u, υ), and (f) QG omega (dPa s−1; contoured every 2 dPa s−1). WRF is initialized at 0000 UTC 30 May.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Model results at 0600 UTC (1400 LT) 30 May from the domain 2 model with a 9-km grid: (a) horizontal distributions of 900-hPa PV [potential vorticity units (PVU, 1 PVU = 10−6 K m2 kg−1 s−1)], geopotential height (gpm, contoured), and winds (full barb represents 10 m s−1) in the CTRL; (b) As in (a), but for the WOLH run. The model is initialized at 0000 UTC 30 May; L and Lt mark the location of frontal cyclone and leeside vortex off the southeastern Taiwan coast, respectively.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Model results at 0600 UTC (1400 LT) 30 May from the domain 2 model with a 9-km grid: (a) horizontal distributions of 900-hPa PV [potential vorticity units (PVU, 1 PVU = 10−6 K m2 kg−1 s−1)], geopotential height (gpm, contoured), and winds (full barb represents 10 m s−1) in the CTRL; (b) As in (a), but for the WOLH run. The model is initialized at 0000 UTC 30 May; L and Lt mark the location of frontal cyclone and leeside vortex off the southeastern Taiwan coast, respectively.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Model results at 0600 UTC (1400 LT) 30 May from the domain 2 model with a 9-km grid: (a) horizontal distributions of 900-hPa PV [potential vorticity units (PVU, 1 PVU = 10−6 K m2 kg−1 s−1)], geopotential height (gpm, contoured), and winds (full barb represents 10 m s−1) in the CTRL; (b) As in (a), but for the WOLH run. The model is initialized at 0000 UTC 30 May; L and Lt mark the location of frontal cyclone and leeside vortex off the southeastern Taiwan coast, respectively.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
b. Analysis over the Taiwan area
In this section we present the island-scale airflow and weather during the passage of the shallow cold front over Taiwan at night followed by the passage of a frontal cyclone above the shallow front during the daytime of 31 May. Our focus is on the impacts of diurnal and local effects on the development of localized heavy rainfall.
1) Nocturnal flow regime during the frontal passage
On 30 May, the surface front moved southeastward over the Taiwan area (Fig. 7a). At 1800 UTC 30 May (0200 LT 31 May) 2008, the surface front was cut by the island terrain into two branches with the eastern branch east of Taiwan propagated faster than the western counterpart (Fig. 7a). The observed (Fig. 7b) and YOTC winds (Fig. 7a) over the island are relatively calm with a very weak downslope wind component over land as a result of nocturnal cooling. The convection over the island is suppressed (Fig. 8a). The low-level westerly/southwesterly flow ahead of the surface front over southwestern Taiwan was deflected by the CMR with splitting airflow over the southwestern Taiwan coast (Fig. 7a). The deflected southerly flow along the western coast converged with the postfrontal northeasterly flow (Fig. 7a) with enhanced convection there (Fig. 8a).
(a) Horizontal distributions of the 2-m temperature (°C) and surface winds (full barb represents 5 m s−1) at 1800 UTC 30 May (0200 LT 31 May) from the YOTC analysis; L and Lt mark the location of frontal cyclone and leeside vortex off the southeastern Taiwan coast, respectively. (b) YOTC analysis of SST (°C) (solid line) and observed surface winds and temperature (°C) from surface stations at 1800 UTC 30 May (0200 LT 31 May). (c) Soundings at Tainan (Yongkang, Station T in Fig. 1e) at 2100 UTC 30 May (0500 LT 31 May) (red) and 2100 UTC 15 Jun (0500 LT 16 Jun) (full barb represents 5 m s−1).
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
(a) Horizontal distributions of the 2-m temperature (°C) and surface winds (full barb represents 5 m s−1) at 1800 UTC 30 May (0200 LT 31 May) from the YOTC analysis; L and Lt mark the location of frontal cyclone and leeside vortex off the southeastern Taiwan coast, respectively. (b) YOTC analysis of SST (°C) (solid line) and observed surface winds and temperature (°C) from surface stations at 1800 UTC 30 May (0200 LT 31 May). (c) Soundings at Tainan (Yongkang, Station T in Fig. 1e) at 2100 UTC 30 May (0500 LT 31 May) (red) and 2100 UTC 15 Jun (0500 LT 16 Jun) (full barb represents 5 m s−1).
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
(a) Horizontal distributions of the 2-m temperature (°C) and surface winds (full barb represents 5 m s−1) at 1800 UTC 30 May (0200 LT 31 May) from the YOTC analysis; L and Lt mark the location of frontal cyclone and leeside vortex off the southeastern Taiwan coast, respectively. (b) YOTC analysis of SST (°C) (solid line) and observed surface winds and temperature (°C) from surface stations at 1800 UTC 30 May (0200 LT 31 May). (c) Soundings at Tainan (Yongkang, Station T in Fig. 1e) at 2100 UTC 30 May (0500 LT 31 May) (red) and 2100 UTC 15 Jun (0500 LT 16 Jun) (full barb represents 5 m s−1).
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
(a) Mosaic radar reflectivities (dBZ) at 0200 LT 31 May (1800 UTC 30 May) and (b) 1600 LT (0800 UTC) 31 May. (c) The 3-h lightning frequency (km−2) during 0200–0500 LT (arrow points to frontal convection region) and (d) 1300–1600 LT 31 May (arrow points to thunderstorm activity over the mountain interior). (Courtesy of Central Weather Bureau.)
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
(a) Mosaic radar reflectivities (dBZ) at 0200 LT 31 May (1800 UTC 30 May) and (b) 1600 LT (0800 UTC) 31 May. (c) The 3-h lightning frequency (km−2) during 0200–0500 LT (arrow points to frontal convection region) and (d) 1300–1600 LT 31 May (arrow points to thunderstorm activity over the mountain interior). (Courtesy of Central Weather Bureau.)
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
(a) Mosaic radar reflectivities (dBZ) at 0200 LT 31 May (1800 UTC 30 May) and (b) 1600 LT (0800 UTC) 31 May. (c) The 3-h lightning frequency (km−2) during 0200–0500 LT (arrow points to frontal convection region) and (d) 1300–1600 LT 31 May (arrow points to thunderstorm activity over the mountain interior). (Courtesy of Central Weather Bureau.)
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
2) Daytime flow regime
At 0600 UTC (1400 LT) 31 May, the upper-level shortwave trough (Fig. 3b) moved over Taiwan (Fig. 9a). At the 850-hPa level, the Taiwan area is under the influence of a cyclonic flow with the frontal cyclone (L) over western Taiwan in both the YOTC analysis (Fig. 9c) and WRF simulation with a 27-km grid (Fig. 9d). Northern Taiwan was under a cold (~23°C) postfrontal northeasterly flow (Fig. 9e). The cold, dry northeasterlies were deflected by CMR leaving warm (~31°C), moist air over the southwestern leeside plain (Fig. 9e), similar to the TAMEX case reported by Chen et al. (1989). The southwesterly flow prevailed above the shallow front off the western Taiwan coast (Figs. 9b,c).
(a) 500-hPa geopotential height (gpm), (b) surface winds (full barb represents 5 m s−1) and 2-m temperature (°C), and (c) 850-hPa geopotential height (gpm) and winds (full barb represents 10 m s−1) from the YOTC analysis at 0600 UTC (1400 LT) 31 May 2008. (d) As in (c), but from the domain 1 model simulation with a 27-km grid. The 500-hPa trough axis is marked by a dashed line in (a). Frontal cyclone was split into a major cyclone (L) off the northwestern Taiwan coast with a secondary mesolow (L′) off the southeastern Taiwan coast in observations (simulations). The model is initialized at 0000 UTC 31 May. (e) SST (°C) and the observed surface winds (full barb represents 5 m s−1) and temperature (°C) from surface stations at 0600 UTC (1400 LT) 31 May 2008.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
(a) 500-hPa geopotential height (gpm), (b) surface winds (full barb represents 5 m s−1) and 2-m temperature (°C), and (c) 850-hPa geopotential height (gpm) and winds (full barb represents 10 m s−1) from the YOTC analysis at 0600 UTC (1400 LT) 31 May 2008. (d) As in (c), but from the domain 1 model simulation with a 27-km grid. The 500-hPa trough axis is marked by a dashed line in (a). Frontal cyclone was split into a major cyclone (L) off the northwestern Taiwan coast with a secondary mesolow (L′) off the southeastern Taiwan coast in observations (simulations). The model is initialized at 0000 UTC 31 May. (e) SST (°C) and the observed surface winds (full barb represents 5 m s−1) and temperature (°C) from surface stations at 0600 UTC (1400 LT) 31 May 2008.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
(a) 500-hPa geopotential height (gpm), (b) surface winds (full barb represents 5 m s−1) and 2-m temperature (°C), and (c) 850-hPa geopotential height (gpm) and winds (full barb represents 10 m s−1) from the YOTC analysis at 0600 UTC (1400 LT) 31 May 2008. (d) As in (c), but from the domain 1 model simulation with a 27-km grid. The 500-hPa trough axis is marked by a dashed line in (a). Frontal cyclone was split into a major cyclone (L) off the northwestern Taiwan coast with a secondary mesolow (L′) off the southeastern Taiwan coast in observations (simulations). The model is initialized at 0000 UTC 31 May. (e) SST (°C) and the observed surface winds (full barb represents 5 m s−1) and temperature (°C) from surface stations at 0600 UTC (1400 LT) 31 May 2008.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
From the time series of Penghu soundings within the central Taiwan Strait (Fig. 10a), the shallow front with a depth <1 km passed the central Taiwan Strait around 0000 UTC (0800 LT). The prefrontal warm, moist potentially unstable southwesterly monsoon flow was replaced by postfrontal stable northeasterly flow. The time series of Tainan soundings along the southwestern coast show that the relatively calm winds with a very weak, relatively cold (θυ < 303 K) downslope flow at night, where θυ is the virtual potential temperature [
Time series of winds (full barb represents 5 m s−1) and equivalent potential temperature (K) at (a) Penghu (Station P in Fig. 1e) from 1500 UTC 30 May to 1200 UTC 31 May and at (b) Banciao (Station B in Fig. 1e) from 1200 UTC 15 Jun to 1800 UTC 16 Jun constructed from sounding data.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Time series of winds (full barb represents 5 m s−1) and equivalent potential temperature (K) at (a) Penghu (Station P in Fig. 1e) from 1500 UTC 30 May to 1200 UTC 31 May and at (b) Banciao (Station B in Fig. 1e) from 1200 UTC 15 Jun to 1800 UTC 16 Jun constructed from sounding data.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Time series of winds (full barb represents 5 m s−1) and equivalent potential temperature (K) at (a) Penghu (Station P in Fig. 1e) from 1500 UTC 30 May to 1200 UTC 31 May and at (b) Banciao (Station B in Fig. 1e) from 1200 UTC 15 Jun to 1800 UTC 16 Jun constructed from sounding data.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Time series of (a) winds (full barb represents 5 m s−1) and virtual potential temperature (K), and (b) equivalent potential temperature (K) at Tainan (Yongkang, Station T in Fig. 1e) from 1500 UTC 30 May to 1200 UTC 31 May constructed from sounding data. Local sunrise is around 0514 LT (2114 UTC, marked R) and sunset is around 1840 LT (1040 UTC, marked S).
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Time series of (a) winds (full barb represents 5 m s−1) and virtual potential temperature (K), and (b) equivalent potential temperature (K) at Tainan (Yongkang, Station T in Fig. 1e) from 1500 UTC 30 May to 1200 UTC 31 May constructed from sounding data. Local sunrise is around 0514 LT (2114 UTC, marked R) and sunset is around 1840 LT (1040 UTC, marked S).
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Time series of (a) winds (full barb represents 5 m s−1) and virtual potential temperature (K), and (b) equivalent potential temperature (K) at Tainan (Yongkang, Station T in Fig. 1e) from 1500 UTC 30 May to 1200 UTC 31 May constructed from sounding data. Local sunrise is around 0514 LT (2114 UTC, marked R) and sunset is around 1840 LT (1040 UTC, marked S).
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
During the passage of the shortwave trough (Fig. 9a), orographic lifting by the warm, moist southwesterlies (Figs. 9c and 11) combined with sea breeze–upslope flow at the surface (Figs. 9e and 11a) is favorable for the development of convective activity over southwestern Taiwan (Fig. 8b). Over western/southwestern Taiwan, in addition to onshore/upslope flow near the surface (Fig. 9e), the southwesterly flow above the shallow postfrontal northeasterlies (Figs. 9b,c and 10) impinged on the central western Central Mountain Range (CMR) with a large angle with radar echoes on the western windward slopes.
Thunderstorm activity occurred in the afternoon hours over southwestern Taiwan (Fig. 8d), where the low-level air was still relatively warm and moist. Note that under the convectively unstable southwesterly flow (Fig. 11b), lightning activity associated with coastal frontal convection and the arc-shaped scattered showers off the southwestern coast in the early morning were also evident (Fig. 8c). The relative lack of lightning over western slopes of Taiwan in the afternoon may be mainly related to the relatively weak convective vigor (or updraft strength) there.
c. Convection-allowing WRF modeling
To assess the effects of thermally driven diurnal circulations and terrain on rainfall, the convection-allowing (3 km) WRF (domain 3) is employed. A model sensitivity test without terrain is also conducted to assess the effects of island terrain on island-scale airflow and horizontal rainfall distributions.
1) Nocturnal flow regime during the frontal passage
At 1800 UTC 30 May (0200 LT 31 May), the cold air surge along the southeast coast of China behind the front is simulated in the 3-km WRF (domain 3; Fig. 12a). The maximum temperature gradient axis associated with the northeast–southwest-oriented front exists both within the Taiwan Strait and northeast of Taiwan in both the model simulation (Fig. 12a) and observations (Figs. 7a,b). The lee vortex (Lt) is simulated off the southeast coast of Taiwan with a northeasterly wind component east of Taiwan (Fig. 12a). The simulated surface flow over land is dominated by a very weak downslope flow because of nocturnal cooling (Fig. 12a), in agreement with observations (Figs. 7b and 11a). The 900-hPa upward motion (Fig. 12c) is simulated over the frontal boundary between the cold northeasterly flow and the orographically deflected warm southwesterly flow off the western Taiwan coast as well as over the southwest coast of Taiwan (Fig. 12a). Convection over land is suppressed in the model simulation (Fig. 12e) in agreement with observations (Fig. 8a). The coastal frontal convection (Fig. 12e) is apparently enhanced by orographic blocking and nocturnal cooling. From the east–west cross section, weak offshore wind component is simulated over southwestern Taiwan plain (Fig. 13a).
Model results from the domain 3 model with a 3-km grid at 1800 UTC 30 May (0200 LT 31 May). (a) 10-m winds (full barb represents 5 m s−1) and 2-m temperature (°C). (b) As in (a), but for WOT run. (c) 900-hPa vertical velocity (cm s−1) and 950-hPa winds (full barb represents 10 m s−1). (d) As in (c), but for WOT run. (e) Rainfall accumulation (mm) during 0000–0800 LT 31 May (1600 UTC 30 May–0000 UTC 31 May). (f) As in (e), but for WOT run. The model is initialized at 0000 UTC 30 May. The frontal cyclone (L) moved eastward and merged with lee vortex (Lt).
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Model results from the domain 3 model with a 3-km grid at 1800 UTC 30 May (0200 LT 31 May). (a) 10-m winds (full barb represents 5 m s−1) and 2-m temperature (°C). (b) As in (a), but for WOT run. (c) 900-hPa vertical velocity (cm s−1) and 950-hPa winds (full barb represents 10 m s−1). (d) As in (c), but for WOT run. (e) Rainfall accumulation (mm) during 0000–0800 LT 31 May (1600 UTC 30 May–0000 UTC 31 May). (f) As in (e), but for WOT run. The model is initialized at 0000 UTC 30 May. The frontal cyclone (L) moved eastward and merged with lee vortex (Lt).
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Model results from the domain 3 model with a 3-km grid at 1800 UTC 30 May (0200 LT 31 May). (a) 10-m winds (full barb represents 5 m s−1) and 2-m temperature (°C). (b) As in (a), but for WOT run. (c) 900-hPa vertical velocity (cm s−1) and 950-hPa winds (full barb represents 10 m s−1). (d) As in (c), but for WOT run. (e) Rainfall accumulation (mm) during 0000–0800 LT 31 May (1600 UTC 30 May–0000 UTC 31 May). (f) As in (e), but for WOT run. The model is initialized at 0000 UTC 30 May. The frontal cyclone (L) moved eastward and merged with lee vortex (Lt).
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Model results from the domain 3 model with a 3-km grid. (a) East–west vertical cross section of vertical velocity (cm s−1, shaded), equivalent potential temperature (K, contoured), u component (<0 m s−1, blue contours), and horizontal wind field (u, υ; full barb represents 10 m s−1) along 22.75°N (black line in Fig. 1e) at 1800 UTC 30 May in the CTRL run and (b) WOT run. The model is initialized at 0000 UTC 30 May. (c),(d) As in (a),(b), but for 0600 UTC 31 May. The model is initialized at 0000 UTC 31 May.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Model results from the domain 3 model with a 3-km grid. (a) East–west vertical cross section of vertical velocity (cm s−1, shaded), equivalent potential temperature (K, contoured), u component (<0 m s−1, blue contours), and horizontal wind field (u, υ; full barb represents 10 m s−1) along 22.75°N (black line in Fig. 1e) at 1800 UTC 30 May in the CTRL run and (b) WOT run. The model is initialized at 0000 UTC 30 May. (c),(d) As in (a),(b), but for 0600 UTC 31 May. The model is initialized at 0000 UTC 31 May.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Model results from the domain 3 model with a 3-km grid. (a) East–west vertical cross section of vertical velocity (cm s−1, shaded), equivalent potential temperature (K, contoured), u component (<0 m s−1, blue contours), and horizontal wind field (u, υ; full barb represents 10 m s−1) along 22.75°N (black line in Fig. 1e) at 1800 UTC 30 May in the CTRL run and (b) WOT run. The model is initialized at 0000 UTC 30 May. (c),(d) As in (a),(b), but for 0600 UTC 31 May. The model is initialized at 0000 UTC 31 May.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
From the model sensitivity test WOT, it is shown that propagation of the shallow cold front, the surface airflow, and the location of deep convection are significantly affected by the presence of CMR. In the WOT run, the simulated surface front moves faster and the cold air surge along the southeastern China coast is weaker than in the CTRL run (Figs. 12a,b). Without orographic blocking, the lee vortex (Lt) is not simulated (Fig. 12b). In contrast to the CTRL run, the surface airflow east of Taiwan has a southerly wind component and converges with the postfrontal northeasterlies off the northeastern coast (Fig. 12b). The 900-hPa upward motion and the convection are orientated in a northeast–southwest direction over the Taiwan area (Figs. 12d,f). In addition, no offshore wind component is simulated over southwestern Taiwan (Fig. 13b).
2) Daytime flow regime
At 0600 UTC (1400 LT) 31 May, the simulated frontal cyclone has moved eastward across Taiwan and merges with the lee vortex (L′; Fig. 14a) in agreement with the YOTC analysis (Fig. 9b). In the meantime, the simulated front advances southward (Fig. 14a). The cold northeasterlies surge southward along eastern Taiwan coast in agreement with the surface data (Fig. 9e). Southwestern Taiwan remains warm and moist with an onshore/upslope wind component at the surface (Fig. 14a). Furthermore, above the shallow postfrontal northeasterlies, the airflow has a large component impinging on the CMR (Figs. 14a,c). The orographic lifting of low-level airflow results in upward motion over the western slopes of the CMR (Fig. 14c). From the vertical cross section, orographic lifting of the combined upslope/westerly flow over southwest slopes of CMR with a hydraulic jump is evident (Fig. 13c). The simulated rainfall accumulation during 0800–1800 LT 31 May exhibits a maximum axis on the western slopes of the CMR (Fig. 14e) in agreement with observations (Fig. 1b).
Model results from the domain 3 model with a 3-km grid at 0600 UTC (1400 LT) 31 May. (a) 10-m winds (full barb represents 5 m s−1) and 2-m temperature (°C). (b) As in (a), but for WOT run. (c) 700-hPa vertical velocity (cm s−1) and 850-hPa winds (full barb represents 10 m s−1). (d) As in (c), but for WOT run. (e) Rainfall accumulation (mm) during 0800–1800 LT 31 May (0000–1000 UTC 31 May). (f) As in (e), but for WOT run. The model is initialized at 0000 UTC 31 May. The frontal cyclone (L) moved eastward and merged with lee vortex (Lt) (Fig. 12c). The merged mesocyclone is marked by L′.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Model results from the domain 3 model with a 3-km grid at 0600 UTC (1400 LT) 31 May. (a) 10-m winds (full barb represents 5 m s−1) and 2-m temperature (°C). (b) As in (a), but for WOT run. (c) 700-hPa vertical velocity (cm s−1) and 850-hPa winds (full barb represents 10 m s−1). (d) As in (c), but for WOT run. (e) Rainfall accumulation (mm) during 0800–1800 LT 31 May (0000–1000 UTC 31 May). (f) As in (e), but for WOT run. The model is initialized at 0000 UTC 31 May. The frontal cyclone (L) moved eastward and merged with lee vortex (Lt) (Fig. 12c). The merged mesocyclone is marked by L′.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Model results from the domain 3 model with a 3-km grid at 0600 UTC (1400 LT) 31 May. (a) 10-m winds (full barb represents 5 m s−1) and 2-m temperature (°C). (b) As in (a), but for WOT run. (c) 700-hPa vertical velocity (cm s−1) and 850-hPa winds (full barb represents 10 m s−1). (d) As in (c), but for WOT run. (e) Rainfall accumulation (mm) during 0800–1800 LT 31 May (0000–1000 UTC 31 May). (f) As in (e), but for WOT run. The model is initialized at 0000 UTC 31 May. The frontal cyclone (L) moved eastward and merged with lee vortex (Lt) (Fig. 12c). The merged mesocyclone is marked by L′.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Our results from a model sensitivity test (WOT) attest that orographic lifting by the moist southwesterlies aloft combined with sea breeze–upslope flow in low levels over southwestern Taiwan is important for the development of heavy precipitation over the western/southwestern slopes during the day. In the WOT run, at the surface, a postfrontal northeasterly flow prevails over northern and central Taiwan (Fig. 14b). Without orographic lifting aloft (Figs. 13d and 14d), only rainfall (Fig. 14f) associated with the frontal cyclone are simulated. There is no rainfall axis simulated along the western slopes of the CMR (Fig. 14f).
4. Intensive observation period 8
a. Favorable conditions for the development for heavy rainfall
1) Evolution of subsynoptic weather patterns
At 0000 UTC 14 June, a deep upper-level trough was over eastern China with a secondary trough extended southwestward to as far south as 20°N (Fig. 15a). As it moved eastward, the secondary trough in the subtropics propagated slower than the midlatitude trough (Fig. 15b). From 1200 to 1800 UTC 15 June, the 500-hPa chart shows the presence of a weak trough with a well-defined cyclonic center off the southwestern coast of Taiwan (Fig. 16a) as the southern part of the trough moved over the moist tongue off the southeastern China coast (Fig. 15c). This well-defined shortwave trough with a vorticity maximum is also clearly evident from both the 500-hPa global analyses of the Japan Meteorological Agency and the CWB (not shown).
(a) Horizontal distributions of 300-hPa geopotential height (gpm) and winds (full barb represents 10 m s−1) at 0000 UTC 14 Jun and (b) 1200 UTC 14 Jun. (c) Total precipitable water (mm) at 0000 UTC 15 Jun from the YOTC analysis. (d) Model results from the domain 1 model with a 27-km grid valid at 1200 UTC 15 Jun (2000 LT 15 Jun): 500-hPa QG omega (dPa s−1) and 500-hPa geopotential height (gpm; contoured). The model is initialized at 0000 UTC 15 Jun. The midlatitude cyclone and the mesolows in the subtropics are marked by Lp and L, respectively.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
(a) Horizontal distributions of 300-hPa geopotential height (gpm) and winds (full barb represents 10 m s−1) at 0000 UTC 14 Jun and (b) 1200 UTC 14 Jun. (c) Total precipitable water (mm) at 0000 UTC 15 Jun from the YOTC analysis. (d) Model results from the domain 1 model with a 27-km grid valid at 1200 UTC 15 Jun (2000 LT 15 Jun): 500-hPa QG omega (dPa s−1) and 500-hPa geopotential height (gpm; contoured). The model is initialized at 0000 UTC 15 Jun. The midlatitude cyclone and the mesolows in the subtropics are marked by Lp and L, respectively.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
(a) Horizontal distributions of 300-hPa geopotential height (gpm) and winds (full barb represents 10 m s−1) at 0000 UTC 14 Jun and (b) 1200 UTC 14 Jun. (c) Total precipitable water (mm) at 0000 UTC 15 Jun from the YOTC analysis. (d) Model results from the domain 1 model with a 27-km grid valid at 1200 UTC 15 Jun (2000 LT 15 Jun): 500-hPa QG omega (dPa s−1) and 500-hPa geopotential height (gpm; contoured). The model is initialized at 0000 UTC 15 Jun. The midlatitude cyclone and the mesolows in the subtropics are marked by Lp and L, respectively.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
(a) Horizontal distributions of 500-hPa geopotential height (gpm) and winds (full barb represents 10 m s−1) at 1800 UTC 15 Jun from the YOTC analysis. Model results from the domain 2 model with a 9-km grid valid at 1800 UTC 15 Jun (0200 LT 16 Jun): (b) 500-hPa temperature difference between the CTRL run and the WOLH run (CTRL−WOLH). (c) 500-hPa PV (PVU), winds (full barb represents 10 m s−1), and geopotential height (gpm). (d) As in (c), but for the WOLH run. The model is initialized at 0000 UTC 15 Jun. The midlevel cyclone is marked by L. The dashed lines mark the locations of the trough axes.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
(a) Horizontal distributions of 500-hPa geopotential height (gpm) and winds (full barb represents 10 m s−1) at 1800 UTC 15 Jun from the YOTC analysis. Model results from the domain 2 model with a 9-km grid valid at 1800 UTC 15 Jun (0200 LT 16 Jun): (b) 500-hPa temperature difference between the CTRL run and the WOLH run (CTRL−WOLH). (c) 500-hPa PV (PVU), winds (full barb represents 10 m s−1), and geopotential height (gpm). (d) As in (c), but for the WOLH run. The model is initialized at 0000 UTC 15 Jun. The midlevel cyclone is marked by L. The dashed lines mark the locations of the trough axes.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
(a) Horizontal distributions of 500-hPa geopotential height (gpm) and winds (full barb represents 10 m s−1) at 1800 UTC 15 Jun from the YOTC analysis. Model results from the domain 2 model with a 9-km grid valid at 1800 UTC 15 Jun (0200 LT 16 Jun): (b) 500-hPa temperature difference between the CTRL run and the WOLH run (CTRL−WOLH). (c) 500-hPa PV (PVU), winds (full barb represents 10 m s−1), and geopotential height (gpm). (d) As in (c), but for the WOLH run. The model is initialized at 0000 UTC 15 Jun. The midlevel cyclone is marked by L. The dashed lines mark the locations of the trough axes.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Figure 17 shows the 700-hPa charts and horizontal distribution of RH during and after the heavy rainfall period (Figs. 17a,b and Figs. 17c,d, respectively). At 0000 UTC 16 June, the 700-hPa trough was west of southwestern Taiwan (Fig. 17a) with a moist southwesterly monsoon flow over Taiwan (Figs. 17a,b). Warm advection at the 700-hPa level is diagnosed within the prefrontal southwesterly monsoon flow (Xu et al. 2012) indicating that the large-scale environment is favorable for rising motion or convection. At 0000 UTC 17 June, as the WPSH extended westward (Fig. 17c), Taiwan was under the influence of a WPSH with drier conditions aloft (Figs. 17d). As a result, heavy localized rainfall diminished over Taiwan.
(a) Geopotential height (gpm) and winds (full barb represents 10 m s−1) and (b) RH (%) at the 700-hPa level for 0000 UTC 16 Jun from the YOTC analysis. (c),(d) As in (a),(b), but for 0000 UTC 17 Jun.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
(a) Geopotential height (gpm) and winds (full barb represents 10 m s−1) and (b) RH (%) at the 700-hPa level for 0000 UTC 16 Jun from the YOTC analysis. (c),(d) As in (a),(b), but for 0000 UTC 17 Jun.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
(a) Geopotential height (gpm) and winds (full barb represents 10 m s−1) and (b) RH (%) at the 700-hPa level for 0000 UTC 16 Jun from the YOTC analysis. (c),(d) As in (a),(b), but for 0000 UTC 17 Jun.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
2) Mesoscale WRF modeling and diagnosis
Our 27-km WRF results valid at 1200 UTC 15 June show that the rising motion at the 500-hPa level diagnosed from the adiabatic QG forcing is on the order of −6 dPa s−1 over the southeastern quadrant of the upper-level low (Fig. 15d), which is comparable to the vertical motion diagnosed from adiabatic QG forcing for a subtropical cutoff low over the North Pacific studied by Knippertz and Martin (2007). Along the cross section (line AB in Fig. 15d), the maximum simulated vertical motion (−50 dPa s−1 at the 500–650-hPa level) over the southeast quadrant of the upper-level low (Fig. 18a) in domain 1 is significantly larger than the maximum dry adiabatic vertical motion diagnosed from QG forcing (Fig. 18b).
(a) The vertical cross section of vertical velocity ω (dPa s−1) and horizontal winds (u, υ) and (b) QG omega (dPa s−1) along line AB in Fig. 15d at 1200 UTC 15 Jun from WRF (domain 1) results. The model is initialized at 0000 UTC 15 Jun.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
(a) The vertical cross section of vertical velocity ω (dPa s−1) and horizontal winds (u, υ) and (b) QG omega (dPa s−1) along line AB in Fig. 15d at 1200 UTC 15 Jun from WRF (domain 1) results. The model is initialized at 0000 UTC 15 Jun.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
(a) The vertical cross section of vertical velocity ω (dPa s−1) and horizontal winds (u, υ) and (b) QG omega (dPa s−1) along line AB in Fig. 15d at 1200 UTC 15 Jun from WRF (domain 1) results. The model is initialized at 0000 UTC 15 Jun.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
From the 9-km WRF results, latent heat release associated with convective activity embedded in the weak trough resulted in an increase in tropospheric temperature (Fig. 16b), cyclonic flow (Fig. 16), and generation of PV (Fig. 16c). It is apparent that abundant moisture provides an energy source for convective feedbacks leading to the deepening of the trough (Figs. 16c,d) as it moved off the southeastern China coast.
b. Analysis over the Taiwan area
In the section, we first present the impacts of rain evaporative cooling and local effects on the planetary boundary layer over southwestern Taiwan and the coastal rainfall at night. Then, we will show the effects of nocturnal windward coastal rainfall on the daytime island-scale airflow and rainfall distribution.
1) Nocturnal flow regime
At 1800 UTC 15 June (0200 LT 16 June), the surface air temperature over land is about 24°–26°C (Fig. 19a), which is lower than the surface air temperature (>27°C) at the South Ship over the open ocean (Fig. 20). Most land stations around the coast show an offshore wind component at night (Fig. 19a). The time series of air temperatures at three coastal stations (Stations K, T, and H in Fig. 1e) showed the presence of cool air (24°–25°C) over southwestern Taiwan (Fig. 20), in agreement with Xu et al. (2012). From 14–16 June, the rain evaporative cooling associated with MCSs over southwestern Taiwan and the upstream ocean resulted in continuous temperature drops. The cold pool over southwestern Taiwan is a combination of rain evaporative cooling and nocturnal cooling.
(a) YOTC SST (°C, thick contour) and observed surface winds (full barb represents 5 m s−1) and temperature (°C) at 1800 UTC 15 Jun (0200 LT 16 Jun). (b) YOTC SST (contoured), observed surface winds (full barb represents 5 m s−1), and temperature from surface stations at 0600 UTC (1400 LT) 16 Jun, and dropsonde winds (full barb represents 5 m s−1), temperature (°C), and pressure (hPa) at the lowest level from 0906 to 1022 UTC (1706–1822 LT).
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
(a) YOTC SST (°C, thick contour) and observed surface winds (full barb represents 5 m s−1) and temperature (°C) at 1800 UTC 15 Jun (0200 LT 16 Jun). (b) YOTC SST (contoured), observed surface winds (full barb represents 5 m s−1), and temperature from surface stations at 0600 UTC (1400 LT) 16 Jun, and dropsonde winds (full barb represents 5 m s−1), temperature (°C), and pressure (hPa) at the lowest level from 0906 to 1022 UTC (1706–1822 LT).
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
(a) YOTC SST (°C, thick contour) and observed surface winds (full barb represents 5 m s−1) and temperature (°C) at 1800 UTC 15 Jun (0200 LT 16 Jun). (b) YOTC SST (contoured), observed surface winds (full barb represents 5 m s−1), and temperature from surface stations at 0600 UTC (1400 LT) 16 Jun, and dropsonde winds (full barb represents 5 m s−1), temperature (°C), and pressure (hPa) at the lowest level from 0906 to 1022 UTC (1706–1822 LT).
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Time series of surface temperature (°C) during 14–16 Jun (LT) at three coastal stations: Kaohisung, Tainan, and Hengchun (Stations K, T, and H in Fig. 1e), and the air temperature (°C) at the lowest level of shipboard soundings launched every 6 h at South Ship.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Time series of surface temperature (°C) during 14–16 Jun (LT) at three coastal stations: Kaohisung, Tainan, and Hengchun (Stations K, T, and H in Fig. 1e), and the air temperature (°C) at the lowest level of shipboard soundings launched every 6 h at South Ship.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Time series of surface temperature (°C) during 14–16 Jun (LT) at three coastal stations: Kaohisung, Tainan, and Hengchun (Stations K, T, and H in Fig. 1e), and the air temperature (°C) at the lowest level of shipboard soundings launched every 6 h at South Ship.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Both the summer trade wind flow over Hawaii and the southwesterly monsoon flow during the early summer rainy season over Taiwan are under a small Fr number flow regime (~0.2) with significant orographic blocking on the windward side (Smolarkiewicz et al. 1988; Li and Chen 1998). However, in sharp contrast to the windward side of the island of Hawaii, southwestern Taiwan is relatively flat due to the presence of a coastal plain. With a flat terrain and infrequent nocturnal showers (Kerns et al. 2010), the land breezes are very weak with a southeasterly wind component ~1–2 m s−1 (red line and wind barb in Fig. 7c and Fig. 11a). From the Tainan (Station T in Fig. 1e) sounding at 2100 UTC 30 May (0500 LT 31 May), the isothermal layer is below ~980 hPa with T ~ 25.5°C (Fig. 7c and Table 2). The virtual potential temperature (θυ) of the offshore flow (Table 2) is about 1–3 K lower than the incoming southwesterly monsoon flow from the open ocean (θυ ~ 305 K).
Vertical profiles of temperature, potential temperature, and virtual potential temperature at Tainan (Yongkang, Station T in Fig. 1e) at 2100 UTC 30 May (0500 LT 31 May).
For the 16 June case, with frequent heavy nocturnal rain showers that moved onshore, the depth of the cold pool over southwestern Taiwan reaches as high as 520 m in the early morning (~950 hPa in Fig. 7c; Xu et al. 2012). The winds within the cold pool have an offshore wind component. From the Tainan sounding at 2100 UTC 15 June (0500 LT 16 June), we found a thin surface layer with T ~ 24.6°C, θυ ~ 300.6 K, and an isothermal layer (23.6°C) between 995 and 975 hPa (Fig. 7c). Between 1000 and 975 hPa, the temperature is more than 1.5°C less compared to the 31 May case (Fig. 7c). For this case, because of the relatively deep offshore flow, the offshore convergence zone is a favorable location for the generation and intensification of convective cells, similar to the windward coast of the island of Hawaii where the depth of the offshore flow in the early morning reaches ~500 m with a thermal deficit (Δθυ) ~ −5 K and extends ~20 km offshore (Feng and Chen 1998).
2) Daytime flow regime
As a result of the heavy early morning rainfall and the cloud cover, temperature drops do not recover during the daytime over southwestern Taiwan (~25°C) (Fig. 20). From the sounding data at Tainan, it is apparent that the low-level virtual potential temperature continued to drop after sunrise (Fig. 21a) and the atmosphere was convectively neutral in the afternoon hours (Fig. 21b) because of rain evaporative cooling. The air temperature near the surface over the ocean off the southwestern coast recovered (Fig. 20) and was warmer than over land due to air–sea fluxes from warm SSTs underneath (Fig. 19b). As a result, sea breezes failed to develop (Fig. 21a). Furthermore, the prevailing south-southwesterly flow aloft was almost parallel to the orientation of CMR (Fig. 21a). Thus, due to the absence of both the sea breeze–upslope flow at the surface and the orographic lifting aloft, there was no afternoon rainfall maximum axis along the western slopes of the CMR (Fig. 22b). The heavy stratiform rainfall over the southwestern Taiwan coast (Xu et al. 2012) was not accompanied by lightning (Fig. 22).
Time series of (a) winds (full barb represents 5 m s−1) and virtual potential temperature (K), and (b) equivalent potential temperature (K) at Tainan (Yongkang, Station T in Fig. 1e) from 1500 UTC 15 Jun to 1200 UTC 16 Jun constructed from sounding data. Local sunrise is around 0514 LT (2114 UTC, marked R) and sunset is around 1840 LT (1040 UTC, marked S).
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Time series of (a) winds (full barb represents 5 m s−1) and virtual potential temperature (K), and (b) equivalent potential temperature (K) at Tainan (Yongkang, Station T in Fig. 1e) from 1500 UTC 15 Jun to 1200 UTC 16 Jun constructed from sounding data. Local sunrise is around 0514 LT (2114 UTC, marked R) and sunset is around 1840 LT (1040 UTC, marked S).
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Time series of (a) winds (full barb represents 5 m s−1) and virtual potential temperature (K), and (b) equivalent potential temperature (K) at Tainan (Yongkang, Station T in Fig. 1e) from 1500 UTC 15 Jun to 1200 UTC 16 Jun constructed from sounding data. Local sunrise is around 0514 LT (2114 UTC, marked R) and sunset is around 1840 LT (1040 UTC, marked S).
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Mosaic radar reflectivities (dBZ) at (a) 1800 UTC 15 Jun (0200 LT 16 Jun) and (b) 0800 UTC (1600 LT) 16 Jun. The 3-h lightning frequency (km−2) during (c) 1300–1600 LT 16 Jun 2008. (Courtesy of CWB.)
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Mosaic radar reflectivities (dBZ) at (a) 1800 UTC 15 Jun (0200 LT 16 Jun) and (b) 0800 UTC (1600 LT) 16 Jun. The 3-h lightning frequency (km−2) during (c) 1300–1600 LT 16 Jun 2008. (Courtesy of CWB.)
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Mosaic radar reflectivities (dBZ) at (a) 1800 UTC 15 Jun (0200 LT 16 Jun) and (b) 0800 UTC (1600 LT) 16 Jun. The 3-h lightning frequency (km−2) during (c) 1300–1600 LT 16 Jun 2008. (Courtesy of CWB.)
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
The northern tip of Taiwan in the afternoon hours of 16 June 2008 was in the leeside wake/convergence zone with flow splitting over the southern tip of Taiwan (Fig. 19b). At 1400 LT 16 June, the Banciao sounding over northern Taiwan is modified by deep convection with convectively unstable stratification earlier (Fig. 10b). With weak winds in the wake zone, relatively cold SSTs offshore, and afternoon solar heating over land, sea-breeze circulation and onshore/upslope flow were well developed over northern Taiwan (Fig. 19b). Heavy rainfall over northern Taiwan associated with thunderstorm activity occurred from late morning into the afternoon hours (Figs. 22b,c). Under favorable large-scale settings, two factors are important for the development of localized heavy rainfall over northern Taiwan: 1) sea-breeze circulations and onshore/upslope flows in the island wake zone and 2) convective instability.
c. Convection-allowing model
1) Nocturnal flow regime
The convection-allowing model is employed to show that under favorable large-scale settings and the presence of preexisting rain cells embedded in the southwesterly monsoon flow (Fig. 22a), the southwestern coast had a nocturnal rainfall maximum. The horizontal distributions of surface temperature and downslope/offshore flow circulation (Fig. 19a) are reproduced by the convection-allowing (3 km) WRF (Fig. 23a). Rising motion is simulated off the southwestern coast (Fig. 23c) where the offshore flow converges with the incoming decelerating southwesterly flow (Fig. 23a). Rising motion (Fig. 23c) and heavy rainfall (Fig. 23e) are simulated off the southwestern coast as the incoming flow decelerates. The convection moved inland and resulted in heavy rainfall along the southwest coast (Fig. 23e) in agreement with the observation (Fig. 1c).
Model results from the domain 3 model with a 3-km grid at 1800 UTC 15 Jun (0200 LT 16 Jun). (a) 10-m winds (full barb represents 5 m s−1) and 2-m temperature (°C). (b) As in (a), but for WOT run. (c) 900-hPa vertical velocity (cm s−1) and 950-hPa winds (full barb represents 10 m s−1). (d) As in (c), but for WOT run. (e) Rainfall accumulation (mm) during 0000–0800 LT 16 Jun (1600 UTC 15 Jun–0000 UTC 16 Jun). (f) As in (e), but for WOT run. The model is initialized at 0000 UTC 15 Jun.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Model results from the domain 3 model with a 3-km grid at 1800 UTC 15 Jun (0200 LT 16 Jun). (a) 10-m winds (full barb represents 5 m s−1) and 2-m temperature (°C). (b) As in (a), but for WOT run. (c) 900-hPa vertical velocity (cm s−1) and 950-hPa winds (full barb represents 10 m s−1). (d) As in (c), but for WOT run. (e) Rainfall accumulation (mm) during 0000–0800 LT 16 Jun (1600 UTC 15 Jun–0000 UTC 16 Jun). (f) As in (e), but for WOT run. The model is initialized at 0000 UTC 15 Jun.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Model results from the domain 3 model with a 3-km grid at 1800 UTC 15 Jun (0200 LT 16 Jun). (a) 10-m winds (full barb represents 5 m s−1) and 2-m temperature (°C). (b) As in (a), but for WOT run. (c) 900-hPa vertical velocity (cm s−1) and 950-hPa winds (full barb represents 10 m s−1). (d) As in (c), but for WOT run. (e) Rainfall accumulation (mm) during 0000–0800 LT 16 Jun (1600 UTC 15 Jun–0000 UTC 16 Jun). (f) As in (e), but for WOT run. The model is initialized at 0000 UTC 15 Jun.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Without terrain, the simulated southwesterly monsoon flow extends over southwestern Taiwan but is weaker than over the open ocean due to nocturnal cooling and land surface friction (Fig. 23b). The land breeze/offshore flow simulated over western and southwestern Taiwan in the CTRL run (Fig. 23a) is absent in the WOT run (Fig. 23b). Without the presence of land breeze/offshore flow, the simulated rising motion along the western/southwestern coast is weaker (Fig. 23d) with less rainfall (Fig. 23f) as compared with those in the CTRL run (Figs. 23c,e). Over land, the simulated convection is suppressed (Fig. 23f).
2) Daytime flow regime
The simulated air temperature over most of the coastal plain of southwestern Taiwan is colder than over the adjacent ocean (Fig. 24a). Thus, sea breeze–upslope flow is not simulated there (Fig. 24a), in agreement with observations (Fig. 19b). The observed turning from the southwesterly flow to the southeasterly flow over southwestern Taiwan as a result of orographic blocking (Fig. 19b) is simulated in the model (Fig. 24a). Rising motion is simulated over the southwestern coast (Fig. 24c) and resulted in heavy rainfall there (Fig. 24e), in agreement with the observed rainfall accumulation (Fig. 1d).
Model results from the domain 3 model with a 3-km grid at 0600 UTC (1400 LT) 16 Jun. (a) 10-m winds (full barb represents 5 m s−1) and 2-m temperature (°C). (b) As in (a), but for WOT run. (c) 700-hPa vertical velocity (cm s−1) and 850-hPa winds (full barb represents 10 m s−1). (d) As in (c), but for WOT run. (e) Rainfall accumulation (mm) during 0800–1800 LT 16 Jun (0000–1000 UTC 16 Jun). (f) As in (e), but for WOT run. The model is initialized at 0000 UTC 15 Jun.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Model results from the domain 3 model with a 3-km grid at 0600 UTC (1400 LT) 16 Jun. (a) 10-m winds (full barb represents 5 m s−1) and 2-m temperature (°C). (b) As in (a), but for WOT run. (c) 700-hPa vertical velocity (cm s−1) and 850-hPa winds (full barb represents 10 m s−1). (d) As in (c), but for WOT run. (e) Rainfall accumulation (mm) during 0800–1800 LT 16 Jun (0000–1000 UTC 16 Jun). (f) As in (e), but for WOT run. The model is initialized at 0000 UTC 15 Jun.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Model results from the domain 3 model with a 3-km grid at 0600 UTC (1400 LT) 16 Jun. (a) 10-m winds (full barb represents 5 m s−1) and 2-m temperature (°C). (b) As in (a), but for WOT run. (c) 700-hPa vertical velocity (cm s−1) and 850-hPa winds (full barb represents 10 m s−1). (d) As in (c), but for WOT run. (e) Rainfall accumulation (mm) during 0800–1800 LT 16 Jun (0000–1000 UTC 16 Jun). (f) As in (e), but for WOT run. The model is initialized at 0000 UTC 15 Jun.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
At 0600 UTC 16 June, in the WOT run without orographic blocking, there is no significant flow deceleration off the southwestern coast (Fig. 24b). The flow splitting over southern/southwestern Taiwan is also absent (Fig. 24b). Thus, the southwesterly flow (Fig. 24b) with embedded rainfall (Fig. 24f) prevails over southwestern Taiwan and the island interior. In the WOT run, simulated upward motion is mainly associated with rainfall within the southwesterly flow (Fig. 24d).
Over northern Taiwan, the thermally driven local circulations and northerly return flow offshore are simulated in the afternoon hours in the CTRL run because the area is within the wake zone (Fig. 24a). The simulated rainfall accumulation during 0800–1800 LT 16 June in the CTRL run exhibits a local maximum over northern Taiwan (Fig. 24e) in agreement with observations (Fig. 1d). In the WOT run, weak sea breezes are simulated along the northwestern and northeastern coasts (Fig. 24b). With sea-breeze circulation over northern Taiwan (Fig. 24b), rainfall is also simulated there (Fig. 24f). In addition, without terrain, scattered rainfall extends from southwestern Taiwan to island interior (Fig. 24f).
From the east–west vertical cross section along 22.75°N (black line in Fig. 1e) (Fig. 25a), in contrast of IOP 3 (Fig. 13c), the upstream flow is blocked and turned to southerlies over southwestern Taiwan without a westerly wind component in agreement with observations (Fig. 21a). As a result, the simulated orographic lifting is much weaker for the IOP 8 case (Fig. 25a) as compared to the IOP 3 case (Fig. 13c). With the winds aloft almost parallel to the orientation of CMR (Fig. 24c), no rainfall maximum is simulated along the western slopes of CMR (Fig. 24e). In the WOT run, without orographic blocking, the simulated incoming flow has a westerly wind component with rising motions over the island interior (Fig. 25b).
Model results from the domain 3 model with a 3-km grid: (a) east–west vertical cross section of vertical velocity (cm s−1, shaded), equivalent potential temperature (K, contoured), u component (<0 m s−1, blue contours) and horizontal wind field (u, υ; full barb represents 10 m s−1) along 22.75°N (black line in Fig. 1e) at 0600 UTC 16 Jun in the CTRL run and (b) WOT run. The model is initialized at 0000 UTC 15 Jun.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Model results from the domain 3 model with a 3-km grid: (a) east–west vertical cross section of vertical velocity (cm s−1, shaded), equivalent potential temperature (K, contoured), u component (<0 m s−1, blue contours) and horizontal wind field (u, υ; full barb represents 10 m s−1) along 22.75°N (black line in Fig. 1e) at 0600 UTC 16 Jun in the CTRL run and (b) WOT run. The model is initialized at 0000 UTC 15 Jun.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
Model results from the domain 3 model with a 3-km grid: (a) east–west vertical cross section of vertical velocity (cm s−1, shaded), equivalent potential temperature (K, contoured), u component (<0 m s−1, blue contours) and horizontal wind field (u, υ; full barb represents 10 m s−1) along 22.75°N (black line in Fig. 1e) at 0600 UTC 16 Jun in the CTRL run and (b) WOT run. The model is initialized at 0000 UTC 15 Jun.
Citation: Monthly Weather Review 142, 7; 10.1175/MWR-D-13-00293.1
5. Discussion
A comparison of daytime heavy rainfall periods between IOP 3 and IOP 8 is shown in Table 3. From our analysis, IOP 3 has relatively strong subsynoptic low-level forcing (Fig. 5f) whereas IOP 8 has weak upper-level forcing (Fig. 18b). However, rainfall accumulation is much higher along the southwestern coast for IOP 8 than along the windward slopes for IOP 3 during the daytime (Figs. 1b,d). Note that the total precipitable water (TPW) reaches 55 mm over the Taiwan area during IOP 3 (Fig. 3c), whereas it is over 65 mm over southwestern Taiwan and the adjacent oceans during IOP 8 (Fig. 15c). These results are in agreement with previous studies that show higher rainfall over the Taiwan area during the late season (1–15 June) of the early summer rainy season than the early season (16–31 May) because of warmer southwesterly monsoon flow with higher moisture content as the season progresses (Chen 1993; Chen and Chen 2003). Yeh and Chen (1998) also found that during the late season (1–27 June) in 1987, the rainfall over the windward slopes was more than twice of that in the early season (10–31 May) due to a more persistent southwesterly monsoon flow that contains higher moisture. Another striking feature for the 16 June case is the existence of extensive rain cells over the ocean embedded in the southwesterly monsoon flow (Fig. 22). Note that the preexisting extensive rain cells embedded in the southwesterly flow over the ocean off the southwestern coast are observed and simulated in the model even without the presence of terrain (WOT runs; Figs. 23f and 24f). As they moved toward the island, they were enhanced by the diurnal and local effects and the convergence at the leading edge of the cold pool. In contrast, for the 31 May case, the preexisting convection was mainly associated with the passage of the shallow frontal system over the western coast at night with suppressed convection over land (Fig. 8a). During the daytime, after the passage of the frontal cyclone, preexisting rain cells over the adjacent oceans were not present (Fig. 8b).
A comparison of daytime heavy rainfall periods between IOP 3 and IOP 8.
Because terrain and local effects are important in determining the location and timing of heavy precipitation under favorable large-scale settings, proper depiction of land surface conditions (terrain and land surface properties) are essential for convection-allowing mesoscale models to provide useful numerical guidance over a mountainous subtropical island and the study of the island-scale climate of rainfall. These localized heavy rainfall events cannot be accurately predicted by using the traditional 24-h synoptic forecast approach. Nowcasting with careful monitoring of these events from radars, satellites, rain gauges, and local observations is needed to provide accurate warnings in a timely manner in real-time operational settings. Furthermore, for the 16 June case, in addition to land surface forcing, it is desirable to include a cold pool in the planetary boundary layer over southwestern Taiwan in the model initial conditions through cycling runs with data assimilation. Work in this area is currently under way.
6. Summary
The diurnal cycle of island-induced circulations and their interactions with the prevailing winds play an important role in determining the locations and timings for the development of clouds and rain in many different parts of the tropical and monsoon regions (e.g., Houze et al. 1981; Johnson and Priegnitz 1981; Carbone et al. 2000; Garrett 1980; Chen and Nash 1994; Kerns et al. 2010; and others). In this study, two distinctly different early summer monsoon localized heavy rainfall events (31 May 2008 during TiMREX IOP 3 and 16 June 2008 during IOP 8) were studied using a combination of observations and numerical simulations. Our results show that the effects of terrain, rain evaporative cooling, local winds, diurnal heating cycle, and preexisting convection play very important roles in determining the timing and location of occurrences of localized heavy rainfall under favorable large-scale conditions.
For the IOP 3 case, after the shallow surface front passed central Taiwan, the postfrontal northeasterly flow is blocked and deflected over northern Taiwan leaving warm, moist, potentially unstable air over southwestern Taiwan. At 0600 UTC (1400 LT) 31 May, the development of a broad convective area over the southwestern slopes of the CMR is related to the orographic lifting of the warm, moist, southwesterly flow above the shallow postfrontal northeasterlies combined with sea breeze–upslope flow on the windward slopes.
For the 15–16 June case, with rain evaporative cooling caused by preexisting rain cells that drifted inland, the depth of the cold pool and offshore flow along the southwestern coast reached as high as 520 m. It is apparent that the early morning coastal rainfall on 16 June is related to a combined effect of nocturnal and rain evaporative cooling and orographic blocking as the preexisting rain cells drift inland.
During the daytime of 16 June, the cold pool remained over southwestern Taiwan. Because of the warm SSTs underneath, the surface air temperature over the ocean off the southwestern coast is warmer than over land. As a result, an offshore wind component along the southwestern coast persisted during the day. Furthermore, low-level winds above the cold pool are parallel to the terrain contour without significant orographic lifting aloft. Because of the presence of an offshore convergence zone, coastal rainfall is favored as rain cells drift inland.
During the daytime of 16 June, northern Taiwan was in the wake zone with weak winds, a northerly return flow, and convectively unstable atmosphere. Furthermore, with relatively cold SSTs offshore and afternoon solar heating over land, daytime sea-breeze circulations and onshore/upslope flows were well developed in the afternoon hours. As a result, heavy rainfall associated with thunderstorm activity occurred there.
Acknowledgments
The authors thank all the participants of the TiMREX for their dedication and efforts. This work is funded jointly by the National Science Foundation under Grant AGS-1142558 and NOAA under Cooperative Agreement NA07NOS4730207 and Taiwan National Science Council (NSC) under Grant NSC-100-2119-M-008-041-MY5. We thank the anonymous reviewers and Dr. Ron McTaggart-Cowan for their helpful comments, Dr. J.-S. Hong (CWB) for providing the land surface data, Hiep V. Nguyen for his assistance, and May Izumi for editing the text.
REFERENCES
Carbone, R. E., J. W. Wilson, T. D. Keenan, and J. M. Hacker, 2000: Tropical island convection in the absence of significant topography. Part I: Life cycle of diurnally forced convection. Mon. Wea. Rev., 128, 3459–3480, doi:10.1175/1520-0493(2000)128<3459:TICITA>2.0.CO;2.
Chen, C.-S., and Y.-L. Chen, 2003: The rainfall characteristics of Taiwan. Mon. Wea. Rev., 131, 1323–1341, doi:10.1175/1520-0493(2003)131<1323:TRCOT>2.0.CO;2.
Chen, C.-S., W.-C. Chen, Y.-L. Chen, P.-L. Lin, and S.-J. Lai, 2005: An investigation of orographic effects on two heavy rainfall events over southwestern Taiwan during the Mei-Yu season. Atmos. Res., 73, 101–130, doi:10.1016/j.atmosres.2004.07.005.
Chen, C.-S., Y.-L. Chen, C.-L. Liu, P.-L. Lin, and W.-C. Chen, 2007: Statistics of heavy rainfall occurrences in Taiwan. Wea. Forecasting, 22, 981–1002, doi:10.1175/WAF1033.1.
Chen, C.-S., Y.-L. Chen, P.-C. Lin, P.-L. Lin, C.-L. Liu, C.-J. Su, and W.-C. Peng, 2008: An investigation of extremely heavy rainfall events over southwestern Taiwan during the Mei-Yu season from 1997-2006 (in Mandarin with English abstract). Atmos. Sci.,35, 287–304.
Chen, F., and J. Dudhia, 2001: Coupling an advanced land surface–hydrology model with the Penn State–NCAR MM5 modeling system. Part I: Model implementation and sensitivity. Mon. Wea. Rev., 129, 569–585, doi:10.1175/1520-0493(2001)129<0569:CAALSH>2.0.CO;2.
Chen, X. A., and Y.-L. Chen, 2002: Kinetic energy budgets of the low-level jet during TAMEX IOP 5. J. Meteor. Soc. Japan, 80, 1–19, doi:10.2151/jmsj.80.1.
Chen, Y.-L., 1993: Some synoptic-scale aspects of the surface fronts over southern China during TAMEX. Mon. Wea. Rev., 121, 50–64, doi:10.1175/1520-0493(1993)121<0050:SSSAOT>2.0.CO;2.
Chen, Y.-L., 2000: Effects of island airflow on rainfall distributions during the Mei-Yu season over Taiwan. Proc. Workshop on Numerical Simulations of Precipitation in Taiwan Area, Chung-Li, Taiwan, National Central University, 5–11.
Chen, Y.-L., and N. B.-F. Hui, 1990: Analysis of a shallow front during TAMEX. Mon. Wea. Rev., 118, 2649–2667, doi:10.1175/1520-0493(1990)118<2649:AOASFD>2.0.CO;2.
Chen, Y.-L., and N. B.-F. Hui, 1992: Analysis of a relative dry front during the Taiwan area mesoscale experiment. Mon. Wea. Rev., 120, 2442–2468, doi:10.1175/1520-0493(1992)120<2442:AOARDF>2.0.CO;2.
Chen, Y.-L., and A. J. Nash, 1994: Diurnal variation of surface airflow and rainfall frequencies on the island of Hawaii. Mon. Wea. Rev., 122, 34–56, doi:10.1175/1520-0493(1994)122<0034:DVOSAA>2.0.CO;2.
Chen, Y.-L., and J. Li, 1995: Characteristic of surface airflow and pressure patterns over the island of Taiwan during TAMEX. Mon. Wea. Rev., 123, 695–716, doi:10.1175/1520-0493(1995)123<0695:COSAAP>2.0.CO;2.
Chen, Y.-L., Y.-X. Zhang, and N. B.-F. Hui, 1989: Analysis of a surface front during the early summer rainy season over Taiwan. Mon. Wea. Rev., 117, 909–931, doi:10.1175/1520-0493(1989)117<0909:AOASFD>2.0.CO;2.
Chen, Y.-L., X. A. Chen, and Y.-X. Zhang, 1994: A diagnostic study of the low-level jet (LLJ) during TAMEX IOP 5. Mon. Wea. Rev., 122, 2257–2284, doi:10.1175/1520-0493(1994)122<2257:ADSOTL>2.0.CO;2.
Chen, Y.-L., X. A. Chen, S. Chen, and Y.-H. Kuo, 1997: A numerical study of the low-level jet during TAMEX IOP 5. Mon. Wea. Rev., 125, 2583–2604, doi:10.1175/1520-0493(1997)125<2583:ANSOTL>2.0.CO;2.
Chou, M.-D., and M. J. Suarez, 1994: An efficient thermal infrared radiation parameterization for use in general circulation models. NASA Tech. Memo. 104606, Vol. 3, 85 pp.
Ciesielski, P. E., R. H. Johnson, and J. Wang, 2009: Correction of humidity biases in Vaisala RS80-H sondes during NAME. J. Atmos. Oceanic Technol., 26, 1763–1780, doi:10.1175/2009JTECHA1222.1.
Ciesielski, P. E., and Coauthors, 2010: Quality-controlled upper-air sounding dataset for TiMREX/SoWMEX: Development and corrections. J. Atmos. Oceanic Technol.,27, 1802–1821, doi:10.1175/2010JTECHA1481.1.
Clark, A. J., W. A. Gallus, M. Xue, and F. Kong, 2009: A comparison of precipitation forecast skill between small convection-allowing and large convection-parameterizing ensembles. Wea. Forecasting, 24, 1121–1140, doi:10.1175/2009WAF2222222.1.
Davis, C. A., and W.-C. Lee, 2012: Mesoscale analysis of heavy rainfall episodes from SoWMEX/TiMREX. J. Atmos. Sci., 69, 521–537, doi:10.1175/JAS-D-11-0120.1.
Feng, J., and Y.-L. Chen, 1998: Evolution of katabatic winds on the island of Hawaii during 10 August 1990. Mon. Wea. Rev., 126, 2185–2199, doi:10.1175/1520-0493(1998)126<2185:EOKFOT>2.0.CO;2.
Garrett, A. J., 1980: Orographic cloud over the eastern slopes of Mauna Loa volcano, Hawaii, related to insolation and wind. Mon. Wea. Rev., 108, 931–941, doi:10.1175/1520-0493(1980)108<0931:OCOTES>2.0.CO;2.
Gemmill, W., B. Katz, and X. Li, 2007: Daily real-time global sea surface temperature: High-resolution analysis at NOAA/NCEP. NOAA/NWS/NCEP/MMAB Office Note 260, 39 pp.
Hong, S.-Y., Y. Noh, and J. Dudhia, 2006: A new vertical diffusion package with an explicit treatment of entrainment processes. Mon. Wea. Rev., 134, 2318–2341, doi:10.1175/MWR3199.1.
Houze, R. A., S. G. Geotis, F. D. Marks, and A. K. West, 1981: Winter monsoon convection in the vicinity of North Borneo. Part I: Structure and time variation of the clouds and precipitation. Mon. Wea. Rev., 109, 1595–1614, doi:10.1175/1520-0493(1981)109<1595:WMCITV>2.0.CO;2.
Hsiao, L.-F., D.-S. Chen, Y.-H. Kuo, Y.-R. Guo, T.-C. Yeh, J.-S. Hong, C.-T. Fong, and C.-S. Lee, 2012: Application of WRF 3DVAR to operational typhoon prediction in Taiwan: Impact of outer loop and partial cycling. Wea. Forecasting, 27, 1249–1263, doi:10.1175/WAF-D-11-00131.1.
Johnson, R. H., and D. L. Priegnitz, 1981: Winter monsoon convection in the vicinity of North Borneo. Part II: Effects on large-scale fields. Mon. Wea. Rev., 109, 1615–1628, doi:10.1175/1520-0493(1981)109<1615:WMCITV>2.0.CO;2.
Johnson, R. H., and J. F. Bresch, 1991: Diagnosed characteristics of precipitation systems over Taiwan during the May–June 1987 TAMEX. Mon. Wea. Rev., 119, 2540–2557, doi:10.1175/1520-0493(1991)119<2540:DCOPSO>2.0.CO;2.
Kain, J. S., 2004: The Kain–Fritsch convective parameterization: An update. J. Appl. Meteor., 43, 170–181, doi:10.1175/1520-0450(2004)043<0170:TKCPAU>2.0.CO;2.
Kerns, J., B. Wesley, Y.-L. Chen, and M.-Y. Chang, 2010: The diurnal cycle of winds, rain, and clouds over Taiwan during the mei-yu, summer, and autumn rainfall regimes. Mon. Wea. Rev., 138, 497–516, doi:10.1175/2009MWR3031.1.
Knippertz, P., and J. E. Martin, 2007: A Pacific moisture conveyor belt and its relationship to a significant precipitation event in the semiarid southwestern United States. Wea. Forecasting, 22, 125–144, doi:10.1175/WAF963.1.
Kuo, Y.-H., and G. T.-J. Chen, 1990: The Taiwan area mesoscale experiments: An overview. Bull. Amer. Meteor. Soc., 71, 488–503, doi:10.1175/1520-0477(1990)071<0488:TTAMEA>2.0.CO;2.
Lai, H.-W., C. A. Davis, and B. J.-D. Jou, 2011: A subtropical oceanic mesoscale convective vortex observed during SoWMEX/TiMREX. Mon. Wea. Rev., 139, 2367–2385, doi:10.1175/2010MWR3411.1.
Laprise, R., 1992: The Euler equations of motion with hydrostatic pressure as an independent variable. Mon. Wea. Rev., 120, 197–207, doi:10.1175/1520-0493(1992)120<0197:TEEOMW>2.0.CO;2.
Li, J., and Y.-L. Chen, 1998: Barrier jets during TAMEX. Mon. Wea. Rev., 126, 959–971, doi:10.1175/1520-0493(1998)126<0959:BJDT>2.0.CO;2.
Li, J., Y.-L. Chen, and W.-C. Lee, 1997: Analysis of a heavy rainfall event during TAMEX. Mon. Wea. Rev., 125, 1060–1082, doi:10.1175/1520-0493(1997)125<1060:AOAHRE>2.0.CO;2.
Liou, Y.-A., and S. K. Kar, 2010: Study of cloud-to-ground lightning and precipitation and their seasonal and geographical characteristics over Taiwan. Atmos. Res., 95, 115–122, doi:10.1016/j.atmosres.2009.08.016.
Loehrer, S. M., T. A. Edmands, and J. A. Moore, 1996: TOGA COARE upper-air sounding data archive: Development and quality control procedures. Bull. Amer. Meteor. Soc., 77, 2651–2671, doi:10.1175/1520-0477(1996)077<2651:TCUASD>2.0.CO;2.
Mlawer, E. J., S. J. Taubman, P. D. Brown, M. J. Iacono, and S. A. Clough, 1997: Radiative transfer for inhomogeneous atmosphere: RRTM, a validated correlated-k model for the longwave. J. Geophys. Res., 102, 16 663–16 682, doi:10.1029/97JD00237.
Moncrieff, M. W., D. E. Waliser, M. J. Miller, M. A. Shapiro, G. R. Asrar, and J. Caughey, 2012: Multiscale convective organization and the YOTC virtual global field campaign. Bull. Amer. Meteor. Soc., 93, 1171–1187, doi:10.1175/BAMS-D-11-00233.1.
Skamarock, W. C., and Coauthors, 2008: A description of the Advanced Research WRF version 3. NCAR Tech Note NCAR/TN-475+STR, 113 pp. [Available online at http://www.mmm.ucar.edu/wrf/users/docs/arw_v3_bw.pdf.]
Smolarkiewicz, P. K., R. M. Rasmussen, and T. L. Clark, 1988: On the dynamics of Hawaiian cloud bands: Island forcing. J. Atmos. Sci., 45, 1872–1905, doi:10.1175/1520-0469(1988)045<1872:OTDOHC>2.0.CO;2.
Tai, S.-L., Y.-C. Liou, J. Sun, S.-F. Chang, and M.-C. Kuo, 2011: Precipitation forecasting using Doppler radar data, a cloud model with adjoint, and the Weather Research and Forecasting Model: Real case studies during SoWMEX in Taiwan. Wea. Forecasting, 26, 975–992, doi:10.1175/WAF-D-11-00019.1.
Tao, W.-K., and J. Simpson, 1993: The Goddard cumulus ensemble model. Part I: Model description. Terr. Atmos. Oceanic Sci., 4, 35–72.
Trier, S. B., D. B. Parsons, and T. J. Matejka, 1990: Observations of a subtropical cold front in a region of complex terrain. Mon. Wea. Rev., 118, 2449–2470, doi:10.1175/1520-0493(1990)118<2449:OOASCF>2.0.CO;2.
Xu, W., E. J. Zipser, Y.-L. Chen, C. Liu, Y.-C. Liou, W.-C. Lee, and B. J.-D. Jou, 2012: An orography-associated extreme rainfall event during TiMREX: Initiation, storm evolution, and maintenance. Mon. Wea. Rev., 140, 2555–2574, doi:10.1175/MWR-D-11-00208.1.
Yeh, H.-C., and Y.-L. Chen, 1998: Characteristics of the rainfall distribution over Taiwan during TAMEX. J. Appl. Meteor., 37, 1457–1469, doi:10.1175/1520-0450(1998)037<1457:CORDOT>2.0.CO;2.
The full sigma levels are 1.0, 0.995, 0.988, 0.98, 0.97, 0.96, 0.945, 0.93, 0.91, 0.89, 0.87, 0.85, 0.82, 0.79, 0.76, 0.73, 0.69, 0.65, 0.61, 0.57, 0.53, 0.49, 0.45, 0.41, 0.37, 0.34, 0.31, 0.28, 0.26, 0.24, 0.22, 0.2, 0.18, 0.16, 0.14, 0.12, 0.1, 0.082, 0.066, 0.052, 0.04, 0.03, 0.02, 0.01, and 0.000.
