• Adams, D. K., , and Souza E. P. , 2009: CAPE and convective events in the Southwest during the North American monsoon. Mon. Wea. Rev., 137 , 8398.

  • Akaeda, K., , Reisner J. , , and Parsons D. , 1995: The role of mesoscale and topographically induced circulations initiating a flash flood observed during the TAMEX project. Mon. Wea. Rev., 123 , 17201739.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Berenguer, M., , Sempere-Torres D. , , Corral C. , , and Sánchez-Diezma R. , 2006: A fuzzy logic technique for identifying nonprecipitating echoes in radar scans. J. Atmos. Oceanic Technol., 23 , 11571180.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Black, M. L., , Burpee R. W. , , and Marks F. D. Jr., 1996: Vertical motion characteristics of tropical cyclones determined with airborne Doppler radial velocities. J. Atmos. Sci., 53 , 18871909.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Caine, S., , Jakob C. , , Siems S. , , and May P. , 2009: Objective classification of precipitating convective regimes using a weather radar in Darwin, Australia. Mon. Wea. Rev., 137 , 15851600.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Carbone, R. E., , Tuttle J. D. , , Ahijevych D. A. , , and Trier S. B. , 2002: Inferences of predictability associated with warm season precipitation episodes. J. Atmos. Sci., 59 , 20332056.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chang, P. L., , Lin P. F. , , Jou B. J-D. , , and Zhang J. , 2009: An application of reflectivity climatology in constructing radar hybrid scans over complex terrains. J. Atmos. Oceanic Technol., 26 , 13151327.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Changnon, S. A., 2001: Damaging thunderstorm activity in the United States. Bull. Amer. Meteor. Soc., 82 , 597608.

  • Chen, C. S., , and Chen Y. L. , 2003: The rainfall characteristic of Taiwan. Mon. Wea. Rev., 131 , 13231341.

  • Chen, G. T-J., , Chou H. C. , , Chang T. C. , , and Liu C. S. , 2001: Frontal and non-frontal convection over northern Taiwan in mei-yu season (in Chinese with English abstract). Atmos. Sci., 29 , 3752.

    • Search Google Scholar
    • Export Citation
  • Chen, T. C., , Yen M. C. , , Hsieh J. C. , , and Arritt R. W. , 1999: Diurnal and seasonal variations of the rainfall measured by the Automatic Rainfall and Meteorological Telemetry System in Taiwan. Bull. Amer. Meteor. Soc., 80 , 22992312.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chen, Y. L., , and Li J. , 1995: Characteristics of surface airflow and pressure patterns over the island of Taiwan during TAMEX. Mon. Wea. Rev., 123 , 695716.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Craven, J., , Jewell R. , , and Brook H. , 2002: Comparison between observed convective cloud-base heights and lifting condensation level for two different lifted parcels. Wea. Forecasting, 17 , 885890.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Dixon, M., , and Wiener G. , 1993: TITAN: Thunderstorm Identification, Tracking, Analysis, and Nowcasting–A radar-based methodology. J. Atmos. Oceanic Technol., 10 , 785797.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Dye, J. E., , Winn W. P. , , Jones J. J. , , and Breed D. W. , 1989: The electrification of New Mexico thunderstorms. Part I: Relationship between precipitation development and the onset of electrification. J. Geophys. Res., 94 , 86438656.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Eilts, M. D., and Coauthors, 1996: Severe weather warning decision support system. Preprints, 18th Conf. on Severe Local Storms, San Francisco, CA, Amer. Meteor. Soc., 536–540.

    • Search Google Scholar
    • Export Citation
  • Fuelberg, H. E., , and Biggar D. G. , 1994: The preconvective environment of summer thunderstorms over the Florida panhandle. Wea. Forecasting, 9 , 316326.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Fulton, R. A., , Breidenbach J. P. , , Seo D. J. , , Miller D. A. , , and O’Bannon T. , 1998: The WSR-88D rainfall algorithm. Wea. Forecasting, 13 , 377395.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Gremillion, M. S., , and Orville R. E. , 1999: Thunderstorm characteristics of cloud-to-ground lightning at Kennedy Space Flight Center, Florida: A study of lightning initiation signatures as indicated by WSR-88D. Wea. Forecasting, 14 , 640649.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Heinselman, P. L., , and Schultz D. M. , 2006: Intraseasonal variability of summer storms over central Arizona during 1997 and 1999. Wea. Forecasting, 21 , 559578.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Henry, S. G., 1993: Analysis of thunderstorm lifetime as a function of size and intensity. Preprints, 26th Conf. on Radar Meteorology, Norman, OK, Amer. Meteor. Soc., 138–140.

    • Search Google Scholar
    • Export Citation
  • Hondl, K. D., , and Eilts M. D. , 1994: Doppler radar signatures of developing thunderstorms and their potential to indicate the onset of cloud-to-ground lightning. Mon. Wea. Rev., 122 , 18181836.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Huntrieser, H., , Schiesser H. H. , , Schmid W. , , and Waldvogel A. , 1997: Comparison of traditional and newly developed thunderstorm indices for Switzerland. Wea. Forecasting, 12 , 108125.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Joe, P., , Burgess D. , , Potts R. , , Keenan T. , , Stumpf G. , , and Treloar A. , 2004: The S2K severe weather detection algorithms and their performance. Wea. Forecasting, 19 , 4363.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Johns, R. H., , and Doswell C. A. III, 1992: Severe local storms forecasting. Wea. Forecasting, 7 , 588612.

  • Johnson, J. T., , Mackeen P. L. , , Witt A. , , Mitchell E. D. , , Stumpf G. J. , , Eilts M. D. , , and Thmas K. W. , 1998: The Storm Cell Identification and Tracking algorithm: An enhanced WSR-88D algorithm. Wea. Forecasting, 13 , 263276.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Johnson, R. H., , and Bresch J. F. , 1991: Diagnosed characteristics of precipitation systems over Taiwan during the May–June 1987 TAMEX. Mon. Wea. Rev., 119 , 25402557.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Jou, B. J-D., 1994: Mountain-originated mesoscale precipitation system in northern Taiwan: A case study 21 June 1991. Terr. Atmos. Oceanic Sci., 5 , 169197.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kerns, B. W. J., , Chen Y. L. , , and Chang M. Y. , 2010: The diurnal cycle of winds, rain, and clouds over Taiwan during the mei-yu, summer, and autumn rainfall regimes. Mon. Wea. Rev., 138 , 497516.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Krajewski, W. F., , and Vignal B. , 2001: Evaluation of anomalous propagation echo detection in WSR-88D data: A large sample case study. J. Atmos. Oceanic Technol., 18 , 807814.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Krehbiel, P. R., , Brook M. , , Lhermitte R. L. , , and Lennon C. L. , 1983: Lightning charge structure in thunderstorms. Proceedings in Atmospheric Electricity, L. H. Ruhnke and J. Latham, Eds., A. Deepak Publishing, 408–410.

    • Search Google Scholar
    • Export Citation
  • Kuo, J. T., , and Orville H. D. , 1973: A radar climatology of summertime convective clouds in the Black Hills. J. Appl. Meteor., 12 , 359368.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lapczak, S., and Coauthors, 1999: The Canadian National Radar Project. Preprints, 29th Int. Conf. on Radar Meteorology, Montreal, QC, Canada, Amer. Meteor. Soc., 327–330.

    • Search Google Scholar
    • Export Citation
  • Li, J., , Chen Y. L. , , and Lee W. C. , 1997: Analysis of a heavy rainfall event during TAMEX. Mon. Wea. Rev., 125 , 10601081.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lin, S. M., , and Kuo H. C. , 1996: A study of summertime afternoon convection in southern Taiwan during 1994 (in Chinese with English abstract). Atmos. Sci., 24 , 249280.

    • Search Google Scholar
    • Export Citation
  • Livingston, E. S., , Nielsen-Gammon J. W. , , and Orville R. E. , 1996: A climatology, synoptic assessment, and thermodynamic evaluation for cloud-to-ground lightning in Georgia: A study for the 1996 Summer Olympics. Bull. Amer. Meteor. Soc., 77 , 14831495.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lynn, B. H., and Coauthors, 2001: Improved simulation of Florida summer convection using the PLACE land model and a 1.5-order turbulence parameterization coupled to the Penn State–NCAR mesoscale model. Mon. Wea. Rev., 129 , 14411461.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • MacGorman, D. R., and Coauthors, 2008: TELEX: The Thunderstorm Electrification and Lightning Experiment. Bull. Amer. Meteor. Soc., 89 , 9971013.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Maddox, R. A., , Zhang J. , , Gourley J. J. , , and Howard K. W. , 2002: Weather radar coverage over the contiguous United States. Wea. Forecasting, 17 , 927934.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Mueller, C., , Saxen T. , , Roberts R. , , Wilson J. , , Betancourt T. , , Dettling S. , , Oien N. , , and Yee J. , 2003: NCAR Auto-Nowcast System. Wea. Forecasting, 18 , 545561.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Murphy, M. S., , and Konrad C. E. II, 2005: Spatial and temporal patterns of thunderstorm events that produce cloud-to-ground lightning in the interior southeastern United States. Mon. Wea. Rev., 133 , 14171430.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • O’Bannon, T., 1997: Using a ‘terrain-based’ hybrid Scan to improve WSR-88D precipitation estimates. Preprints, 28th Conf. on Radar Meteorology, Austin, TX, Amer. Meteor. Soc., 506–507.

    • Search Google Scholar
    • Export Citation
  • Overeem, A., , Holleman I. , , and Buishand A. , 2009: Derivation of a 10-year radar-based climatology of rainfall. J. Appl. Meteor. Climatol., 48 , 14481463.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Parker, M. D., , and Knievel J. C. , 2005: Do meteorologists suppress thunderstorms?: Radar-derived statistics and the behavior of moist convection. Bull. Amer. Meteor. Soc., 86 , 341358.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Reap, R. M., 1986: Evaluation of cloud-to-ground lightning data from the western United States for the 1983–84 summer seasons. J. Climate Appl. Meteor., 25 , 785799.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Reap, R. M., , and MacGorman D. R. , 1989: Cloud-to-ground lightning: Climatological characteristics and relationships to model fields, radar observations, and severe local storms. Mon. Wea. Rev., 117 , 518535.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Rickenbach, T. M., , and Rutledge S. A. , 1998: Convection in TOGA COARE: Horizontal scale, morphology, and rainfall production. J. Atmos. Sci., 55 , 27152729.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Rosenfeld, D., , Wolff D. B. , , and Atlas D. , 1993: General probability-matched relations between radar reflectivity and rain rate. J. Appl. Meteor., 32 , 5072.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Saxen, T. R., and Coauthors, 2008: The operational mesogamma-scale analysis and forecast system of the U.S. Army Test and Evaluation Command. Part IV: The White Sands Missile Range Auto-Nowcast System. J. Appl. Meteor. Climatol., 47 , 11231139.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Shafer, P. E., , and Fuelberg H. E. , 2006: A statistical procedure to forecast warm season lightning over portions of the Florida peninsula. Wea. Forecasting, 21 , 851868.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Steenburgh, W. J., , Halvorson S. F. , , and Onton D. J. , 2000: Climatology of lake-effect snowstorms of the Great Salt Lake. Mon. Wea. Rev., 128 , 709727.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Steiger, S. M., , Orville R. E. , , and Carey L. D. , 2007: Total lightning signatures of thunderstorm intensity over North Texas. Part I: Supercells. Mon. Wea. Rev., 135 , 32813302.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Steiner, M., , Houze R. A. Jr., , and Yuter S. E. , 1995: Climatological characterization of three-dimensional storm structure from operational radar and rain gauge data. J. Appl. Meteor., 34 , 19782007.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Szoke, E. J., , Mueller C. K. , , Wilson J. W. , , and Zipser E. J. , 1985: Development of convection and severe weather along outflow boundaries in northeast Colorado: Diagnosing the above-surface environment with a mobile sounding system. Preprints, 14th Conf. on Severe Local Storms, Indianapolis, IN, Amer. Meteor. Soc., 386–389.

    • Search Google Scholar
    • Export Citation
  • Tapia, A., , Smith J. A. , , and Dixon M. , 1998: Estimation of convective rainfall from lightning observations. J. Appl. Meteor., 37 , 14971509.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Toracinta, E. R., , Mohr K. I. , , Zipser E. J. , , and Orville R. E. , 1996: A comparison of WSR-88D reflectivities, SSM/I brightness temperatures, and lightning for mesoscale convective system in Texas. Part I: Radar reflectivity and lightning. J. Appl. Meteor., 35 , 902918.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Vivekanandan, J., , Zrnic D. S. , , Ellis S. M. , , Oye R. , , Ryzhkov A. V. , , and Straka J. , 1999: Cloud microphysics retrieval using S-band dual-polarization radar measurements. Bull. Amer. Meteor. Soc., 80 , 381388.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wang, C. C., , Chen G. T-J. , , and Carbone R. E. , 2005: Variability of warm-season cloud episodes over East Asia based on GMS infrared brightness temperature observations. Mon. Wea. Rev., 133 , 14781500.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Watson, A. I., , López R. E. , , and Holle R. L. , 1994: Diurnal cloud-to-ground lightning patterns in Arizona during the southwest monsoon. Mon. Wea. Rev., 122 , 17161725.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Weckwerth, T. M., 2000: The effect of small-scale moisture variability on thunderstorm initiation. Mon. Wea. Rev., 128 , 40174030.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wilson, J. W., , and Schreiber W. E. , 1986: Initiation of convective storms at radar-observed boundary-layer convergence lines. Mon. Wea. Rev., 114 , 25162536.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Yuter, S. E., , and Houze R. A. Jr., 1995: Three-dimensional kinematic and microphysical evolution of Florida cumulonimbus. Part II: Frequency distributions of vertical velocity, reflectivity, and differential reflectivity. Mon. Wea. Rev., 123 , 19411963.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zehnder, J. A., , Zhang L. , , Hansford D. , , Radzan A. , , Selover N. , , and Brown C. M. , 2006: Using digital cloud photogrammetry to characterize the onset and transition from shallow to deep convection over orography. Mon. Wea. Rev., 134 , 25272546.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zhang, J., , Howard K. , , and Gourley J. J. , 2005: Constructing three-dimensional multiple-radar reflectivity mosaics: Examples of convective storms and stratiform rain echoes. J. Atmos. Oceanic Technol., 22 , 3042.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • View in gallery

    Distributions of observation stations in Taiwan. Gray shades represent terrain heights. Locations of radar sites are marked with triangles and their respective abbreviations. Surface stations and rain gauges are denoted by blue plus signs and pink circles, respectively. Surface stations used in this study are labeled with the station numbers for reference.

  • View in gallery

    (a) Distribution of rainfall amounts (mm) on undisturbed days during the warm seasons during 2005–08. Gauge stations with local precipitation maximum are labeled with the station numbers for reference. (b) Hourly average rainfall (mm) for all rain gauges. The dashed lines indicate the crest lines of two major mountain ranges (Fig. 1).

  • View in gallery

    Frequency of occurrence (%) of (a) reflectivity >40 dBZ and (b) CG lightning during 1200–2100 LST on undisturbed days. Seven TLDS sites are denoted by plus signs.

  • View in gallery

    The accumulated rainfall (mm) according to different rainfall rates (mm h−1) at rain gauges (a) C0D360 and (b) C0U710. The numbers indicate the percentages (%) of total rainfall on undisturbed days as shown in the top right-hand corner.

  • View in gallery

    Frequency of occurrence (%) for reflectivity >40 dBZ at (a) 1400, (b) 1500, (c) 1600, (d) 1700, (e) 1800, and (f) 1900 LST on undisturbed days. The four inset boxes in (b) indicate the subdomains for calculating the Hovmöller diagrams and the movement of storm cells in Figs. 6 and 7.

  • View in gallery

    Hovmöller diagrams of the frequency of occurrence (%) of reflectivity >40 dBZ on undisturbed days for subdomains (a) N, (c) C, (e) S, and (g) E. (b),(d),(f), and (h) As in (a),(c),(e), and (g), but for the CG lightning. In each Hovmöller diagram, the frequency in the left (right) panel is averaged across (along) the long side of the subdomain as indicated in Fig. 5b. The average topographic profile is also indicated at the top of each panel.

  • View in gallery

    Histograms of the movement of storm cells identified by the SCIT algorithm between 1200 and 2100 LST on undisturbed days in the subdomains (a) N, (b) C, (c) S, and (d) E. The corresponding average speeds are plotted as the thick black lines.

  • View in gallery

    Climatological CFADs of radar reflectivity for (a) RCWF data collected at 6-min intervals and (b) RCCG data collected at 10-min intervals during 1200–2100 LST on undisturbed days. The CFAD bin size is 5 dBZ and is shaded with colors at intervals of 2% (dBZ)−1 km−1.

  • View in gallery

    Vertical profiles of mean temperature (solid line) and dewpoint temperature (dashed line) differences between the mean profiles from undisturbed days vs TSA (heavy line) and non-TSA (thin line) days (Table 2). The average profiles were noted at the Panchiao station (46692) in northern Taiwan taken at 0800 LST (0000 UTC).

  • View in gallery

    Box-and-whiskers plot of T–Td from 1000 to 400 hPa for the TSA days (gray) and non-TSA days (white). The bottom and top of the box are the value of the first (Q1) and third (Q3) quartiles, respectively. The line in the box represents the median value. Outliers are the points that fall below Q1 − 1.5(IQR) or above Q3 + 1.5(IQR) (as the length of whiskers), where the IQR (interquartile range) is equal to the difference between Q3 and Q1.

  • View in gallery

    Histograms of the frequencies of the observed wind directions between (a) 0–3 and (c) 3–6 km by radiosonde observations at Panchiao at 0800 LST (0000 UTC) on TSA days. (b),(d) As in (a),(c), but for non-TSA days. The corresponding average wind speeds are plotted with thick black lines.

  • View in gallery

    Hourly average surface wind at 1200 LST on (a) TSA and (b) non-TSA days in northern Taiwan. Full-wind barbs correspond to 1 m s−1 and half barbs correspond to 0.5 m s−1. The terrain heights are also indicated with gray shading. Hourly average (c) temperature (°C) and (d) dewpoint temperature (°C) are shown for stations 46690 (Danshui), 46692 (Taipei), and 46694 (Keelung). The locations of the surface stations are indicated in Fig. 1. TSA and non-TSA days are indicated with solid and dashed lines, respectively.

  • View in gallery

    As in Fig. 12, but representing central Taiwan. The locations of stations 46742 (Yungkang), 46748 (Chiayi), and 46778 (Chigu) are indicated in Fig. 1.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 152 152 48
PDF Downloads 136 136 42

Warm Season Afternoon Thunderstorm Characteristics under Weak Synoptic-Scale Forcing over Taiwan Island

View More View Less
  • 1 Department of Atmospheric Sciences, National Taiwan University, and Central Weather Bureau, Taipei, Taiwan
  • | 2 Central Weather Bureau, Taipei, Taiwan
  • | 3 Department of Atmospheric Sciences, National Taiwan University, Taipei, Taiwan
  • | 4 National Center for Atmospheric Research, Boulder, Colorado
© Get Permissions
Full access

Abstract

The spatial and temporal characteristics and distributions of thunderstorms in Taiwan during the warm season (May–October) from 2005 to 2008 and under weak synoptic-scale forcing are documented using radar reflectivity, lightning, radiosonde, and surface data. Average hourly rainfall amounts peaked in midafternoon (1500–1600 local solar time, LST). The maximum frequency of rain was located in a narrow strip, parallel to the orientation of the mountains, along the lower slopes of the mountains. Significant diurnal variations were found in surface wind, temperature, and dewpoint temperature between days with and without afternoon thunderstorms (TSA and non-TSA days). Before thunderstorms occurred, on TSA days, the surface temperature was warmer (about 0.5°–1.5°C) and the surface dewpoint temperature was moister (about 0.5°–2°C) than on non-TSA days. Sounding observations from northern Taiwan also showed warmer and higher moisture conditions on TSA days relative to non-TSA days. The largest average difference was in the 750–550-hPa layer where the non-TSA days averaged 2.5°–3.5°C drier. These preconvective factors associated with the occurrences of afternoon thunderstorms could be integrated into nowcasting tools to enhance warning systems and decision-making capabilities in real-time operations.

Corresponding author address: Ben Jong-Dao Jou, Dept. of Atmospheric Sciences, National Taiwan University, No. 1, Sec. 4, Roosevelt Road, Taipei 10772, Taiwan. Email: jouben@ntu.edu.tw

Abstract

The spatial and temporal characteristics and distributions of thunderstorms in Taiwan during the warm season (May–October) from 2005 to 2008 and under weak synoptic-scale forcing are documented using radar reflectivity, lightning, radiosonde, and surface data. Average hourly rainfall amounts peaked in midafternoon (1500–1600 local solar time, LST). The maximum frequency of rain was located in a narrow strip, parallel to the orientation of the mountains, along the lower slopes of the mountains. Significant diurnal variations were found in surface wind, temperature, and dewpoint temperature between days with and without afternoon thunderstorms (TSA and non-TSA days). Before thunderstorms occurred, on TSA days, the surface temperature was warmer (about 0.5°–1.5°C) and the surface dewpoint temperature was moister (about 0.5°–2°C) than on non-TSA days. Sounding observations from northern Taiwan also showed warmer and higher moisture conditions on TSA days relative to non-TSA days. The largest average difference was in the 750–550-hPa layer where the non-TSA days averaged 2.5°–3.5°C drier. These preconvective factors associated with the occurrences of afternoon thunderstorms could be integrated into nowcasting tools to enhance warning systems and decision-making capabilities in real-time operations.

Corresponding author address: Ben Jong-Dao Jou, Dept. of Atmospheric Sciences, National Taiwan University, No. 1, Sec. 4, Roosevelt Road, Taipei 10772, Taiwan. Email: jouben@ntu.edu.tw

1. Introduction

The purpose of this study is to investigate the characteristics of afternoon thunderstorms on the island of Taiwan during the warm season (May–October 2005–2008) for days when synoptic forcing was weak. It is particularly challenging to forecast thunderstorms in Taiwan, which is a mountainous island characterized by the Central Mountain Range (CMR) running across most of it in a north-northeast–south-southwest orientation at an average height of about 2 km (Fig. 1). Mountains not only generate local circulations but also interact with large-scale low-level winds to produce localized rain showers (Akaeda et al. 1995; Li et al. 1997). The sea–land breeze and anabatic–katabatic flows are important in moistening the boundary layer to provide favorable conditions for the initiation of afternoon thunderstorms (Johnson and Bresch 1991; Chen and Li 1995). Hence, in the complex terrain of Taiwan, the orographic effects play a significant role on the spatial rainfall variations throughout the year (Chen and Chen 2003). During the summer months in Taiwan, most of the convection is triggered by diurnal heating rather than by synoptic systems; thus, the contrast in rainfall occurrence between daytime–nighttime is largest during the summer (Kerns et al. 2010).

Thunderstorms can produce lightning, high winds, heavy rainfall, damaging hail, and tornadoes, which lead to the loss of property and human life (Huntrieser et al. 1997; Changnon 2001). However, it is difficult to accurately forecast the precise location and time of thunderstorms because they occur on small temporal and spatial scales that are not well resolved by observational data (Weckwerth 2000) and numerical models (Lynn et al. 2001). To provide guidance for forecasting the occurrence of thunderstorms, some preconvective indices based on thermodynamic and kinematic features such as stability, wind shear, and relative humidity (Fuelberg and Biggar 1994; Huntrieser et al. 1997) have been proposed. Others have used thunderstorm climatologies based on reflectivity and lightning data to improve the accuracy of forecasting the timing and location of thunderstorms (Shafer and Fuelberg 2006; Saxen et al. 2008).

Jou (1994) found that thunderstorms in northern Taiwan initiated on the mountain peaks and propagated down the terrain slope and brought heavy rain to the basin and plain areas. Downslope propagation was aided by the storm outflows moving down the mountains colliding with mountain upslope winds. In addition, major intensification occurred when the outflows collided with the sea breeze near the base of the mountains. Others have noted similar enhancement of convective activity when mountain-induced flows collide with thunderstorm outflows (Szoke et al. 1985; Wilson and Schreiber 1986). As a result of these situations, the maximum storm activity, during the afternoon, often occurs along the windward slopes of the mountains rather than at higher elevations farther inland (Johnson and Bresch 1991; Lin and Kuo 1996; Chen et al. 2001).

Although previous studies have analyzed the characteristics of heavy rainfall in Taiwan, specific studies about the space–time distributions of afternoon thunderstorms is limited owing to the absence of a dense data network and long-term data collection. Radar is the primary dataset for examining the characteristics. Radiosonde and surface station data are used to investigate differences in the preconvective environment between thunderstorm and nonthunderstorm days. The data and methodology used in this paper are described in section 2. The spatial and temporal variations of thunderstorms in different parts of Taiwan are presented in section 3. Preconvective environments on days with and without afternoon thunderstorms are analyzed in section 4, followed by our conclusions in section 5.

2. Data and methodology

In this study, in order to assure that the selected cases represent the impacts of local geographic factors, cases influenced by synoptic-scale perturbations (fronts, tropical cyclones, etc.) were eliminated based on examination of daily weather maps and weather outlooks issued by the Central Weather Bureau (CWB). Accordingly, from the 4 yr (2005–2008) of warm season (May–October) data, 277 days out of a total of 736 were selected as synoptically undisturbed days.

On undisturbed days, rain gauge data are used to examine the spatial and diurnal distributions of rainfall amounts. Radar reflectivity and cloud-to-ground (CG) lightning-based climatologies are used to investigate the spatial and temporal variabilities of rainfall. Surface station data and radiosonde observations from Panchiao, Taiwan, are used to analyze the variations in local-circulation and large-scale atmospheric conditions favorable for the development of thunderstorms. The distributions of rain gauges, surface observation stations, and radar sites are shown in Fig. 1.

Radar data were collected from four operational Doppler radars: Wu-Fan San (RCWF), Hual-Lien (RCHL), Chi-Gu (RCCG), and Ken-Ting (RCKT). The radars were installed by the CWB of Taiwan and the data have been available since 2001. The reflectivity observations from the individual radars were then combined to generate 3D reflectivity mosaic grids (Zhang et al. 2005). The mosaic grid has a spatial resolution of 0.0125° on the latitude–longitude coordinate system and a 10-min update cycle. Before the mosaic was developed, reflectivity observations were quality controlled (Chang et al. 2009) to remove nonprecipitation echoes. The complex terrain around the radars results in beam blockage in some directions and ground clutter in other directions and creates difficulties in using radar to characterize thunderstorms over the CMR and eastern Taiwan. Chang et al. (2009), following the procedures of others (Krajewski and Vignal 2001; Overeem et al. 2009), conducted a comprehensive study using the Taiwan radars and rain gauges to identify radar beams that are blocked or partially blocked by terrain. Chang’s study used radar reflectivity data from the four CWB radars and rainfall data from at least 370 rain gauges for the 3-yr period from 2005 to 2007. This data were used to identify the blocked beams and produce hybrid scans for each radar. Hybrid scans are a set of radar bins, from the lowest elevation angle, that do not have significant blockage or clutter (O’Bannon 1997; Maddox et al. 2002). In the current study, reflectivity observations are quality controlled by the hybrid scans constructed by Chang et al. (2009). These data were then used to identify and characterize afternoon thunderstorms.

CG lightning data were obtained from the Vaisala Total Lightning Detection System (TLDS), provided by the Taiwan Power Company (TPC), and used along with radar observations to investigate the convective features of thunderstorms (e.g., Watson et al. 1994).

a. Rainfall distribution

Figure 2 shows the total accumulated rainfall for the 277 undisturbed days. Figure 2 was created by using an inverse distance-weighting function with a 30-km radius of influence (Chang et al. 2009). The areas in which the total rainfall accumulation exceeds 1000 mm are principally located on the western side of the Snow Mountain Range (SMR) and the CMR (Fig. 2a). A local maximum rainfall of about 3500 mm (A in Fig. 2a) was on the slope of the middle portion of the CMR. Secondary maxima of about 2400 mm were located at areas B and C, which were on the western slope of the southern CMR and the northern tip of the CMR, respectively. Another local maximum (D in Fig. 2a) of approximately 1800 mm was located on the west slope of the SMR. Overall, there was much less rainfall in coastal areas.

Figure 2b shows the hourly averaged rainfall. The hourly average rainfall increased significantly during the afternoon and reached a maximum between 1500 and 1700 local solar time (LST). This afternoon maximum suggests that the rainfall is associated with thunderstorms. This has also been reported by Johnson and Bresch (1991), Lin and Kuo (1996), and Chen et al. (2001).

b. Reflectivity and lightning climatology

In this present study the frequencies of occurrence of reflectivity and lightning above set thresholds are used to construct diurnal climatology maps of thunderstorms during undisturbed days.

Radar reflectivity climatology maps have been used to study spatial distributions of precipitation in other regions (Kuo and Orville 1973; Steenburgh et al. 2000; Heinselman and Schultz 2006). Here, radar reflectivity greater than 40 dBZ is used as the lower threshold for defining the presence of a thunderstorm, similar to other studies (Reap 1986; Dye et al. 1989; Reap and MacGorman 1989; Toracinta et al. 1996; Tapia et al. 1998; MacGorman et al. 2008). A reflectivity threshold of greater than 40 dBZ is considered a reasonable criteria for convective activity (Livingston et al. 1996) as it has been shown to discriminate between convective and stratiform rain (e.g., Rickenbach and Rutledge 1998; Parker and Knievel 2005).

To produce the radar reflectivity climatology maps, the frequencies of occurrence of reflectivity greater than 40 dBZ for each 0.0125° (grid point) at 10-min intervals are computed. For CG lightning data, one lightning occurrence was counted if one or more strikes of CG lightning were detected in 10 min; the temporal resolution of the CG lightning observations was downscaled in time to match up with the radar temporal observations. The frequency of lightning occurrence was then computed similarly to the frequency of radar occurrence.

Comparison of the spatial distribution of reflectivity frequencies greater than 40 dBZ in Fig. 3a with the spatial distribution of CG lightning frequencies in Fig. 3b shows close correspondence. Therefore, similar to Murphy and Konrad (2005), both reflectivity and lightning climatology maps can be used to address spatial and temporal distributions of thunderstorms.

3. Characteristics of thunderstorms

a. Temporal variations

Figure 3a shows the distribution of reflectivity frequencies greater than 40 dBZ from 1200 to 2100 LST on undisturbed days. It shows two local maxima along the mountain-slope regions approximately parallel to the SMR and CMR. These local maxima had frequencies of approximately 4% and 2%. Low-frequency areas were found near the coastal and the eastern parts of Taiwan that coincide with the low rainfall amounts shown in Fig. 2a.

While there is general agreement in the overall appearance of Fig. 2a (rain gauge accumulation) with Fig. 3a (>40-dBZ reflectivity frequency), there are some exceptions, particularly at location C in Fig. 2a, where the local maximum in rainfall accumulation does not have a corresponding maximum in the reflectivity frequency plot (Fig. 3a). To help clarify this apparent disagreement, rain gauge stations C0D360 and C0U710 are examined; these gauges are located in Fig. 2a at D and C, respectively. At location D the small maximum in rainfall accumulation is also suggested in the reflectivity frequency map (Fig. 3a). The discrepancy at station COU710 could arise if a relatively greater portion of the accumulated rainfall occurred at reflectivity intensities of less than 40 dBZ. A reflectivity of 40 dBZ corresponds to a rainfall rate of approximately 22 mm h−1 based on the empirical ZR relationship, where Z = 250R1.2 (Rosenfeld et al. 1993; Fulton et al. 1998), which is for tropical convective systems. Figure 4 shows that the amount of rainfall that fell at 5-mm increments at rain gauges C0D360 and C0U710. Station C0D360 had 47.8% of the accumulated rain contributed by rates below 20 mm h−1 compared to 63.5% for station C0U710. Thus, the majority of the rainfall at station C0U710 occurred at rates below the 40-dBZ reflectivity level, thus explaining the absence of a maximum in reflectivity frequency at location C in Fig. 3a.

The evolutionary patterns of reflectivity frequencies greater than 40 dBZ during 1400–1900 LST on undisturbed days are shown in Fig. 5. Similar to Fig. 2a, the regions of maximum frequency are primarily along the western slopes of the SMR and CMR with the maximum frequency occurring in mid- to late afternoon. The four boxed regions in Fig. 5b enclose the maximum frequency locations; in general they correspond closely to the areas of maximum rainfall in Fig. 2a. As observed for the precipitation accumulation map in Fig. 2a, the frequency of greater than 40 dBZ is significantly higher in the central-southern region of Taiwan compared to northern Taiwan.

b. Hovmöller diagram over the mountain slopes

Using the Hovmöller diagram (e.g., Carbone et al. 2002; Wang et al. 2005), the evolution and motion characteristics of afternoon thunderstorms on undisturbed days were further investigated in four subdomains (N, C, S, and E in Fig. 5b). Hovmöller diagrams are presented for both the reflectivity and CG lightning frequency fields. The frequencies were averaged in these subdomains along the short and long axes of the polygons. The polygons were aligned such that the short axes are roughly perpendicular to the SMR and CMR ridgelines and the long axes are parallel to the ridgelines. In Fig. 6, the average perpendicular component (highlighting E–W variations) is shown in the left panel of each figure and the average parallel component (N–S variations) is shown in the right panel. The reflectivity frequencies are presented in the left column and lightning frequencies in the right column of Fig. 6 and both are plotted with respect to time. The average terrain is shown at the top of each Hovmöller diagram.

1) Subdomain N

The perpendicular component for the subdomain N (Fig. 6a) shows the frequency greater than 2.0% started for the hour 1300–1400 LST near an average height of 300 m on the windward slope of the SMR and reached a maximum frequency of about 3.5% at 1500–1600 LST. For the parallel component, the spatial frequency distribution revealed slightly higher frequencies toward the north. Both components showed the duration of frequencies greater than 2.0% was between 1300 and 1700 LST. There were no significant time trends to suggest propagation.

2) Subdomain C

The spatial distribution of the perpendicular component in subdomain C (Fig. 6c) is similar to that in subdomain N (Fig. 6a) except with a higher maximum frequency. The zone of frequency greater than 2.0% was at an average height of 400 m on the windward slope of the CMR. The frequency greater than 2.0% started at the hour 1400–1500 LST, which was 1 h later than for subdomain N, and reached a maximum frequency of about 6.5% during 1500–1600 LST. The parallel component showed slightly lower frequencies in the north. Both components showed the period of frequency greater than 2.0% was from 1400 to 1900 LST, and the duration was about 4–5 h, which was 1 h longer than in subdomain N. Again, there were no significant time trends to suggest propagation.

3) Subdomain S

Figure 6e shows the frequency distributions of the perpendicular and parallel components in subdomain S, which has a steeper mountain slope than in subdomains N and C. The perpendicular component showed the zone of frequency greater than 2.0% was mainly located below 100 m in the foothills of the southern CMR. The time when frequency greater than 2.0% started and reached its maximum was the same as for subdomain C, but the maximum frequency was lower with a value of about 4%. The perpendicular and parallel components showed the duration of frequency greater than 2.0% was about 1 h shorter than in the subdomain C, which was from 1400 to 1800 LST. The time trend of the parallel component indicated a northward propagation of the maximum, which also increased as it moved northward.

4) Subdomain E

Figure 6g shows relatively lower frequencies. The perpendicular and parallel components, showed frequencies greater than 2.0% started at 1400–1500 LST and ended at 1500–1600, which was shorter than in the other three subdomains. It is noteworthy that the zone of frequency greater than 2.0% was principally located at much higher elevations. Previous analysis in section 3a suggested that the rainfall in this subdomain fell at lower rainfall rates, implying that it was more stratiform in nature.

5) Lightning frequency

The CG lightning frequencies for all subdomains were lower (0.1%–0.2%) than those of the reflectivity; however, the local maximum regions were highly correlated. The Hovmöller diagrams for both reflectivity and lightning (Fig. 6) showed that thunderstorms were most frequent along the lower slopes of the mountains and occurred primarily during the afternoon hours. The Hovmöller diagrams, with the exception of subdomain S, showed little evidence of storm motion.

c. Movement of storms

Several existing techniques (Dixon and Wiener 1993; Eilts et al. 1996; Lapczak et al. 1999) have been used to detect, classify, and track thunderstorms. Joe et al. (2004) has shown that various cell-tracking algorithms generally provide similar cell motions. The Storm Cell Identification and Tracking (SCIT) algorithm, one of the components in the Warning Decision Support System (WDSS; Johnson et al. 1998), is used in the current study to examine cell characteristics associated with afternoon thunderstorms. Considering the serious beam blockages encountered by RCKT and RCHL (Chang et al. 2009), only RCWF and RCCG data were used in this study. To be consistent with the reflectivity climatology, a reflectivity threshold of 40 dBZ was used to define storm cells.

On undisturbed days, a total of 35 542 storm cells were identified over land by the SCIT algorithm from RCWF observations, and 93.3% of them occurred between 1200 and 2100 LST. Meanwhile, from RCCG observations, 47 283 were identified and 89% of these occurred between 1200 and 2100 LST. The peak in storm cell occurrence for both radars was between 1500 and 1600 LST (Table 1). As seen in Figs. 5 and 6, the greatest reflectivity frequencies greater than 40 dBZ occurred in northern and central to southern Taiwan. The diurnal cycle of the storm cell frequency was very similar to the results of Henry (1993) and Saxen et al. (2008).

Storm-cell movements obtained from SCIT for subdomains N, C, S, and E (Fig. 5b) from 1200 to 2100 LST are shown in Fig. 7. Movements of storm cells with tracking durations of 30 min or longer were divided into eight directions, and the corresponding speeds were calculated. Cell motions in all subdomains were in all directions however, with a tendency for the preferred direction to be toward the northeast. In the subdomain N (Fig. 7a), 32% of the movement was to the northeast with a mean speed of around 6 m s−1, which was parallel to the orientation of the SMR (northeast–southwest) in northern Taiwan. Similarly, the movements of storm cells in subdomains C and S (Figs. 7b and 7c) were largely parallel to the orientations of the middle and southern portions of the CMR. Cell speeds were generally near 6 m s−1 with little variation in direction of motion or between subdomains. The movements of storm cells in each subdomain suggested that the topography may have an impact not only on the development but also the movement of thunderstorms.

d. Composite vertical structures

The vertical structures of the afternoon thunderstorms were analyzed using the climatological contoured frequency by altitude diagrams (CFADs), which are the frequency distributions of any given reflectivity at a given altitude (Yuter and Houze 1995). The algorithm has been used to investigate the detailed evolution of vertical motion and reflectivity fields in different weather systems (Yuter and Houze 1995; Black et al. 1996) and precipitation regimes (Caine et al. 2009). In this study, CFADs were computed in convective regions that were roughly defined as areas within 10-km radius of the storm-cell locations (Steiner et al. 1995) identified by the SCIT algorithm (Johnson et al. 1998).

The reflectivity range for computing CFADs was from −10 to 65 dBZ with an interval of 5 dBZ, and the vertical domain was defined between 2 and 17 km MSL with an interval of 1.0 km. RCWF data were analyzed for thunderstorms in northern Taiwan and RCCG data for central to southern Taiwan. The convective CFADs during 1200–2100 LST on undisturbed days are shown in Fig. 8. Similar to what was presented in Steiner et al. (1995), the CFAD from RCWF (Fig. 8a) showed a high reflectivity frequency (>16%) of 35–45 dBZ at a height of 2.5–6.5 km, and the reflectivity frequency for 10–20 dBZ was also high (>18%) above 9.5 km. The reflectivity frequency (>14%) decreased with increasing height at a rate of approximate 2.7 dBZ km−1. The 2% contour at an altitude of 2 km was between −2 and 55 dBZ in northern Taiwan. Below 5 km, a nearly vertically aligned contour with a strong gradient was seen within the range of frequencies between 45 and 55 dBZ. This vertically aligned convective profile is different from the brightband maximum at the freezing level in the stratiform profile (not shown), as has been documented in numerous previous studies (e.g., Steiner et al. 1995; Yuter and Houze 1995). Similar distributions were found in southern Taiwan (observed by the RCCG radar; Fig. 8b). The local maxima at 45 dBZ in reflectivity and 2 km in height showed a higher frequency (>18%) than that in northern Taiwan (∼16%; Fig. 8a), indicating stronger afternoon thunderstorms in southern Taiwan than in the north.

The composite vertical structures (Fig. 8) showed that a climatological vertical extent of 40 dBZ can reach to an altitude higher than 10 km MSL, which typically corresponds to a temperature of lower than −20°C in warm seasons over the Taiwan area. This suggests that the great majority of storms could potentially produce CG lightning (e.g., Krehbiel et al. 1983; Hondl and Eilts 1994; Gremillion and Orville 1999), as shown in Figs. 3 and 6.

4. Preconvective environments

The ambient, environmental conditions for the initiation of thunderstorms have been summarized by Johns and Doswell (1992), which include the presence of conditional instability, a moist layer in the lower or middle troposphere, and a lifting source. Huntrieser et al. (1997) have discussed various thermodynamic and kinematic parameters derived from soundings that are useful for forecasting thunderstorms. These parameters have been utilized to identify preconvective environments conducive to thunderstorm development (Fuelberg and Biggar 1994; Huntrieser et al. 1997; Adams and Souza 2009).

Sounding and station parameters based on the above studies are investigated here to determine their utility in Taiwan. Undisturbed days (identified in section 2) were separated into TSA non-TSA days according to radar reflectivity and lightning data. Sounding and surface observations for TSA and non-TSA days are analyzed and compared to determine if differences exist.

a. Definition of thunderstorm days

As discussed in section 3a, reflectivity values greater than 40 dBZ and CG lightning are used here to specify spatial and temporal distributions of thunderstorms over Taiwan (Tapia et al. 1998; Murphy and Konrad 2005; Steiger et al. 2007). Accordingly, a TSA day is declared if reflectivity greater than 40 dBZ and CG lightning occurs between 1200 and 2100 LST on undisturbed days. Further, TSA days are identified for the four subdomains defined in Fig. 5b. Table 2 shows there are 89 TSA days identified in northwestern Taiwan (N), 145 days in central Taiwan (C), 94 days in southern Taiwan (S), and only 49 days in northeastern Taiwan (E). The variations in TSA days among different geographic regions generally agree with the results of the reflectivity and lightning climatology shown in section 3. In addition, more TSA days were identified in July and August than in any other warm months (Table 2), indicating a greater response to maximum solar heating during this time of the year.

b. Mean sounding profiles

There are two operational sounding sites, Panchiao (46692) and Hualien (46699), which are located in northern and eastern Taiwan, respectively (Fig. 1). Hualien is located very close (∼10 km) to the foothills; thus, the low-level winds were frequently affected by terrain blocking that resulted in the winds being unrepresentative of the preconvective environment. Therefore, the Panchiao sounding, launched at approximately 0800 LST (0000 UTC), was selected to represent the large-scale preconvective environment. For all 277 undisturbed days, average temperature (T) and dewpoint temperature (Td) profiles were obtained for the 89 TSA days and 188 non-TSA days. Differences between the average of all undisturbed days and the averages for the TSA days and non-TSA days are shown in Fig. 9.

Figure 9 shows that on TSA days, the near-surface layer was relatively warmer (+0.5°C in mean temperature) and moister (+1.0°C in mean dewpoint) than the average of all undisturbed days. In contrast, on non-TSA days, the near-surface layer was relatively cooler (−0.2°C) and drier (−0.5°C) than the average of all undisturbed days. In the lower to middle troposphere, the profiles of dewpoint temperature showed noticeable differences between TSA and non-TSA days. Non-TSA days were drier than TSA days, with a difference of 1.5°–3.5°C from the near-surface layer to 500 hPa. The largest average difference was in the 750–550-hPa layer where the non-TSA days averaged 2.5°–3.5°C drier.

Figure 10 is a box-and-whiskers plot comparing the dewpoint depression (TTd) versus pressure levels for TSA and non-TSA days. While there is significant spread in the data, the non-TSA days had larger median (TTd) differences. The largest difference between TSA and non-TSA days was in the 850–600-hPa layer. This tendency for more moist midlevel conditions on TSA days generally agrees with Fuelberg and Biggar (1994), who showed the average relative humidity at midlevels was higher on days with strong convection.

Jou (1994) has shown that convective systems in northern Taiwan develop along the mean low- to middle-tropospheric wind. Histograms of wind directions between altitudes of 0–3 and 3–6 km for each sounding on TSA and non-TSA days are shown in Fig. 11. While the most common wind direction was SW for both TSA and non-TSA days, it was more pronounced for the TSA days. The tendency for southwesterly winds on TSA days likely contributed to the higher frequency of northeastward-moving storms shown in Fig. 7. Winds from the southwest quadrant are likely to bring relatively warmer and moister conditions, which are more likely to trigger moist convection when encountering the mountains. Also more moist conditions at midlevels will reduce the entrainment of dry air into growing cumulus (Chen et al. 2001; Zehnder et al. 2006). Similar results were documented in Fuelberg and Biggar (1994); they found that strong convection was coupled with southwesterly winds over the Florida panhandle. In addition, the stratiform regions of the thunderstorms over the western sides of the SMR might be advected by the low- to midlevel southwesterly winds to potentially increase the rainfall accumulation of area C (Fig. 2a) on the eastern sides of the SMR.

c. Diurnal variations

Surface station observations are examined for diurnal variations in wind, temperature, and dewpoint between TSA and non-TSA days. Due to the relative sparsity of surface stations (blue plus signs in Fig. 1) in southern (subdomain S) and northeastern (subdomain E) Taiwan, the surface station analyses covers only parts of central and northern Taiwan, particularly in the plains area.

Diurnal variations in average surface winds on TSA and non-TSA days revealed a convergence pattern from 1000 to 1400 LST (1200 LST is shown in Figs. 12a and 12b), which was caused by the northwestern flow from the Danshui River valley and northeastern flow from the Keelung River valley in northern Taiwan. Sea breezes were also present on both TSA and non-TSA days, but the wind directions in Taipei Basin had a more northeasterly component on non-TSA days (Fig. 12b). The results are consistent with other studies in undisturbed environments over Taiwan area (Johnson and Bresch 1991; Chen et al. 1999; Kerns et al. 2010). The surface mean temperatures from Danshui (46690), Taipei (46692), and Keelung (46694) stations (Fig. 12c) were generally 0.5°–1.5°C higher on TSA days than on non-TSA days before afternoon thunderstorms occurred at about 1300 LST. The temperature at Taipei was cooler on TSA days than on non-TSA days during afternoon hours. As reported by Fuelberg and Biggar (1994) for Florida, this temperature difference was likely due to the evaporative cooling and increased cloud cover associated with the storm activity. In contrast, temperatures at Danshui and Keelung were less affected by evaporative cooling during afternoon hours because of less storm activity (Fig. 5). Surface mean dewpoint temperatures at Danshui and Keelung (Fig. 12d) showed significant differences with a daily mean value of approximately 1°–2°C higher for TSA days during the morning and early afternoon.

In central Taiwan (Fig. 13a), a slight southerly component to the westerly surface winds in Tainan County was observed compared to the westerly winds in Chiayi County on TSA days. In contrast, a slight northern component to the average surface winds in Tainan was observed on non-TSA days (Fig. 13b). Further studies are required to determine if there is any significance to this small difference. The surface mean temperatures (Fig. 13c) were generally 0.5°–1°C warmer on TSA days than on non-TSA days during the morning and early afternoon. More pronounced were the higher surface mean dewpoint temperatures (Fig. 13d) on TSA days, which were about 0.5°–1.5°C higher than on non-TSA days during the morning and early afternoon. In northern and central Taiwan, the relatively warmer and more moist conditions could provide a more favorable environment for the initiation and development of moist convection, as discussed in Craven et al. (2002).

5. Conclusions

In this study, 4 yr (2005–08) of rain gauge, radar, lightning, surface, and sounding data, during the warm season (May–October), were analyzed to investigate the spatial and temporal variations of afternoon thunderstorms over Taiwan. The reflectivity climatology indicated that the maximum thunderstorm frequencies in different parts of Taiwan were all during 1500–1600 LST. The highest frequency of thunderstorms occurred along the lower mountain slopes approximately parallel to the ridges of the SMR and CMR. Similar distributions were found in CG lightning frequencies. Both the reflectivity and CG lightning climatologies showed that storm activity was strongly dependent on geographic features. More thunderstorm activity was found in central to southern Taiwan than in northern and eastern Taiwan. Although thunderstorms occurred earlier in northern Taiwan, the duration of thunderstorm activity in central to southern Taiwan was longer.

The preconvective environments associated with the occurrences of afternoon thunderstorms observed from soundings showed the mean and dewpoint temperatures in the near-surface layer were relatively warmer (+0.5°C) and moister (+1.0°C) on TSA days than on all undisturbed days. In the lower to middle troposphere, the temperature profile on non-TSA days was drier than TSA days with differences of 1.5°–3.5°C. The largest average difference was in the 750–550-hPa layer where the non-TSA days averaged 2.5°–3.5°C drier. The wind directions in the 0–3- and 3–6-km layers also indicated that relatively humid southwesterly flows in the lower to middle troposphere potentially provided more favorable storm initiation conditions because of less dry-air entrainment during the storm development phase.

Diurnal variations in local-circulation characteristics, such as sea breezes and upslope winds, dominate the boundary layer wind patterns under weak synoptic conditions. In northern Taiwan, the northwestern flow from the Danshui River valley and northeastern flow from the Keelung River valley converge during the morning and early afternoon periods. In central Taiwan, a confluence of flow patterns between Chiayi and Tainan Counties may be a contributing factor to new storm development. In addition low- and midlevel winds impinging on the mountains may have a considerable effect on storm initiation and location. Further studies are required to better understand the effects of these local circulations and topography on storm occurrence and distribution. Surface station data showed that prior to the occurrence of afternoon thunderstorms on the TSA days were generally 0.5°–1.5°C warmer and 0.5°–2°C more moist than on non-TSA days. These preconvective features may be integrated in real-time nowcasting tools using fuzzy logic approaches (e.g., Vivekanandan et al. 1999; Mueller et al. 2003; Berenguer et al. 2006) to improve operational predictions of afternoon thunderstorms in Taiwan.

Acknowledgments

The authors thank the Central Weather Bureau for providing the radar data and computer resources. Dr. Jian Zhang provided valuable comments that greatly improved the manuscript. The authors also thank two anonymous reviewers for their helpful comments on the manuscript. This research is supported by the National Science Council of Taiwan, Republic of China, under Grants 97-2625-M-052-005 and 98-2625-M-052-005.

REFERENCES

  • Adams, D. K., , and Souza E. P. , 2009: CAPE and convective events in the Southwest during the North American monsoon. Mon. Wea. Rev., 137 , 8398.

  • Akaeda, K., , Reisner J. , , and Parsons D. , 1995: The role of mesoscale and topographically induced circulations initiating a flash flood observed during the TAMEX project. Mon. Wea. Rev., 123 , 17201739.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Berenguer, M., , Sempere-Torres D. , , Corral C. , , and Sánchez-Diezma R. , 2006: A fuzzy logic technique for identifying nonprecipitating echoes in radar scans. J. Atmos. Oceanic Technol., 23 , 11571180.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Black, M. L., , Burpee R. W. , , and Marks F. D. Jr., 1996: Vertical motion characteristics of tropical cyclones determined with airborne Doppler radial velocities. J. Atmos. Sci., 53 , 18871909.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Caine, S., , Jakob C. , , Siems S. , , and May P. , 2009: Objective classification of precipitating convective regimes using a weather radar in Darwin, Australia. Mon. Wea. Rev., 137 , 15851600.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Carbone, R. E., , Tuttle J. D. , , Ahijevych D. A. , , and Trier S. B. , 2002: Inferences of predictability associated with warm season precipitation episodes. J. Atmos. Sci., 59 , 20332056.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chang, P. L., , Lin P. F. , , Jou B. J-D. , , and Zhang J. , 2009: An application of reflectivity climatology in constructing radar hybrid scans over complex terrains. J. Atmos. Oceanic Technol., 26 , 13151327.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Changnon, S. A., 2001: Damaging thunderstorm activity in the United States. Bull. Amer. Meteor. Soc., 82 , 597608.

  • Chen, C. S., , and Chen Y. L. , 2003: The rainfall characteristic of Taiwan. Mon. Wea. Rev., 131 , 13231341.

  • Chen, G. T-J., , Chou H. C. , , Chang T. C. , , and Liu C. S. , 2001: Frontal and non-frontal convection over northern Taiwan in mei-yu season (in Chinese with English abstract). Atmos. Sci., 29 , 3752.

    • Search Google Scholar
    • Export Citation
  • Chen, T. C., , Yen M. C. , , Hsieh J. C. , , and Arritt R. W. , 1999: Diurnal and seasonal variations of the rainfall measured by the Automatic Rainfall and Meteorological Telemetry System in Taiwan. Bull. Amer. Meteor. Soc., 80 , 22992312.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chen, Y. L., , and Li J. , 1995: Characteristics of surface airflow and pressure patterns over the island of Taiwan during TAMEX. Mon. Wea. Rev., 123 , 695716.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Craven, J., , Jewell R. , , and Brook H. , 2002: Comparison between observed convective cloud-base heights and lifting condensation level for two different lifted parcels. Wea. Forecasting, 17 , 885890.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Dixon, M., , and Wiener G. , 1993: TITAN: Thunderstorm Identification, Tracking, Analysis, and Nowcasting–A radar-based methodology. J. Atmos. Oceanic Technol., 10 , 785797.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Dye, J. E., , Winn W. P. , , Jones J. J. , , and Breed D. W. , 1989: The electrification of New Mexico thunderstorms. Part I: Relationship between precipitation development and the onset of electrification. J. Geophys. Res., 94 , 86438656.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Eilts, M. D., and Coauthors, 1996: Severe weather warning decision support system. Preprints, 18th Conf. on Severe Local Storms, San Francisco, CA, Amer. Meteor. Soc., 536–540.

    • Search Google Scholar
    • Export Citation
  • Fuelberg, H. E., , and Biggar D. G. , 1994: The preconvective environment of summer thunderstorms over the Florida panhandle. Wea. Forecasting, 9 , 316326.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Fulton, R. A., , Breidenbach J. P. , , Seo D. J. , , Miller D. A. , , and O’Bannon T. , 1998: The WSR-88D rainfall algorithm. Wea. Forecasting, 13 , 377395.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Gremillion, M. S., , and Orville R. E. , 1999: Thunderstorm characteristics of cloud-to-ground lightning at Kennedy Space Flight Center, Florida: A study of lightning initiation signatures as indicated by WSR-88D. Wea. Forecasting, 14 , 640649.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Heinselman, P. L., , and Schultz D. M. , 2006: Intraseasonal variability of summer storms over central Arizona during 1997 and 1999. Wea. Forecasting, 21 , 559578.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Henry, S. G., 1993: Analysis of thunderstorm lifetime as a function of size and intensity. Preprints, 26th Conf. on Radar Meteorology, Norman, OK, Amer. Meteor. Soc., 138–140.

    • Search Google Scholar
    • Export Citation
  • Hondl, K. D., , and Eilts M. D. , 1994: Doppler radar signatures of developing thunderstorms and their potential to indicate the onset of cloud-to-ground lightning. Mon. Wea. Rev., 122 , 18181836.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Huntrieser, H., , Schiesser H. H. , , Schmid W. , , and Waldvogel A. , 1997: Comparison of traditional and newly developed thunderstorm indices for Switzerland. Wea. Forecasting, 12 , 108125.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Joe, P., , Burgess D. , , Potts R. , , Keenan T. , , Stumpf G. , , and Treloar A. , 2004: The S2K severe weather detection algorithms and their performance. Wea. Forecasting, 19 , 4363.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Johns, R. H., , and Doswell C. A. III, 1992: Severe local storms forecasting. Wea. Forecasting, 7 , 588612.

  • Johnson, J. T., , Mackeen P. L. , , Witt A. , , Mitchell E. D. , , Stumpf G. J. , , Eilts M. D. , , and Thmas K. W. , 1998: The Storm Cell Identification and Tracking algorithm: An enhanced WSR-88D algorithm. Wea. Forecasting, 13 , 263276.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Johnson, R. H., , and Bresch J. F. , 1991: Diagnosed characteristics of precipitation systems over Taiwan during the May–June 1987 TAMEX. Mon. Wea. Rev., 119 , 25402557.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Jou, B. J-D., 1994: Mountain-originated mesoscale precipitation system in northern Taiwan: A case study 21 June 1991. Terr. Atmos. Oceanic Sci., 5 , 169197.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kerns, B. W. J., , Chen Y. L. , , and Chang M. Y. , 2010: The diurnal cycle of winds, rain, and clouds over Taiwan during the mei-yu, summer, and autumn rainfall regimes. Mon. Wea. Rev., 138 , 497516.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Krajewski, W. F., , and Vignal B. , 2001: Evaluation of anomalous propagation echo detection in WSR-88D data: A large sample case study. J. Atmos. Oceanic Technol., 18 , 807814.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Krehbiel, P. R., , Brook M. , , Lhermitte R. L. , , and Lennon C. L. , 1983: Lightning charge structure in thunderstorms. Proceedings in Atmospheric Electricity, L. H. Ruhnke and J. Latham, Eds., A. Deepak Publishing, 408–410.

    • Search Google Scholar
    • Export Citation
  • Kuo, J. T., , and Orville H. D. , 1973: A radar climatology of summertime convective clouds in the Black Hills. J. Appl. Meteor., 12 , 359368.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lapczak, S., and Coauthors, 1999: The Canadian National Radar Project. Preprints, 29th Int. Conf. on Radar Meteorology, Montreal, QC, Canada, Amer. Meteor. Soc., 327–330.

    • Search Google Scholar
    • Export Citation
  • Li, J., , Chen Y. L. , , and Lee W. C. , 1997: Analysis of a heavy rainfall event during TAMEX. Mon. Wea. Rev., 125 , 10601081.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lin, S. M., , and Kuo H. C. , 1996: A study of summertime afternoon convection in southern Taiwan during 1994 (in Chinese with English abstract). Atmos. Sci., 24 , 249280.

    • Search Google Scholar
    • Export Citation
  • Livingston, E. S., , Nielsen-Gammon J. W. , , and Orville R. E. , 1996: A climatology, synoptic assessment, and thermodynamic evaluation for cloud-to-ground lightning in Georgia: A study for the 1996 Summer Olympics. Bull. Amer. Meteor. Soc., 77 , 14831495.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lynn, B. H., and Coauthors, 2001: Improved simulation of Florida summer convection using the PLACE land model and a 1.5-order turbulence parameterization coupled to the Penn State–NCAR mesoscale model. Mon. Wea. Rev., 129 , 14411461.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • MacGorman, D. R., and Coauthors, 2008: TELEX: The Thunderstorm Electrification and Lightning Experiment. Bull. Amer. Meteor. Soc., 89 , 9971013.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Maddox, R. A., , Zhang J. , , Gourley J. J. , , and Howard K. W. , 2002: Weather radar coverage over the contiguous United States. Wea. Forecasting, 17 , 927934.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Mueller, C., , Saxen T. , , Roberts R. , , Wilson J. , , Betancourt T. , , Dettling S. , , Oien N. , , and Yee J. , 2003: NCAR Auto-Nowcast System. Wea. Forecasting, 18 , 545561.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Murphy, M. S., , and Konrad C. E. II, 2005: Spatial and temporal patterns of thunderstorm events that produce cloud-to-ground lightning in the interior southeastern United States. Mon. Wea. Rev., 133 , 14171430.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • O’Bannon, T., 1997: Using a ‘terrain-based’ hybrid Scan to improve WSR-88D precipitation estimates. Preprints, 28th Conf. on Radar Meteorology, Austin, TX, Amer. Meteor. Soc., 506–507.

    • Search Google Scholar
    • Export Citation
  • Overeem, A., , Holleman I. , , and Buishand A. , 2009: Derivation of a 10-year radar-based climatology of rainfall. J. Appl. Meteor. Climatol., 48 , 14481463.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Parker, M. D., , and Knievel J. C. , 2005: Do meteorologists suppress thunderstorms?: Radar-derived statistics and the behavior of moist convection. Bull. Amer. Meteor. Soc., 86 , 341358.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Reap, R. M., 1986: Evaluation of cloud-to-ground lightning data from the western United States for the 1983–84 summer seasons. J. Climate Appl. Meteor., 25 , 785799.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Reap, R. M., , and MacGorman D. R. , 1989: Cloud-to-ground lightning: Climatological characteristics and relationships to model fields, radar observations, and severe local storms. Mon. Wea. Rev., 117 , 518535.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Rickenbach, T. M., , and Rutledge S. A. , 1998: Convection in TOGA COARE: Horizontal scale, morphology, and rainfall production. J. Atmos. Sci., 55 , 27152729.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Rosenfeld, D., , Wolff D. B. , , and Atlas D. , 1993: General probability-matched relations between radar reflectivity and rain rate. J. Appl. Meteor., 32 , 5072.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Saxen, T. R., and Coauthors, 2008: The operational mesogamma-scale analysis and forecast system of the U.S. Army Test and Evaluation Command. Part IV: The White Sands Missile Range Auto-Nowcast System. J. Appl. Meteor. Climatol., 47 , 11231139.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Shafer, P. E., , and Fuelberg H. E. , 2006: A statistical procedure to forecast warm season lightning over portions of the Florida peninsula. Wea. Forecasting, 21 , 851868.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Steenburgh, W. J., , Halvorson S. F. , , and Onton D. J. , 2000: Climatology of lake-effect snowstorms of the Great Salt Lake. Mon. Wea. Rev., 128 , 709727.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Steiger, S. M., , Orville R. E. , , and Carey L. D. , 2007: Total lightning signatures of thunderstorm intensity over North Texas. Part I: Supercells. Mon. Wea. Rev., 135 , 32813302.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Steiner, M., , Houze R. A. Jr., , and Yuter S. E. , 1995: Climatological characterization of three-dimensional storm structure from operational radar and rain gauge data. J. Appl. Meteor., 34 , 19782007.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Szoke, E. J., , Mueller C. K. , , Wilson J. W. , , and Zipser E. J. , 1985: Development of convection and severe weather along outflow boundaries in northeast Colorado: Diagnosing the above-surface environment with a mobile sounding system. Preprints, 14th Conf. on Severe Local Storms, Indianapolis, IN, Amer. Meteor. Soc., 386–389.

    • Search Google Scholar
    • Export Citation
  • Tapia, A., , Smith J. A. , , and Dixon M. , 1998: Estimation of convective rainfall from lightning observations. J. Appl. Meteor., 37 , 14971509.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Toracinta, E. R., , Mohr K. I. , , Zipser E. J. , , and Orville R. E. , 1996: A comparison of WSR-88D reflectivities, SSM/I brightness temperatures, and lightning for mesoscale convective system in Texas. Part I: Radar reflectivity and lightning. J. Appl. Meteor., 35 , 902918.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Vivekanandan, J., , Zrnic D. S. , , Ellis S. M. , , Oye R. , , Ryzhkov A. V. , , and Straka J. , 1999: Cloud microphysics retrieval using S-band dual-polarization radar measurements. Bull. Amer. Meteor. Soc., 80 , 381388.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wang, C. C., , Chen G. T-J. , , and Carbone R. E. , 2005: Variability of warm-season cloud episodes over East Asia based on GMS infrared brightness temperature observations. Mon. Wea. Rev., 133 , 14781500.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Watson, A. I., , López R. E. , , and Holle R. L. , 1994: Diurnal cloud-to-ground lightning patterns in Arizona during the southwest monsoon. Mon. Wea. Rev., 122 , 17161725.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Weckwerth, T. M., 2000: The effect of small-scale moisture variability on thunderstorm initiation. Mon. Wea. Rev., 128 , 40174030.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wilson, J. W., , and Schreiber W. E. , 1986: Initiation of convective storms at radar-observed boundary-layer convergence lines. Mon. Wea. Rev., 114 , 25162536.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Yuter, S. E., , and Houze R. A. Jr., 1995: Three-dimensional kinematic and microphysical evolution of Florida cumulonimbus. Part II: Frequency distributions of vertical velocity, reflectivity, and differential reflectivity. Mon. Wea. Rev., 123 , 19411963.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zehnder, J. A., , Zhang L. , , Hansford D. , , Radzan A. , , Selover N. , , and Brown C. M. , 2006: Using digital cloud photogrammetry to characterize the onset and transition from shallow to deep convection over orography. Mon. Wea. Rev., 134 , 25272546.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zhang, J., , Howard K. , , and Gourley J. J. , 2005: Constructing three-dimensional multiple-radar reflectivity mosaics: Examples of convective storms and stratiform rain echoes. J. Atmos. Oceanic Technol., 22 , 3042.

    • Crossref
    • Search Google Scholar
    • Export Citation

Fig. 1.
Fig. 1.

Distributions of observation stations in Taiwan. Gray shades represent terrain heights. Locations of radar sites are marked with triangles and their respective abbreviations. Surface stations and rain gauges are denoted by blue plus signs and pink circles, respectively. Surface stations used in this study are labeled with the station numbers for reference.

Citation: Weather and Forecasting 26, 1; 10.1175/2010WAF2222386.1

Fig. 2.
Fig. 2.

(a) Distribution of rainfall amounts (mm) on undisturbed days during the warm seasons during 2005–08. Gauge stations with local precipitation maximum are labeled with the station numbers for reference. (b) Hourly average rainfall (mm) for all rain gauges. The dashed lines indicate the crest lines of two major mountain ranges (Fig. 1).

Citation: Weather and Forecasting 26, 1; 10.1175/2010WAF2222386.1

Fig. 3.
Fig. 3.

Frequency of occurrence (%) of (a) reflectivity >40 dBZ and (b) CG lightning during 1200–2100 LST on undisturbed days. Seven TLDS sites are denoted by plus signs.

Citation: Weather and Forecasting 26, 1; 10.1175/2010WAF2222386.1

Fig. 4.
Fig. 4.

The accumulated rainfall (mm) according to different rainfall rates (mm h−1) at rain gauges (a) C0D360 and (b) C0U710. The numbers indicate the percentages (%) of total rainfall on undisturbed days as shown in the top right-hand corner.

Citation: Weather and Forecasting 26, 1; 10.1175/2010WAF2222386.1

Fig. 5.
Fig. 5.

Frequency of occurrence (%) for reflectivity >40 dBZ at (a) 1400, (b) 1500, (c) 1600, (d) 1700, (e) 1800, and (f) 1900 LST on undisturbed days. The four inset boxes in (b) indicate the subdomains for calculating the Hovmöller diagrams and the movement of storm cells in Figs. 6 and 7.

Citation: Weather and Forecasting 26, 1; 10.1175/2010WAF2222386.1

Fig. 6.
Fig. 6.

Hovmöller diagrams of the frequency of occurrence (%) of reflectivity >40 dBZ on undisturbed days for subdomains (a) N, (c) C, (e) S, and (g) E. (b),(d),(f), and (h) As in (a),(c),(e), and (g), but for the CG lightning. In each Hovmöller diagram, the frequency in the left (right) panel is averaged across (along) the long side of the subdomain as indicated in Fig. 5b. The average topographic profile is also indicated at the top of each panel.

Citation: Weather and Forecasting 26, 1; 10.1175/2010WAF2222386.1

Fig. 7.
Fig. 7.

Histograms of the movement of storm cells identified by the SCIT algorithm between 1200 and 2100 LST on undisturbed days in the subdomains (a) N, (b) C, (c) S, and (d) E. The corresponding average speeds are plotted as the thick black lines.

Citation: Weather and Forecasting 26, 1; 10.1175/2010WAF2222386.1

Fig. 8.
Fig. 8.

Climatological CFADs of radar reflectivity for (a) RCWF data collected at 6-min intervals and (b) RCCG data collected at 10-min intervals during 1200–2100 LST on undisturbed days. The CFAD bin size is 5 dBZ and is shaded with colors at intervals of 2% (dBZ)−1 km−1.

Citation: Weather and Forecasting 26, 1; 10.1175/2010WAF2222386.1

Fig. 9.
Fig. 9.

Vertical profiles of mean temperature (solid line) and dewpoint temperature (dashed line) differences between the mean profiles from undisturbed days vs TSA (heavy line) and non-TSA (thin line) days (Table 2). The average profiles were noted at the Panchiao station (46692) in northern Taiwan taken at 0800 LST (0000 UTC).

Citation: Weather and Forecasting 26, 1; 10.1175/2010WAF2222386.1

Fig. 10.
Fig. 10.

Box-and-whiskers plot of T–Td from 1000 to 400 hPa for the TSA days (gray) and non-TSA days (white). The bottom and top of the box are the value of the first (Q1) and third (Q3) quartiles, respectively. The line in the box represents the median value. Outliers are the points that fall below Q1 − 1.5(IQR) or above Q3 + 1.5(IQR) (as the length of whiskers), where the IQR (interquartile range) is equal to the difference between Q3 and Q1.

Citation: Weather and Forecasting 26, 1; 10.1175/2010WAF2222386.1

Fig. 11.
Fig. 11.

Histograms of the frequencies of the observed wind directions between (a) 0–3 and (c) 3–6 km by radiosonde observations at Panchiao at 0800 LST (0000 UTC) on TSA days. (b),(d) As in (a),(c), but for non-TSA days. The corresponding average wind speeds are plotted with thick black lines.

Citation: Weather and Forecasting 26, 1; 10.1175/2010WAF2222386.1

Fig. 12.
Fig. 12.

Hourly average surface wind at 1200 LST on (a) TSA and (b) non-TSA days in northern Taiwan. Full-wind barbs correspond to 1 m s−1 and half barbs correspond to 0.5 m s−1. The terrain heights are also indicated with gray shading. Hourly average (c) temperature (°C) and (d) dewpoint temperature (°C) are shown for stations 46690 (Danshui), 46692 (Taipei), and 46694 (Keelung). The locations of the surface stations are indicated in Fig. 1. TSA and non-TSA days are indicated with solid and dashed lines, respectively.

Citation: Weather and Forecasting 26, 1; 10.1175/2010WAF2222386.1

Fig. 13.
Fig. 13.

As in Fig. 12, but representing central Taiwan. The locations of stations 46742 (Yungkang), 46748 (Chiayi), and 46778 (Chigu) are indicated in Fig. 1.

Citation: Weather and Forecasting 26, 1; 10.1175/2010WAF2222386.1

Table 1.

Hourly numbers of storm cells identified by the SCIT algorithm from RCWF and RCCG radar data during 1200–2100 LST on undisturbed days. The peaked number of the storm cells is in boldface.

Table 1.
Table 2.

Total days per month with (without) afternoon thunderstorms in northern (N), central (C), southern (S) and eastern (E) Taiwan during the warm seasons (May–October) of 2005–08.

Table 2.
Save