• Blumen, W., 1990: Atmospheric Processes over Complex Terrain. Meteor. Monogr. No. 45, Amer. Meteor. Soc., 323 pp.

  • Brunt, A. T., 1968: Space-time relations of cyclone rainfall in the northeast Australian region. Civil Eng. Trans. Inst. Eng. Australia, April issue, 40–46. [Available from Inst. of Eng. Australia, 11 National Circuit, Barton A.C.T. 2600, Australia.].

    • Search Google Scholar
    • Export Citation
  • Chang, C-P., , T-C. Yeh, , and J. M. Chen, 1993: Effects of terrain on the surface structure of typhoons over Taiwan. Mon. Wea. Rev., 121 , 734752.

    • Search Google Scholar
    • Export Citation
  • Chang, P-L., 2000: Analysis of evolution of landfalling typhoon circulations: A case study of Typhoon Herb (1996). Ph.D. dissertation, National Taiwan University, 158 pp.

  • Colle, B. A., 2004: Sensitivity of orographic precipitation to changing ambient conditions and terrain geometries: An idealized modeling perspective. J. Atmos. Sci., 61 , 588606.

    • Search Google Scholar
    • Export Citation
  • Colle, B. A., , and Y. Zeng, 2004: Bulk microphysical sensitivities within the MM5 for orographic precipitation. Part II: Impact of barrier width and freezing level. Mon. Wea. Rev., 132 , 28022815.

    • Search Google Scholar
    • Export Citation
  • Colle, B. A., , C. F. Mass, , and K. J. Westrick, 2000: MM5 precipitation verification over the Pacific Northwest during the 1997–99 cool seasons. Wea. Forecasting, 15 , 730744.

    • Search Google Scholar
    • Export Citation
  • Cressman, G. P., 1959: An operational objective analysis system. Mon. Wea. Rev., 87 , 367374.

  • Durran, D. R., , and J. B. Klemp, 1982: On the effects of moisture on the Brunt-Väisälä frequency. J. Atmos. Sci., 39 , 21522158.

  • Geerts, B., , G. M. Heymsfield, , L. Tian, , J. B. Halverson, , A. Guillory, , and M. I. Mejia, 2000: Hurricane Georges’s landfall in the Dominican Republic: Detailed airborne Doppler radar imagery. Bull. Amer. Meteor. Soc., 81 , 9991018.

    • Search Google Scholar
    • Export Citation
  • Georgis, J-F., , F. Roux, , M. Chong, , and S. Pradier, 2003: Triple-Doppler radar analysis of the heavy rain event observed in the Lago Maggiore region during MAP IOP 2b. Quart. J. Roy. Meteor. Soc., 129 , 495522.

    • Search Google Scholar
    • Export Citation
  • Hamuro, M., and Coauthors, 1969: Precipitation bands of Typhoon Vera in 1959 (Part 1). J. Meteor. Soc. Japan, 47 , 298308.

  • Hope, J. R., 1975: Atlantic hurricane season of 1974. Weatherwise, 28 , 1218.

  • Houze Jr, R. A., 1993: Cloud Dynamics. Academic Press, 573 pp.

  • Jiang, Q., , and R. B. Smith, 2003: Cloud timescales and orographic precipitation. J. Atmos. Sci., 60 , 15431559.

  • Lee, C-S., , and D-L. Tsai, 1995: An analysis of rainbands associated with four landfalling typhoons using CAA Doppler radar data (in Chinese with English abstract). Atmos. Sci., 23 , 209235.

    • Search Google Scholar
    • Export Citation
  • Lee, C-S., , L-R. Huang, , H-S. Shen, , and S-T. Wang, 2006: A climatology model for forecasting typhoon rainfall in Taiwan. Nat. Hazards, 37 , 87105.

    • Search Google Scholar
    • Export Citation
  • Lin, Y-L., , S. Chiao, , T-A. Wang, , M. L. Kaplan, , and R. P. Weglarz, 2001: Some common ingredients for heavy orographic rainfall. Wea. Forecasting, 16 , 633660.

    • Search Google Scholar
    • Export Citation
  • Lin, Y-L., , D. B. Ensley, , S. Chiao, , and C-Y. Huang, 2002: Orographic influences on rainfall and track deflection associated with the passage of a tropical cyclone. Mon. Wea. Rev., 130 , 29292950.

    • Search Google Scholar
    • Export Citation
  • Lin, Y-L., , S-Y. Chen, , C. M. Hill, , and C-Y. Huang, 2005a: Control parameters for the influence of a mesoscale mountain range on cyclone track continuity and deflection. J. Atmos. Sci., 62 , 18491866.

    • Search Google Scholar
    • Export Citation
  • Lin, Y-L., , H. D. Reeves, , S-Y. Chen, , and S. Chiao, 2005b: Formation mechanisms for convection over the Ligurian Sea during MAP IOP-8. Mon. Wea. Rev., 133 , 22272245.

    • Search Google Scholar
    • Export Citation
  • Mass, C. F., , and G. K. Ferber, 1990: Surface pressure perturbations produced by an isolated mesoscale topographic barrier. Part I: General characteristics and dynamics. Mon. Wea. Rev., 118 , 25792596.

    • Search Google Scholar
    • Export Citation
  • Mohr, C. G., , and L. J. Miller, 1983: CEDRIC—A software package for Cartesian space editing, synthesis, and display of radar fields under interactive control. Preprints, 21st Conf. on Radar Meteorology, Edmonton, AB, Canada, Amer. Meteor. Soc., 569–574.

  • Neiman, P. J., , F. M. Ralph, , A. B. White, , D. E. Kingsmill, , and P. O. G. Persson, 2002: The statistical relationship between upslope flow and rainfall in California’s coastal mountains: Observations during CALJET. Mon. Wea. Rev., 130 , 14681492.

    • Search Google Scholar
    • Export Citation
  • Neiman, P. J., , P. O. G. Persson, , F. M. Ralph, , D. P. Jorgensen, , A. B. White, , and D. E. Kingsmill, 2004: Modification of fronts and precipitation by coastal blocking during an intense landfalling winter storm in southern California: Observations during CALJET. Mon. Wea. Rev., 132 , 242273.

    • Search Google Scholar
    • Export Citation
  • Nettleton, L., , S. Daud, , R. Neitzel, , C. Burghart, , W. C. Lee, , and P. Hildebrand, 1993: SOLO: A program to peruse and edit radar data. Preprints, 26th Int. Conf. on Radar Meteorology, Norman, OK, Amer. Meteor. Soc., 338–339.

  • Oye, R., , C. Mueller, , and S. Smith, 1995: Software for radar translation, visualization, editing, and interpolation. Preprints, 27th Conf. on Radar Meteorology, Vail, CO, Amer. Meteor. Soc., 359–361.

  • Parrish, J. R., , R. W. Burpee, , F. D. Marks Jr., , and R. Grebe, 1982: Rainfall patterns observed by digitized radar during the landfall of Hurricane Frederic (1979). Mon. Wea. Rev., 110 , 19331944.

    • Search Google Scholar
    • Export Citation
  • Powell, M. D., 1982: The transition of the Hurricane Frederic boundary-layer wind field from the open Gulf of Mexico to landfall. Mon. Wea. Rev., 110 , 19121932.

    • Search Google Scholar
    • Export Citation
  • Ray, P. S., , C. L. Ziegler, , W. Bumgarner, , and R. J. Serafin, 1980: Single- and multiple-Doppler radar observations of tornadic storms. Mon. Wea. Rev., 108 , 16071625.

    • Search Google Scholar
    • Export Citation
  • Roe, G. H., , and M. B. Baker, 2006: Microphysical and geometrical controls on the pattern of orographic precipitation. J. Atmos. Sci., 63 , 861880.

    • Search Google Scholar
    • Export Citation
  • Roeloffzen, J. C., , W. D. Van Den Berg, , and J. Oerlemans, 1986: Frictional convergence at coastlines. Tellus, 38A , 397411.

  • Shieh, S-L., , S-T. Wang, , M-D. Cheng, , and T-C. Yeh, 1996: User’s guide (1) for typhoon forecasting in the Taiwan area (in Chinese). Res. Rep. CWB84-1M-01, 356 pp.

  • Sinclair, M. R., , D. S. Wratt, , R. D. Henderson, , and W. R. Gray, 1997: Factors affecting the distribution and spillover of precipitation in the Southern Alps of New Zealand—A case study. J. Appl. Meteor., 36 , 428442.

    • Search Google Scholar
    • Export Citation
  • Smith, R. B., 1979: The influence of mountains on the atmosphere. Advances in Geophysics, Vol. 21, Academic Press, 87–230.

  • Tuleya, R. E., , M. A. Bender, , and Y. Kurihara, 1984: A simulation study of the landfall of tropical cyclones using a movable nested-mesh model. Mon. Wea. Rev., 112 , 124136.

    • Search Google Scholar
    • Export Citation
  • Wang, S-T., 1980: Prediction of the behavior and intensity of typhoons in Taiwan and its vicinity (in Chinese). Research Rep. 018, NSC-67M-0202-01(01), National Science Council of Taiwan, 100 pp. [Available from S.-T. Wang, Central Weather Bureau, 64, Kung-Yuan Rd., Taipei 100, Taiwan.].

  • Wang, S-T., 1989: Track, intensity, structure, wind and precipitation characteristics of typhoons affecting Taiwan (in Chinese). Disaster Mitigation Research Rep. 80-73, NSC 80-04140-P052-02B, National Science Council of Taiwan, 285 pp. [Available from S.-T. Wang, Central Weather Bureau, 64, Kung-Yuan Rd., Taipei 100, Taiwan.].

  • Willoughby, H. E., , and P. G. Black, 1996: Hurricane Andrew in Florida: Dynamics of a disaster. Bull. Amer. Meteor. Soc., 77 , 543549.

  • Wu, C-C., , and Y-H. Kuo, 1999: Typhoons affecting Taiwan: Current understanding and future challenges. Bull. Amer. Meteor. Soc., 80 , 6780.

    • Search Google Scholar
    • Export Citation
  • Wu, C-C., , T-H. Yen, , Y-H. Kuo, , and W. Wang, 2002: Rainfall simulation associated with Typhoon Herb (1996) near Taiwan. Part I: The topographic effect. Wea. Forecasting, 17 , 10011015.

    • Search Google Scholar
    • Export Citation
  • Yu, C-K., , and B. F. Smull, 2000: Airborne Doppler observations of a landfalling cold front upstream of steep coastal orography. Mon. Wea. Rev., 128 , 15771603.

    • Search Google Scholar
    • Export Citation
  • Yu, C-K., , and N. A. Bond, 2002: Airborne Doppler observations of a cold front in the vicinity of Vancouver Island. Mon. Wea. Rev., 130 , 26922708.

    • Search Google Scholar
    • Export Citation
  • Yu, C-K., , D. P. Jorgensen, , and F. Roux, 2007: Multiple precipitation mechanisms over mountains observed by airborne Doppler radar during MAP IOP5. Mon. Wea. Rev., 135 , 955984.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    Topographic features of northern Taiwan. Terrain height (m MSL) is indicated by shading. Locations of Snow Mountain Range, DT, NKR, and YTR are marked. Locations of the WFS Doppler radar and the CAA Doppler radar at Taoyuan International Airport are denoted by triangles. Locations of surface observing stations, automatic meteorological observing stations, automatic rain gauge, and sounding station are denoted by squares, hollow circles, solid circles, and asterisk, respectively. The surface observing station at Keelung (KE) and the sounding station at Banciao (BA) are also marked. The inset box immediately off the northern coast of Taiwan denotes the dual-Doppler synthesis domain (45 × 15 km2) adopted in this study.

  • View in gallery

    Track of Typhoon Xangsane (2000). Typhoon center is indicated every 6 h from 0000 UTC 26 Oct to 1800 UTC 1 Nov 2000. Filled (open) typhoon symbols indicate that the maximum wind speed of the typhoon is greater (smaller) than ∼33 m s−1 (adapted from the Central Weather Bureau of Taiwan).

  • View in gallery

    The Central Weather Bureau surface analysis at 0000 UTC 1 Nov 2000. Full wind barbs correspond to 5 m s−1; half barbs correspond to 2.5 m s−1.

  • View in gallery

    Low-level PPI scan (1.4° elevation) from the WFS radar (location marked by +) at 0329 UTC 1 Nov 2000. (a) Radar reflectivity (dBZ, shading); (b) radial velocity (m s−1) with a contour interval of 10 m s−1. Terrain height in (a) is indicated by contours at 300, 600, and 1800 m while terrain height in (b) is indicated by shading (key at top of panel). In (a), two inset boxes (A and B) encompassing the DT and the northern portion of SMR indicate the area for calculating the mean radar reflectivities shown in Fig. 6.

  • View in gallery

    Time–height cross section of Doppler-derived winds averaged over the dual-Doppler synthesis domain (box in Fig. 1) during the influence of Xangsane (from 1500 UTC 31 Oct to 0600 UTC 1 Nov 2000). Wind flags correspond to 25 m s−1; full wind barbs correspond to 5 m s−1; half barbs correspond to 2.5 m s−1.

  • View in gallery

    Temporal variation of precipitation in the vicinity of northern Taiwan during 1500 UTC 31 Oct to 0600 UTC 1 Nov 2000. The low-level PPI scan (1.4° elevation) of radar reflectivity from the WFS radar within the elongated boxes (shown in Fig. 4) was averaged in a direction normal to the orientation of the box and plotted as a function of time and alongbox distance. Results for box A and B are shown in (a) and (b), respectively. The right panel of each figure shows the corresponding mean terrain height along the box. Low-level mean Doppler-derived winds (averaged below 1 km over the dual-Doppler synthesis domain indicated in Fig. 1) are also indicated (full wind barbs correspond to 5 m s−1; half barbs correspond to 2.5 m s−1).

  • View in gallery

    Ten-hour accumulated rainfall (mm) observed by rain gauge over northern Taiwan from 2000 UTC 31 Oct to 0600 UTC 1 Nov 2000. Terrain height is indicated by contours with an interval of 300 m (MSL).

  • View in gallery

    Skew T–logp plot. (a) The Banciao sounding (location in Fig. 1) taken at 0000 UTC 1 Nov 2000 and (b) the corresponding vertical profiles of potential temperature and equivalent potential temperature. In (a), full wind barbs correspond to 5 m s−1; half barbs correspond to 2.5 m s−1.

  • View in gallery

    (a) Horizontal distribution of cumulative reflectivity (contours with 150-dBZ interval) derived from the low-level PPI scan (1.4° elevation) of the WFS radar during the 10-h period from 2000 UTC 31 Oct to 0600 UTC 1 Nov 2000. Terrain height (m MSL) is indicated by shading. (b) As in (a) except showing the frequency distribution of heavy precipitation (>40 dBZ) with contour interval of 10%. Line segments D1, D2, and N mark location of vertical cross sections shown in Figs. 11 –13.

  • View in gallery

    (a) Frequency distribution of topographically forced vertical velocities (>1 m s−1) during the 10-h period from 2000 UTC 31 Oct to 0600 UTC 1 Nov 2000, with a contour interval of 30%. Shading indicates terrain height (m MSL). (b) As in (a), except that shading indicates the frequency distribution of heavy precipitation (>40 dBZ).

  • View in gallery

    Mean vertical structures of reflectivity (dBZ, shading) along D1 in Fig. 9b, obtained from different intervals of low-level oncoming flow component along the section. Shown is the oncoming flow component at (a) 10–15, (b) 15–20, (c) 20–25, (d) 25–30, and (e) 30–35 m s−1. For clarity, regions of reflectivity greater than 40 dBZ are also contoured with a 1-dBZ interval. Shading and arrow in lower portion of each panel indicate topography and mountain peak along the section, respectively. The windward (i.e., northern) side is on the left of each panel.

  • View in gallery

    As in Fig. 11, but along D2 in Fig. 9b.

  • View in gallery

    As in Fig. 11, but along N in Fig. 9b, showing the oncoming flow component at (a) 5–10, (b) 10–15, (c) 15–20, (d) 20–25, (e) 25–30, and (f) 30–35 m s−1. Note that the blank region near the right end of each panel (from X = ≈14 to X = ≈20 km) was largely influenced by the interactions between topography over the southern slopes of the northern portion of SMR (cf. Fig. 9b) and landfalling typhoon rainbands coming from the eastern coast of Taiwan. Because radar echoes in this region were much less relevant to the orographic effects associated with NKR discussed in the text, for clear illustration they have been precluded in the analysis.

  • View in gallery

    Display of the lowest-level (750 m MSL) reflectivity (dBZ, shading) of the WFS radar as a function of oncoming flow component along (a) D1, (b) D2, and (c) N in Fig. 9b. Thick line in each panel denotes the location of strongest radar reflectivity found at the given oncoming flow components. Shading and arrow in the lower portion of each panel indicates topography and mountain peak along the section, respectively.

  • View in gallery

    Distance (Dp, km) between the location of the lowest-level (750 m MSL) reflectivity maximum and the mountain peak along the vertical cross section (D1, D2, and N in Fig. 9b) as a function of the low-level oncoming flow component. The negative (positive) value of Dp denotes distance upstream (downstream) of the mountain peak. Vertical line marks location of mountain peak.

  • View in gallery

    Schematic diagram illustrating the downstream shift of hydrometeors over (a) DT and (b) NKR due to changes in upstream oncoming flow. Shading denotes the main region of heavy precipitation with darker shading representing stronger precipitation intensity. Solid (dashed) arrows indicate the trajectory of hydrometeors in the weak (strong) oncoming flow condition. Open arrows denote airflow patterns over mountains. The hD and hN represent the altitude for hydrometeors starting their descent to the ground over DT and NKR, respectively. The “ΔxD” and “ΔxN” represent the distance of downstream shift of hydrometeors over DT and NKR, respectively.

  • View in gallery

    (a) The location of four boxes (A–D) used to calculate their differential reflectivity (see details in text). Shading indicates topographic features over northern Taiwan. (b) The differences in mean reflectivity (dBZ; contours) calculated over A and B (B − A) as a function of height and low-level oncoming flow. (c) The differences in mean reflectivity (dBZ, contours) calculated over C and D (D − C) as a function of height and low-level oncoming flow. In (b) and (c), regions of positive differential reflectivity are shaded.

  • View in gallery

    As in Figs. 17b,c but showing the differences in mean reflectivity factor (Z; mm6 m−3): (a) over A and B, and (b) over C and D.

  • View in gallery

    Differential reflectivity factors (Z × 103 mm6 m−3) averaged in the lowest 1 km (MSL) as a function of low-level oncoming flow. Dashed (solid) curve denotes results calculated over A and B (C and D) shown in Fig. 17a.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 35 35 19
PDF Downloads 32 32 22

Radar Observations of Intense Orographic Precipitation Associated with Typhoon Xangsane (2000)

View More View Less
  • 1 Department of Atmospheric Sciences, Chinese Culture University, Taipei, Taiwan
© Get Permissions
Full access

Abstract

With measurements from two ground-based Doppler radars located in northern Taiwan, this study documents the detailed aspects of intense orographic precipitation associated with Typhoon Xangsane (2000) as it moved northward immediately off the eastern coast of Taiwan, bringing strong low-level northeasterly to north-northeasterly winds impinging on the mountainous northern coast. With relatively good, persistent coverage of radar echoes on both inland and upstream regions, this particular event provides a unique description of the orographic precipitation and its relationship with orographic geometry, strong upstream oncoming flow, and the precipitation inherently associated with typhoon circulations. In this case, the heaviest precipitation was observed to occur primarily over two coastal mountain barriers: Mount Da-Tun (DT) and the Nangang-Keelung Range (NKR). Barrier DT is an approximately 3D mountain barrier, and the NKR, adjacent to the southeast of DT, is a relatively lower, narrower 2D mountain range. In particular, the distinct distribution and intensity of precipitation over the two barriers were observed. Analyses of vertical cross sections passing through the major regions of heavy precipitation over DT and NKR indicate the region of low-level heavy precipitation tended to shift downstream as the low-level oncoming flow intensified, and the precipitation exhibited a deeper, wider extent and stronger intensity at stronger oncoming flow regimes. However, changes in the location of maximum precipitation over DT (NKR) were confined mainly to regions over windward slopes (near and downstream of the mountain crest). The degree of downstream shift of low-level heaviest precipitation with respect to different magnitudes of oncoming flow was relatively limited (∼8 km) over NKR, as compared with a larger downstream shift of ∼15–17 km over DT. This contrasting aspect can be understood as a consequence of a longer “lifting section” and relatively lower fall speed of hydrometeors over the windward slope of DT. In addition, the precipitation inherently associated with the typhoon circulations was found to be an important contributor to the observed variations in precipitation intensity over DT and NKR. Stronger background typhoon precipitation and a shorter downstream shift of precipitation (i.e., a quasi-stationary precipitation feature) over NKR may explain the fact of larger precipitation accumulation observed over this narrower, lower barrier.

Corresponding author address: Professor Cheng-Ku Yu, Department of Atmospheric Sciences, Chinese Culture University, 55 Hwa-Kang Road, Yang-Ming-Shan, Taipei 111, Taiwan. Email: yuku@faculty.pccu.edu.tw

Abstract

With measurements from two ground-based Doppler radars located in northern Taiwan, this study documents the detailed aspects of intense orographic precipitation associated with Typhoon Xangsane (2000) as it moved northward immediately off the eastern coast of Taiwan, bringing strong low-level northeasterly to north-northeasterly winds impinging on the mountainous northern coast. With relatively good, persistent coverage of radar echoes on both inland and upstream regions, this particular event provides a unique description of the orographic precipitation and its relationship with orographic geometry, strong upstream oncoming flow, and the precipitation inherently associated with typhoon circulations. In this case, the heaviest precipitation was observed to occur primarily over two coastal mountain barriers: Mount Da-Tun (DT) and the Nangang-Keelung Range (NKR). Barrier DT is an approximately 3D mountain barrier, and the NKR, adjacent to the southeast of DT, is a relatively lower, narrower 2D mountain range. In particular, the distinct distribution and intensity of precipitation over the two barriers were observed. Analyses of vertical cross sections passing through the major regions of heavy precipitation over DT and NKR indicate the region of low-level heavy precipitation tended to shift downstream as the low-level oncoming flow intensified, and the precipitation exhibited a deeper, wider extent and stronger intensity at stronger oncoming flow regimes. However, changes in the location of maximum precipitation over DT (NKR) were confined mainly to regions over windward slopes (near and downstream of the mountain crest). The degree of downstream shift of low-level heaviest precipitation with respect to different magnitudes of oncoming flow was relatively limited (∼8 km) over NKR, as compared with a larger downstream shift of ∼15–17 km over DT. This contrasting aspect can be understood as a consequence of a longer “lifting section” and relatively lower fall speed of hydrometeors over the windward slope of DT. In addition, the precipitation inherently associated with the typhoon circulations was found to be an important contributor to the observed variations in precipitation intensity over DT and NKR. Stronger background typhoon precipitation and a shorter downstream shift of precipitation (i.e., a quasi-stationary precipitation feature) over NKR may explain the fact of larger precipitation accumulation observed over this narrower, lower barrier.

Corresponding author address: Professor Cheng-Ku Yu, Department of Atmospheric Sciences, Chinese Culture University, 55 Hwa-Kang Road, Yang-Ming-Shan, Taipei 111, Taiwan. Email: yuku@faculty.pccu.edu.tw

1. Introduction

Orographic precipitation can occur in a wide variety of synoptic conditions. However, our knowledge on this subject is obtained mainly from previous studies of midlatitude fall season and wintertime precipitation systems as they encounter topography. Two primary factors—the dynamical interaction between the airflow and topography and its associated microphysical processes—have long been recognized as crucial in determining the intensity and location of precipitation (Smith 1979; Blumen 1990; Colle 2004). Particularly, with analyses of detailed Doppler radar measurements collected over mountains from recent field experiments, a number of observational and modeling studies have further improved our understanding on how the orographically influenced circulations modulate the precipitation associated with midlatitude synoptic disturbances (e.g., Yu and Smull 2000; Yu and Bond 2002; Neiman et al. 2004; Lin et al. 2005b; Yu et al. 2007).

In contrast to orographic precipitation associated with midlatitude weather disturbances, which has previously been largely explored, we have very little knowledge about orographic precipitation occurring in the tropical cyclone environment. Fundamentally, the thermodynamical characteristics and flow regime associated with tropical cyclones are distinctly different from those of midlatitude weather systems (e.g., synoptic fronts and cyclones). Regions in the vicinity of tropical cyclones are usually characterized by abundant moisture and extremely high winds at low levels. This implies that a relatively large Froude number (Fr = U/NH; U is the upstream wind speed, N is the Brunt–Väisälä frequency, and H is the mountain height) flow regime can easily happen, which particularly favors orographic lifting and contributes to precipitation enhancement over the windward slopes (e.g., Hamuro et al. 1969; Lin et al. 2001; Wu et al. 2002). In addition, a tropical cyclone is an approximately circular, intense vortex, and as it approaches a mountain barrier, the upstream oncoming flow in terms of its wind speed and incident angle is expected to change considerably with time. Multiple flow regimes usually occur during the approach of a tropical cyclone and substantially complicate the dynamical interaction of tropical cyclone circulations with topography (Shieh et al. 1996; Lin et al. 2005a). Furthermore, the precipitation generated by orographic forcing and usually embedded within highly variable precipitation and/or rainbands inherently associated with tropical cyclones makes this scientific subject highly challenging to investigate.

As tropical cyclones approach or move across topography, continuous torrential rains can frequently result in severe flooding (e.g., Hope 1975; Wu and Kuo 1999). Observations of landfalling tropical cyclones occurring over different geographical locations have shown a generally strong relationship between the total rainfall and local distribution of topography (Brunt 1968; Hamuro et al. 1969; Parrish et al. 1982). For tropical cyclones affecting Taiwan, the importance of orography on modulating mesoscale rainfall distributions has been well known. For example, Chang et al. (1993) used surface observations from 82 typhoons occurring during a 20-yr period to document the effects of Taiwan’s terrain on the surface features of typhoons. Their analyses showed that the location of the typhoon circulation center relative to topography was crucial in controlling precipitation patterns over Taiwan. Wang (1980, 1989) comprehensively examined various observational aspects of typhoons as they interacted with Taiwan topography and found that the precipitation distributions associated with typhoons were strongly modulated by topography, especially with an obvious trend linked to the typhoon track. Based on Wang’s results, a reasonable rainfall forecast over Taiwan could be made when the typhoon track is predicted well. Lee et al. (2006) analyzed rainfall data from conventional surface stations and automatic rain gauges for 58 typhoons affecting Taiwan during 1989–2001. Their results showed that a large average typhoon rainfall occurred generally over the mountainous regions and that the averaged rainfall amount considerably increased with surface-station elevation. Nevertheless, owing to the relatively coarse conventional observations (primarily from surface stations), these previous studies provide only a gross view of the relationship between the typhoon’s precipitation and topography, and in particular, their associated physical processes are poorly documented.

With analysis of Doppler radar measurements collected over Taiwan and other mountainous geographical regions, a few observational studies have provided detailed but limited aspects of orographic precipitation occurring in the vicinity of typhoons. For example, Lee and Tsai (1995) used a ground-based Doppler radar located in northern Taiwan to investigate the characteristics of rainbands associated with four landfalling typhoons. Their results indicated that, in addition to relatively fast-moving typhoon rainbands, occurrence of strong orographic precipitation was evident in all of their analyzed typhoon events and its embedded rainbands tended to be quasi-stationary and persistent over the mountain slopes. Stationary characteristics of orographically produced precipitation were also noted by Chang (2000) for another landfalling typhoon event in Taiwan. Geerts et al. (2000) used airborne Doppler radar to investigate Hurricane Georges making landfall on the mountainous island of Hispaniola. Their analyses suggest that the orographic lifting of boundary layer air would play an important role in contributing to the occurrence of deep convection within the hurricane’s eye and the enhancement of radar reflectivity at low levels. However, although these radar-related studies have described some interesting features of orographic precipitation associated with landfalling tropical storms, none of them have explicitly addressed the physical connection between the precipitation distribution and intensity, topographic features, and incident flow associated with tropical cyclone circulations.

Owing to the increasing capability of mesoscale models in realistically producing the detailed mesoscale structures associated with tropical cyclones, numerical simulations have been quite helpful in providing some insight into the modifications of tropical cyclones by topography. Most of the modeling work has focused on the track deflection and change in intensity as a tropical cyclone approaches or moves over topography, not on the orographically modified precipitation near tropical cyclones. Lin et al. (2002) used a nonhydrostatic mesoscale model to investigate the orographic rain accompanying Typhoon Bilis (2000) as it passed Taiwan topography. Their analyses indicate that the rainfall distribution over the mountainous regions was strongly controlled by the mechanism of orographic lifting rather than by the original rainbands associated with the typhoon. Recent numerical work by Wu et al. (2002), who simulated the landfalling of Typhoon Herb (1996) with a special emphasis on its associated rainfall distribution and intensity, suggest that the forced upslope lifting associated with the interaction of the typhoon’s circulations and Taiwan topography substantially increased the total rainfall amounts over mountains. It is clear that our knowledge obtained from the numerical studies is largely hampered by inherent uncertainties and complexities associated with parameterized processes over topography, high sensitivities of model and terrain horizontal resolution on simulated orographic rainfall (e.g., Wu et al. 2002; Colle et al. 2000), and the difficulties in model validation due to the lack of adequate observations over mountainous regions.

The primary objective of this study is to explore our knowledge of orographic precipitation occurring in the typhoon environment by analyzing Doppler radar measurements collected when Typhoon Xangsane affected Taiwan on 31 October–1 November 2000. Xangsane was one of the most unforgettable tropical cyclones invading Taiwan. In addition to Nari (2001) and Toraji (2001), severe flooding, landslides, and debris flow associated with this typhoon event caused the most serious loss of human life (64 deaths and 25 missing) in the past 30 yr. Moreover, a Singapore Airlines commercial aircraft crashed at Taoyuan International Airport while attempting to take off in the stormy weather during the influence of Xangsane, and 82 passengers were killed by this unfortunate accident. Scientifically, because this particular event has relatively good, persistent coverage of radar echoes on both inland and upstream regions, it provides a unique description of the orographic precipitation and its physical relationship with orographic geometry, strong upstream oncoming flow, and the precipitation inherently associated with typhoon circulations. In this study, observations from two ground-based Doppler radars located in northern Taiwan were used to document the detailed aspects of the precipitation distribution as Xangsane moved northward immediately off the eastern coast of Taiwan, bringing strong, low-level northeasterly to north-northeasterly winds impinging on the mountainous northern coast. How these observed spatial and temporal variations of precipitation relate to the topographic features, upstream airflow, and typhoon precipitation was the particular focus of this study.

2. Data and methodology

The primary datasets used in this study were provided by the Central Weather Bureau of Taiwan operational S-band (10 cm) Doppler radar (WSR-88D) on Wu-Fen-San (WFS) and the Civil Aeronautics Administration (CAA) operational C-band (5 cm) Doppler radar located at Taoyuan International Airport. Locations of the two radar sites are indicated in Fig. 1. Both radars provide volumetric distributions of reflectivity and radial velocity with a temporal interval of ∼6 min (for WFS radar) and ∼30 min (for CAA radar) between each volume. Details on characteristics of both WFS and CAA radars are summarized in Table 1. The WFS radar is located just ∼10 km inland from the northern coast and has longer wavelength, which can provide better data coverage and less attenuation than the CAA radar. Because the lowest plan-position indicator (PPI) scan (0.4° elevation) of the WFS radar frequently contains contamination due to mountain clutter and/or blockage of the radar beam, the low-level precipitation patterns presented in this paper are derived from the radar reflectivities from the 1.4° elevation PPI scan of the WFS radar. For this second scanning elevation, the contaminated echoes are not present over northern Taiwan (i.e., the major study region); however, they still can be found near the southern portion of Snow Mountain Range (Fig. 1) where terrain height is generally above 2000 m (mean sea level; MSL). In addition to radar data, other data sources used in this study, including routine surface and sounding observations, are indicated and summarized in Fig. 1.

The dual-Doppler synthesis from the WFS and CAA radars derived from multiple-view reflectivity and radial velocity data (Ray et al. 1980) is also applied to retrieve the 3D wind field off the northern coast of Taiwan, upstream of the barriers. The inset box in Fig. 1 marks the synthesis domain extending from the northern coastline of Taiwan to ∼15 km offshore. This derived offshore kinematic information allows us to investigate the possible physical link between the upstream oncoming flow and orographic precipitation. Owing to the inherent limitation of synthesized geometry between the two radars, we did not attempt to perform wind synthesis over Taiwan landmass in this study. A significant portion of northern Taiwan is close to the baseline zone of the two radars, where the cross-beam angles are too large or small, and it would cause substantial uncertainties and errors in the synthesized winds.

The National Center for Atmospheric Research (NCAR) “SOLO” software (Nettleton et al. 1993) is used to unfold the radial velocities and to remove sea clutter and unreasonable or incorrect values of radar reflectivity and radial velocity data. The NCAR “REORDER” software (Oye et al. 1995) is used to interpolate reflectivities and radial velocities from raw PPI scans to Cartesian coordinates with horizontal grid spacing of 1 km and vertical grid spacing of 500 m, over a volume of 45 × 15 km2 in the horizontal (box in Fig. 1) and 10 km in the vertical, with the lowest analysis level located at 500 m (MSL). At grid points where multiple reflectivity data are available, the maximum observed reflectivity value is retained to mitigate the effects of attenuation. Synthesis of the gridded radial velocities into horizontal wind fields is performed using NCAR Custom Editing and Display of Reduced Information in Cartesian Space (CEDRIC; Mohr and Miller 1983). Vertical air motions are obtained through variational adjustment of the anelastic continuity equation with boundary conditions of zero vertical motions near the surface and at echo top. Owing to a relatively higher altitude of the WFS radar site (∼766 m MSL), the detailed wind information near the surface cannot be adequately sampled by the radar. Moreover, the highest elevation of radar beams is sometimes unable to reach the actual cloud top of deep convection associated with the typhoon precipitation. These inherent uncertainties of the lower (upper) boundary condition may represent a significant source of error in deducing vertical air velocity. In view of this, only dual-Doppler synthesized horizontal winds are analyzed and presented in this paper. In this study, there are a total of 21 sets of synthesized winds derived from 2000 UTC 31 October to 0600 UTC 1 November with a time interval of ∼30 min between each wind field set.

3. Case overview

Typhoon Xangsane (2000) initially moved westward after its formation over the tropical western Pacific Ocean on 26 October 2000 (Fig. 2). It started to head north immediately west of the Philippines on 30 October 2000 and kept its northward journey over the Bashai Strait over the next two days. Xangsane did not make landfall on the landmass of Taiwan, but passed nearshore, immediately off the eastern coast during 31 October to 1 November 2000.

The synoptic surface analysis at 0000 UTC 1 November 2000 (Fig. 3) shows a continental anticyclone centered over northern China near 35°N, 110°E and a northeast–southwest-oriented stationary front located northeast of the low pressure associated with Typhoon Xangsane. During this time, wind fields in the vicinity of Taiwan were dominated by the typhoon cyclonic circulations whose pressure center was just located immediately adjacent to the eastern coast of northern Taiwan. Strong north-northeasterlies (>20 m s−1) prevailed over northern Taiwan. An elongated zone of enhanced pressure gradient oriented approximately northeast–southwest between the continental high and the typhoon’s depression was also evident. As Xangsane progressed further northeastward into the East China Sea at a later time, low-level prevailing winds over the Taiwan area were still generally from the north and were mainly influenced by the arrival of large-scale anticyclonic circulations associated with the approaching continental high.

During the passage of Xangsane, heavy rain occurred and appeared to be persistent over northern Taiwan. Figure 4 shows the 0329 UTC 1 November PPI display of radar reflectivity and radial velocity at 1.4° elevation from the WFS radar to illustrate the major precipitation and airflow patterns in the vicinity of northern Taiwan. During this time, Xangsane’s center with near-zero reflectivities was located ∼65 km offshore east of the WFS radar, and the primary precipitation associated with the typhoon circulation was confined to regions north and west of the typhoon center (Fig. 4a). The highly asymmetric distribution of typhoon precipitation was mainly due to the influence of Taiwan topography. The typhoon eyewall, with maximum reflectivities reaching above 40–45 dBZ, were also evident. The outer typhoon precipitation bands typically moved onshore and made landfall on the northern and/or eastern coast of Taiwan. Northerly winds, characterized by intense approaching radial velocities (greater than 30 m s−1), prevailed immediately upstream of the Taiwan landmass (Fig. 4b).

Some important evolving aspects of the upstream oncoming flow associated with the typhoon circulations can be clearly depicted by a sequence of dual-Doppler observations from the synthesis domain in Fig. 1. At a given analysis level, winds at each grid point within the synthesis domain are averaged to obtain a mean value to represent the upstream oncoming flow at that altitude. The standard deviations of the mean winds calculated at different analysis levels are found to be generally small (<4 m s−1) during the study period, suggesting a relatively horizontal uniform nature of the upstream oncoming flow over regions immediately off the northern coast of Taiwan. The time–height cross section of the mean dual-Doppler-derived winds reveals a general veering of low-level winds with height and an anticlockwise transition of winds from relatively weak easterlies to strong north-northeasterlies during the approach of Xangsane (Fig. 5). The low-level northerly flow off the northern coast reached its maximum intensity (∼40 m s−1) at around 0200 UTC 1 November, and afterward decreased continuously as Xangsane moved further northeastward away from northern Taiwan (cf. Fig. 2). Note that midtropospheric winds became northwesterly and/or westerly after 0400 UTC 1 November, consistent with the influence of the typhoon circulation, whose center had moved to regions ∼100 km northeast of Taiwan. In distinct contrast, northerly winds still prevailed in the low troposphere during this later period and were primarily influenced by the arrival of the continental anticyclonic circulation, as described earlier.

As shown in Fig. 4a, there was obvious evidence of orographically enhanced precipitation (maximum reflectivity >40–45 dBZ) immediately inland along the northern coast of Taiwan, namely adjacent to and over the windward slopes of Mount Da-Tun (DT) and near the northern end of Snow Mountain Range (SMR). This zone of enhanced precipitation appeared to be quasi-stationary, distinct from the generally fast-moving characteristics of typhoon rainbands. This evolving aspect can be clearly seen from a sequence of low-level radar reflectivities averaged within the elongated boxes (locations in Fig. 4a) encompassing the two mountain barriers (i.e., the DT and the northern portion of SMR) and oriented approximately parallel to the oncoming north-northeasterly flow, as shown in Fig. 6. Before about 2000 UTC 31 October, northern Taiwan was mainly influenced by the passage of outer typhoon rainbands, with heavy precipitation extending from inland to regions well offshore. There was no obvious evidence of orographically enhanced precipitation during this early period. With time, Xangsane approached northern Taiwan, and the low-level oncoming winds over this region became northeasterly after 2000 UTC and started to have an appreciable flow component perpendicular to the northern slopes of coastal mountain barriers. Precipitation was significantly enhanced immediately inland; in particular, the heaviest precipitation (>40 dBZ) occurred over the windward slopes of DT and SMR and tended to persist for several hours. The enhanced orographic precipitation started to weaken in association with the decrease in the oncoming northerly flow after 0300 UTC as Xangsane moved farther away from the northern coast of Taiwan. Persistent and intense precipitation over the two mountain barriers caused a significant amount of accumulated rainfall. The rainfall map (Fig. 7), which was produced objectively using rain gauge data with the interpolation algorithm of Cressman (1959), shows the most prominent rainfall nearby DT and the northern portion of SMR. More than 400 mm of rainfall over a 10-h period from 2000 UTC 31 October to 0600 UTC 1 November was observed over these mountainous regions.

The vertical thermodynamic characteristics over northern Taiwan can be revealed by a sounding taken at Banciao (location in Fig. 1). It should be noted that the Banciao sounding is located on the lee of DT, ∼30 km from the northern coast of Taiwan, and it would be probably less representative of thermodynamic conditions upstream of the DT and SMR. A modified sounding profile can be obtained to check this possibility if the Banciao sounding data below 1 km (i.e., close to the height of DT) is replaced by a linear interpolation using surface measurements at Keelung, located at the northern coastline (location in Fig. 1). It was found that this sounding profile (not shown) exhibited thermodynamic features highly similar to those seen in the Banciao sounding (cf. Fig. 8). Hence, the influence of leeside effects on thermodynamics did not appear significant at Banciao. The thermodynamic profile shown in Fig. 8a reveals saturated conditions throughout the low–midtroposphere. Nearly neutral convective instability was present in the lowest 2.5 km, with stable convective instability aloft (Fig. 8b). Low-level winds veered with height from northerly near the surface to easterly near the midtroposphere. These wind patterns were basically similar to those derived from dual-Doppler synthesis off the northern coast of Taiwan at around 0000 UTC 1 November (cf. Fig. 5), but their intensities, particularly at low levels, were much weaker, presumably because of the influence of relatively large surface friction over land.

The heights of coastal topography over northern Taiwan (including DT and the northern portion of SMR) are generally less than ∼1 km (MSL; cf. Fig. 1). The dry Brunt–Väisälä frequency (N) calculated below 1 km is equal to 1.1 × 10−2 s−1. Given a saturated condition in the lower troposphere (Fig. 8), the saturated Brunt–Väisälä frequency (Nm) derived by Durran and Klemp (1982) would be more suitable to approximate static stability for this case. Its value is calculated to be equal to 3.9 × 10−3 s−1. With these values, a critical magnitude of upstream wind speed for a Froude number equal to 1 in the dry and saturated conditions is found to be ∼11 and ∼4 m s−1, respectively. Given low-level oncoming northerly winds associated with Xangsane are generally stronger than 20 m s−1 during the occurrence of intense orographic precipitation (cf. Figs. 5 and 6), a Froude number greater than 1 would be easily achieved for this case. In this flow regime, oncoming winds are expected to climb over mountains, instead of flowing around (Smith 1979). On the other hand, given the saturated and convectively neutral environment in the lower troposphere (Fig. 8), the buoyant force on air parcels being displaced by orographic lifting would be rather small, implying the insignificance of mountain-wave motions (Blumen 1990). The pattern and intensity of vertical motions over mountains would be primarily determined by the forced lifting produced as oncoming winds flow over their underlying mountain slopes. The possible importance of topographically forced vertical motions on the observed orographic precipitation will be discussed further in section 4.

4. Characteristics of precipitation distribution

In northern Taiwan, most of surface observing stations are located in lowland regions, and the distribution of rain gauge stations is rather sparse, particularly over higher-terrain regions (cf. Fig. 1). The high temporal and spatial resolution of measurements from the low-level PPI scan (1.4° elevation) of the WFS radar reflectivity are analyzed to provide a detailed view of trends and extremes in precipitation intensity over both upstream and mountainous regions. As described in section 3, the occurrence of the prominent orographic precipitation was confined to between 2000 UTC 31 October and 0600 UTC 1 November, and hence subsequent analyses of radar observations will focus only on this 10-h period.

Figure 9a shows the horizontal distribution of accumulated radar reflectivity during this interval. General precipitation patterns over land seen in Fig. 9a were similar to those indicated by surface rain gauge (cf. Fig. 7) but with more detailed aspects resolved. There was a general onshore increase in the accumulated reflectivity. Two local accumulated reflectivity maxima (∼4000 dBZ) were observed over the northern slopes (i.e., the windward side) of DT, but with a sharp decrease on the lee. A localized precipitation minimum located in the low foothills immediately southwest of the barrier was also evident. As clearly seen in Fig. 6a, a localized region of relatively weak radar echoes were present on the lee slope of DT and appeared to persist for ∼10 h during the occurrence of strong oncoming northerly flow. The horizontal extent of this observed feature was approximately comparable to the half-width of the DT barrier (cf. Figs. 6a and 9a). Its appearance is consistent with the influences of topography; namely, the prevailing north-northeasterly winds flow over the windward slopes of DT and descend on its leeside to favor the evaporation of hydrometeors. Several previous studies from other geographical locations have reported the frequent occurrence of so-called rain shadow on the lee of an isolated, 3D mesoscale barrier (e.g., Mass and Ferber 1990). The previously documented rain shadow is usually characterized by relatively dry, precipitation-free features. In contrast, considerable reflectivity accumulation was still present near the local minimum of precipitation on the lee of DT, presumably due to contribution from the background typhoon precipitation.

Another accumulated reflectivity maximum with values (>4300 dBZ) even greater than those seen over DT was found along the first, northern hill of SMR, called the Nangang-Keelung Range (NKR; Fig. 1). In contrast to the precipitation patterns over DT, where peak reflectivity values were confined to the windward slopes, accumulated reflectivity with the maximum value of 4300 dBZ spillover into the lee of NKR was clearly evident. Gradually reduced precipitation was observed, however, over higher mountains and slopes further south.

To check whether the patterns of reflectivity accumulation shown in Fig. 9a can reasonably represent the realistic distribution and intensity of precipitation, the frequency distribution of strong reflectivity (>40 dBZ) during the 10-h period is also calculated for comparison and shown in Fig. 9b. The frequency distributions were highly similar to the accumulated reflectivity pattern shown in Fig. 9a. Two localized maxima with frequencies of 60%–70% were coincident with the peaks of accumulated reflectivity value (cf. Fig. 9a). The largest frequency of heavy precipitation was found over NKR and reached above 80%, suggesting precipitation over this mountain range to be more intense and stably persistent. There was also some evidence of precipitation enhancement upstream of the coastal barriers; particularly, a rapid onshore increase in either reflectivity accumulation or frequency, starting from ∼5–10 km offshore, was located along the northern coastline. However, precipitation enhancement was not obvious farther offshore, where much lower values (<30%) and relatively uniform frequency distributions were observed (Fig. 9b). Note that slightly larger frequency values (∼10%–20%) found in offshore regions northeast of Taiwan were related to the passage of rainbands associated with the inner core of the typhoon (cf. Fig. 4a). Given an unblocked flow regime for the present case, as described in section 3, upstream blocking by topography would probably not be a primary cause of the precipitation enhancement near the coast. Instead, the boundary layer convergence due to the differential surface roughness between land and ocean (e.g., Roeloffzen et al. 1986) would be more likely to provide a zone of upward vertical motions along the coast, contributing to the observed signatures. Pronounced discontinuities in low-level winds due to the change in surface roughness at the coastline have been well documented by previous modeling and observational studies of landfalling tropical cyclones (Powell 1982; Tuleya et al. 1984). The coastal frictional convergence has also been observed to play a role in forming intense convective cells near the eyewall of landfalling Hurricane Andrew (Willoughby and Black 1996).

An interesting but uncertain issue regards whether the distributions of orographically enhanced precipitation shown in Fig. 9 have some physical connection to the topographically forced vertical motions. To provide some insight on this, the topographically forced vertical motions are estimated and then compared with the precipitation patterns shown in Fig. 9. Considering the horizontal winds flowing over a mountain, the topographically forced vertical velocity is proportional to the steepness of the mountain slope along the wind direction of oncoming flow and can be approximated by the expression
i1520-0493-136-2-497-e1
where h is the terrain height; and u and υ represent the east–west and south–north flow components of upstream oncoming winds, respectively. Equation (1) has been used in many previous studies to evaluate the relative importance of orographic lifting and other convective forcings associated with synoptic and/or mesoscale systems (e.g., Lin et al. 2001; Neiman et al. 2002; Wu et al. 2002; Georgis et al. 2003). To obtain more representative magnitudes for Wterrain at different slope heights and time periods, u and υ in (1) are not a constant value but a function of the terrain height h and time t. In our calculations, u and υ values at different heights and time periods are obtained from the Doppler-derived wind profile shown in Fig. 5. Note that the patterns of topographically forced vertical velocity for each calculated time period are slightly different because of the temporal variation of upstream oncoming flow. For the convenience of comparison with Fig. 5, the frequency values of topographically forced vertical velocity are then computed during the 10-h interval of interest.

Figure 10 shows the frequency distribution of topographically forced vertical velocity (Wterrain) greater than 1 m s−1. Consistent with prevailing north-northeasterly oncoming flow, the regions of larger frequency values were generally found over the northern slopes of mountains (Fig. 10a). Two primary areas of maximum frequency of Wterrain (>90%) are located over the windward slopes of DT. These frequency maxima did not exactly coincide with the regions of maximum precipitation accumulation and frequency (Fig. 10b), but the local maxima of precipitation frequency tends to be located near and immediately downstream of these local frequency maxima of Wterrain. When considering the realistic advection effect of water droplets by ambient flow, the majority of precipitation particles reaching the ground in a position downwind of major orographic lifting are reasonable (e.g., Blumen 1990), although residence time scale for hydrometeor growth aloft remains uncertain for this case. The zone of the most enhanced precipitation over NKR generally coincided very well with the maximum frequency of Wterrain (Fig. 10b), but still with a trend of shifting slightly downstream of these frequency maxima of Wterrain. Given a nearly neutral convective instability at low levels, as shown in Fig. 8, the above results suggest that condensation due to orographic lifting (instead of release of convective instability) may have played an important role in contributing to the precipitation enhancement over the mountain slopes (Houze 1993).

There was no obvious evidence of the local maximum frequencies of Wterrain over the more distant southern mountains of SMR [e.g., Ying-Tzu Range (YTR) marked in Fig. 1] associated with the local maxima of precipitation frequency. Low correlation between the degree of orographic lifting and the precipitation intensity over these secondary, downwind mountain ranges implies the significance of other factors (such as the leeside drying and blocking of low-level moisture supply by NKR) influencing precipitation development in these regions.

5. Relation of precipitation to low-level upstream oncoming flow

In this section, we investigate how the precipitation structure and intensity over DT and NKR change as the intensities of the low-level oncoming flow change with the approach of typhoon circulations. In pursuing the investigation of these important aspects, several representative vertical cross sections of radar reflectivity across the major regions of intense orographic precipitation obtained from the different intensities of the oncoming northeasterly-to-northerly flow are analyzed. For DT, a wider, approximately 3D barrier (∼10 km in half-width), two vertical sections (D1 and D2 indicated in Fig. 9b) are selected to pass through the cores of maximum reflectivity accumulation and frequency and be oriented roughly parallel to the oncoming flow. For NKR, a narrower, approximately 2D barrier (∼5 km in half-width), 16 vertical cross sections (N indicated in Fig. 9b) running through the zone of strongest reflectivity accumulation and frequency and normal to the barrier1 are selected. Given a roughly 2D precipitation pattern over NKR, a representative vertical section is then obtained by averaging these different, respective vertical sections.

The temporal interval of the WFS radar between each volumetric data is ∼6 min, which gives a set of about 100 vertical cross sections for each selected section (i.e., D1, D2, and N) during the study period (∼10 h). Because the temporal resolution of upstream oncoming flow retrieved from the dual-Doppler synthesis (cf. Fig. 5) is ∼30 min, these Doppler-derived winds are first interpolated into a 6-min interval and the oncoming flow component parallel to the orientation of selected cross sections is then calculated. Radar reflectivities along each selected section (i.e., D1, D2, and N) observed from different time periods are averaged within an arbitrary interval (5 m s−1 used herein) of low-level mean oncoming wind component computed below the maximum height of mountains [∼1 km (MSL)] over the study region. This averaging procedure may largely mitigate the relatively rapid temporal variety of precipitation due to the passage of the individual typhoon’s rainbands so as to amplify the precipitation variations exclusively relevant to the changes in upstream oncoming flow.

Mean vertical structures of precipitation along D1 and D2 at their corresponding intervals of low-level oncoming flow component are shown in Figs. 11 and 12, respectively. It is obvious that the trends of precipitation change with increasing oncoming flow seen from these two vertical cross sections are quite similar. As the oncoming flow increased, the region of low-level heavy precipitation (>40 dBZ) tended to have an obvious shift toward the mountain peak from regions either over the lower mountain slope (Fig. 11a) or slightly upstream of the mountain (Fig. 12a). Horizontal extent of 40-dBZ reflectivity was rather limited (only <5 km in width) at weaker oncoming flow regimes (Figs. 11a,b and 12a,b). However, a much wider region (∼10 km) of heavy precipitation (>40 dBZ) could be seen when the oncoming flow was greater than 20 m s−1 (Figs. 11c–e and 12c–e). Reflectivity trough or minimum on the lee of the barrier, as similarly seen in Fig. 9a, was more evident at these stronger oncoming flow regimes, consistent with the influence of the increasing leeside drying effect. The value of low-level observed strongest reflectivity was also generally increased at stronger oncoming flow regimes. Vertical extent of heavy precipitation (>40 dBZ) was confined to the lowest 2 km (MSL) but it became deeper (with the 40-dBZ contour reaching above 2 km) at stronger oncoming flow regimes (>25 m s−1). This deepening precipitation feature would be probably related to the deepening layer of terrain-induced rising motions over the windward slope as the low-level oncoming flow intensifies (e.g., Colle 2004).

Similar to DT, there was evidence of precipitation enhancement over NKR when the oncoming flow intensified, and the most enhanced precipitation was also confined to the lowest levels (Fig. 13). The vertical extent of heavy precipitation was generally deeper at stronger oncoming flow regimes, with the 40-dBZ contour extending to a height of ≥4 km (Figs. 13c–f). In contrast to the precipitation structures observed over DT, the coverage of low-level heavy precipitation (>40 dBZ) over NKR was much wider, particularly not only over the windward slope but also on the lee side (Fig. 13). This spillover characteristic was evident even in the relatively weak oncoming flow regimes (Figs. 13a,b). The location of the low-level strongest reflectivity, as the oncoming flow increased, moved slightly downstream from the mountain crest at X ≃ 4 km (Fig. 13b) to the lee at X ≃ 8 km (Figs. 13d,e). Note that the degree of the downstream shift of low-level heaviest precipitation due to the changes in upstream oncoming flow was relatively limited, compared to a larger downstream shift observed over DT (cf. Figs. 11 and 12).

Some further aspects regarding the relationship between the upstream oncoming flow and the low-level precipitation can be obtained if reflectivities from the lowest analysis level [750 m (MSL)] along D1, D2, and N are plotted as a function of low-level oncoming flow, as shown in Fig. 14. Note that no averaging procedure was applied in the reflectivity field shown in Fig. 14 and thus a more continuous aspect regarding how the low-level precipitation evolves with the change in the intensity of oncoming flow can be seen. Over DT, a prominent downstream shift of heavy precipitation as the oncoming flow intensified was evident (Figs. 14a,b). This variation was also accompanied by a pronounced widening of the major precipitation region; the horizontal extent of heavy precipitation (>40 dBZ) was generally increased from ∼3 km at a relatively weak oncoming flow regime (<15 m s−1) to at least ∼10 km at a stronger oncoming flow regime (>20 m s−1). Over NKR, similar features of precipitation variation were found (Fig. 14c), but with relatively small changes in the location of the low-level precipitation maximum (i.e., heavy lines in Fig. 14).

Figure 15 shows the distance (denoted by Dp) of the reflectivity maximum relative to the mountain peak along the vertical cross sections as a function of low-level oncoming flow.2 It is clear that, over DT, the position of the low-level reflectivity maximum changed prominently with increasing oncoming flow, from a Dp of −15 km at ∼13 m s−1 to a Dp of ∼0 km at 32 m s−1. The variety of precipitation locations along D1 and D2 exhibited a rather similar trend and were confined to regions over the windward slopes, upstream of the mountain peak. The position of the reflectivity maximum relative to the mountain peak along N was situated in a region near and downstream of the mountain crest, entirely different from the curves describing the changes in precipitation location over DT. Particularly, a lesser distance of downstream shift (≤∼8 km) was evident as the oncoming flow intensified from 8 m s−1 to ∼30 m s−1. This observed behavior, distinct from that over DT, is schematically illustrated in Fig. 16 and qualitatively interpreted below.

Despite the complexity of microphysical processes contributing to the hydrometeor growth over topography, the distance of downstream shift of precipitation particles (Δx) due to changes in upstream oncoming flow component normal to the barrier (Δu) would be primarily related to the differential advection effect of hydrometeors by ambient wind (e.g., Sinclair et al. 1997) and can be simply approximated by the expression
i1520-0493-136-2-497-e2
where tr is the residence time of hydrometeors above the ground, h is the altitude for hydrometeors starting their descent to the ground, wair is the vertical motions of air, and υt is the terminal fall speed of hydrometeors. Because there was no prominent difference in Δu (∼20 m s−1) between DT and NKR (Fig. 15), their difference in the degree of downstream shift should be mainly affected by h and the fall speed of hydrometeors (wair + υt).

Hydrometeors formed over DT would experience a longer “lifting section” (Blumen 1990), since DT has about two times the width of NKR, and potentially, they would have more significant vertical displacement (i.e., hD > hN in Fig. 16). Note that the magnitudes of topographically forced vertical velocity calculated over the two barriers are indeed comparable (not shown). As shown in Figs. 11 –13, the intensities of observed reflectivities over NKR were generally stronger than those over DT,3 implying a larger terminal fall speed of hydrometeors over NKR. In addition, given a large Fr regime and a saturated, nearly neutral environment for this case as described in section 3, upstream air should easily flow over topography with prevailing ascending (descending) motions over the windward (lee) slope. Therefore, a relatively small (large) fall speed of hydrometeors over the windward (lee) slope of DT (NKR) can be anticipated (Fig. 16). According to (2), these characteristics described above imply a larger downstream shift of hydrometeors over DT than NKR (i.e., ΔxD > ΔxN in Fig. 16), which are consistent with our radar observations showing more (less) changes in precipitation location over the windward (lee) slope of DT (NKR; cf. Fig. 15).

6. Difference in precipitation intensity between DT and NKR

Consistent with greater precipitation accumulation and frequency over NKR than DT (cf. Fig. 9), the reflectivity structures displayed in Figs. 11 –14 indicate that precipitation intensity over NKR was obviously stronger than that over DT. There are two possible factors, both of which would be probably related to the observed difference in the precipitation intensity between DT and NKR: the orographic geometry and the influence of ambient typhoon precipitation.

It has been generally recognized that the orographic geometry has a strong influence on the airflow, thermodynamic structures, and microphysical processes, hence further altering precipitation intensity and distribution in the vicinity of mountains (e.g., Jiang and Smith 2003; Colle 2004; Colle and Zeng 2004; Roe and Baker 2006). For the present case, the heights of both DT (∼1 km MSL) and NKR (∼700 m MSL) are much lower than the melting level (0°C level corresponding to a height of ∼5.5 km; cf. Fig. 8a), which implies the dominance of warm-rain processes on the low-level precipitation enhancement over the two barriers. Besides, given that the slope steepness for DT and NKR has no significant difference, a longer “lifting section” for DT would suggest the occurrence of more condensation over its windward slope. It is also difficult to reasonably explain our observations showing relatively weak radar echoes (compared to NKR) over this barrier. Although the mountain geometry cannot be conclusively ruled out as a potential factor in influencing precipitation intensity over DT and NKR, a crucial physical link between them did not appear.

Instead, the typhoon precipitation would probably have a more direct impact in this respect. As shown in Figs. 4a and 11 –13, orographic precipitation over NKR and DT were embedded within the larger-scale coverage of precipitation associated with Xangsane. In real observations like the present case, however, it is almost impossible to separate the typhoon’s precipitation from the precipitation/cloud generated by orographically forced lifting because the two kinds of precipitation are expected to be mixed together and interact microphysically with each other. Nevertheless, an alternative way to evaluate the possible influences of typhoon precipitation on the observed orographic precipitation can be performed with calculating mean reflectivities averaged within the four boxes indicated in Fig. 17. Boxes A and B shown in Fig. 17a encompass the major area of orographically enhanced precipitation over DT and NKR (cf. Fig. 9), respectively, while reflectivity averaged over boxes C and D represents mean precipitation intensity upstream of DT and NKR, respectively. The reason box C (D) was selected to be located northeast of DT (NKR) is because it approximately encompasses the main path of typhoon rainbands influencing DT (NKR). If the inherent evolution of typhoon precipitation and/or rainbands as it moves from box C (D) to A (B), compared to the modification of precipitation by topography, is assumed to be relatively minor, the mean reflectivity calculated over box C and D can be a reasonable estimate, at least in an ensemble sense, for the strength of background precipitation associated with typhoon circulations. As a consequence, the differential reflectivity values among these select boxes may give some clues as to the influence of typhoon precipitation on the observed intensity of orographic precipitation. It is important to note that the intensity of orographic precipitation observed over DT and NKR evidently increases with increasing low-level oncoming flow (cf. Figs. 11 –13). To preclude this effect and to isolate the signals more related to background typhoon precipitation, the differential reflectivities are computed only when their associated oncoming flow components have the same magnitudes, rather than at the same observing time.

Figure 17b shows the differential values in the mean reflectivity (dBZ) between boxes A and B, which are plotted as a function of height and the intensity of upstream oncoming flow. At low levels, the precipitation intensity over NKR is generally stronger than that over DT, with positive differential reflectivities of ∼0–20 dBZ below 2 km (MSL). Positive differential reflectivities even with larger values also prevailed aloft, except for relatively small regions having negative values seen near ∼5 km (MSL) when the upstream oncoming flow lies between 20 and 24 m s−1. These results basically reflect the stronger precipitation intensity over NKR than DT, as indicated by the mean vertical structures of mountainous precipitation shown in Figs. 11 –13.

Similar to Fig. 17b, the differential values in the mean reflectivity between boxes C and D are shown in Fig. 17c. For most of the observations, the intensity of low-level background typhoon precipitation over NKR is stronger than over DT (i.e., the positive differential values). This result is not surprising because the location of NKR is closer to Xangsane’s center and thus is frequently influenced by the passage of more intense typhoon rainbands. As evident in Fig. 6, the low-level offshore reflectivities upstream of SMR were generally stronger than those upstream of DT during intense orographic precipitation. Although the values of the differential reflectivity shown in Figs. 17b,c are not exactly the same, their patterns appear roughly similar.

For comparison, the differences in the mean reflectivity factor (Z; mm6 m−3) between selected boxes in Fig. 17a are also calculated and shown in Fig. 18. Unlike the radar reflectivity (dBZ), which is in log scale, the magnitude of reflectivity factor is linearly proportional to the radar receiving power and thus its differential values would allow for a more clear and correct depiction on the observed difference of intensity in radar signals between the boxes in Fig. 17a. The differential values of reflectivity factor between boxes A and B (Fig. 18a) exhibit a clear pattern similar to those between boxes C and D (Fig. 18b). The correlation coefficient is calculated to be ∼0.8 below 4 km (MSL) and reduced to ∼0.6 at higher altitudes. Larger correlation coefficient at lower altitudes may imply a more direct and significant impact of background typhoon precipitation on the intensity of low-level precipitation developing over DT and NKR. Figure 19 further shows the mean values of the differential reflectivity factor averaged below 1 km (MSL) from Figs. 18a,b. The differential reflectivity factor between DT and NKR (dashed curve) exhibits an obvious trend of variations highly corresponding to the fluctuations in their associated differential background reflectivity factor (solid curve). These analysis results strongly suggest the influences of typhoon precipitation to be an important factor in modifying the precipitation intensity over mountains for this case.

Analyses presented above also provide an important implication. Assuming that the degree of precipitation enhancement (relative to the background value) over DT and NKR is comparable, the differential values of reflectivity factor between DT and NKR would be approximately identical to those of their background precipitation. Obviously, this is not the case. As shown in Fig. 19, the dashed curve tends to be suited generally well above the solid curve over different oncoming flow regimes. Similar tendencies can be found at higher altitudes (cf. Fig. 18). Ideally, the difference between the two curves shown in Fig. 19 (abbreviated by DIFF) can represent the differential amount of precipitation enhancement between NKR and DT. This relationship can be more easily realized with the help of the following mathematical expression:
i1520-0493-136-2-497-e3
where I represents the differential value of reflectivity factor between boxes A and B (cf. Fig. 18a), II represents the differential value of reflectivity factor between boxes C and D (cf. Fig. 18b), III represents the net amount of precipitation enhancement over NKR, and IV represents the net amount of precipitation enhancement over DT. According to (3) and Fig. 19, it is clear that the degree of precipitation enhancement over NKR was persistently more prominent than DT (i.e., III > IV; DIFF > 0), regardless of different oncoming flow magnitudes. This observed characteristic would be probably attributed to a stronger background typhoon precipitation over NKR that substantially favors the collision and coalescence processes over this barrier. However, the lack of microphysical observations in this case, unfortunately, prevents a close inspection on the relevant processes described above. In the future, the detailed kinematic and microphysical measurements over mountains plus the high-resolution mesoscale or cloud models will be required to investigate this issue, and particularly to clarify the relative importance of the mountain geometry and the background typhoon precipitation on the precipitation enhancement over topography.

7. Conclusions

With measurements from two ground-based Doppler radars located in northern Taiwan, this study has documented the detailed aspects of the intense orographic precipitation associated with Typhoon Xangsane (2000) as it moved northward immediately off the eastern coast of Taiwan and brought strong low-level northeasterly to north-northeasterly winds impinging on the mountainous northern coast. The heaviest precipitation occurred primarily over two mountain barriers: Mount Da-Tun and the Nangang-Keelung Range. The DT is an approximately 3D mountain barrier with peak mountain heights of ∼1000 m and a half-width of ∼10 km, and the NKR (i.e., the northernmost mountains of Snow Mountain Range), adjacent to the southeast of DT, is a relatively lower and narrower mountain range (a half-width of ∼5 km) oriented roughly southwest–northeast. Distinct aspects of precipitation, in terms of its distribution, intensity, and relation to upstream oncoming flow over the two mountain barriers, are observed.

The patterns of accumulated reflectivity over DT were characterized by two maxima located over the northern (i.e., windward) slopes and by a sharp decrease near the mountain peak and on the lee. A localized minimum of precipitation was located at low foothills southwest of the barrier. The precipitation patterns observed over NKR are different from those over DT. A precipitation maximum with even larger accumulated reflectivity values was observed near and downwind of the mountain crest (i.e., the spillover characteristic), followed by gradually reduced precipitation on the farther lee. Analyses also indicate that local maxima of precipitation frequency over the two barriers were located near and immediately downstream of the frequency maxima of calculated topographically forced vertical motions.

Analyses of vertical cross sections passing through the maximum of accumulated reflectivity over DT and NKR show that the most pronounced precipitation existed in the lowest analysis level (750 m MSL). Changes in precipitation location, structure, and intensity over DT and NKR with evolving upstream oncoming flow were also evident. The region of low-level heavy precipitation tended to shift downstream as the low-level oncoming flow intensified, and the precipitation exhibited a deeper, wider extent and stronger intensity at stronger oncoming flow regimes. However, the changes in the location of major heavy precipitation over DT (NKR) were confined primarily to regions over windward slopes (near and downstream of the mountain crest). The degree of downstream shift of low-level heaviest precipitation with respect to different magnitudes of oncoming winds was relatively limited (∼8 km) over NKR, in comparison with a larger downstream shift (∼15–17 km) over DT. As elaborated in section 5, this contrasting characteristic of precipitation between DT and NKR can be understood as a consequence of a longer “lifting section” and relatively lower fall speed of hydrometeors over the windward slope of DT.

In addition to the significance of the upstream oncoming flow on modulating the orographic precipitation over DT and NKR, the influence of precipitation inherently associated with the typhoon circulations was found to be an important factor in contributing to the observed variations of intensity of orographic precipitation. The differences in mean precipitation intensity calculated over DT and NKR had an obvious trend closely related to the differential intensity of their associated background typhoon precipitation. Stronger background typhoon precipitation and a shorter downstream shift of precipitation (i.e., a quasi-stationary precipitation feature) over NKR may explain the fact of larger precipitation accumulation observed over this narrower, lower barrier.

This study has shown that the complicated interactions between typhoon precipitation, orography, and the precipitation exclusively produced by orographic forcings appear to be crucial in determining the distribution and intensity of the observed orographic precipitation. Additional detailed kinematic and microphysical observations over mountains, plus the high-resolution mesoscale or cloud models, will be required to investigate these processes and to clarify the relative importance of the mountain geometry and the ambient typhoon precipitation on the precipitation enhancement over topography.

Acknowledgments

Wu-Fen-San radar and Taoyuan International Airport radar data used in this study were provided by the Central Weather Bureau and Civil Aeronautics Administration of Taiwan, respectively. We thank Dr. Nolan Atkins for his review of the manuscript and appreciate Dr. Ming-Jen Yang and anonymous reviewers for providing detailed, helpful comments that improved the manuscript. We also thank Mrs. Candace Gudmundson for help with the editing and Shao-Yu Chang for drawing the schematic figure. This study is supported by the National Science Council of Taiwan under Grants NSC 95-2111-M-034-001 and NSC 96-2111-M-034-001-MY3.

REFERENCES

  • Blumen, W., 1990: Atmospheric Processes over Complex Terrain. Meteor. Monogr. No. 45, Amer. Meteor. Soc., 323 pp.

  • Brunt, A. T., 1968: Space-time relations of cyclone rainfall in the northeast Australian region. Civil Eng. Trans. Inst. Eng. Australia, April issue, 40–46. [Available from Inst. of Eng. Australia, 11 National Circuit, Barton A.C.T. 2600, Australia.].

    • Search Google Scholar
    • Export Citation
  • Chang, C-P., , T-C. Yeh, , and J. M. Chen, 1993: Effects of terrain on the surface structure of typhoons over Taiwan. Mon. Wea. Rev., 121 , 734752.

    • Search Google Scholar
    • Export Citation
  • Chang, P-L., 2000: Analysis of evolution of landfalling typhoon circulations: A case study of Typhoon Herb (1996). Ph.D. dissertation, National Taiwan University, 158 pp.

  • Colle, B. A., 2004: Sensitivity of orographic precipitation to changing ambient conditions and terrain geometries: An idealized modeling perspective. J. Atmos. Sci., 61 , 588606.

    • Search Google Scholar
    • Export Citation
  • Colle, B. A., , and Y. Zeng, 2004: Bulk microphysical sensitivities within the MM5 for orographic precipitation. Part II: Impact of barrier width and freezing level. Mon. Wea. Rev., 132 , 28022815.

    • Search Google Scholar
    • Export Citation
  • Colle, B. A., , C. F. Mass, , and K. J. Westrick, 2000: MM5 precipitation verification over the Pacific Northwest during the 1997–99 cool seasons. Wea. Forecasting, 15 , 730744.

    • Search Google Scholar
    • Export Citation
  • Cressman, G. P., 1959: An operational objective analysis system. Mon. Wea. Rev., 87 , 367374.

  • Durran, D. R., , and J. B. Klemp, 1982: On the effects of moisture on the Brunt-Väisälä frequency. J. Atmos. Sci., 39 , 21522158.

  • Geerts, B., , G. M. Heymsfield, , L. Tian, , J. B. Halverson, , A. Guillory, , and M. I. Mejia, 2000: Hurricane Georges’s landfall in the Dominican Republic: Detailed airborne Doppler radar imagery. Bull. Amer. Meteor. Soc., 81 , 9991018.

    • Search Google Scholar
    • Export Citation
  • Georgis, J-F., , F. Roux, , M. Chong, , and S. Pradier, 2003: Triple-Doppler radar analysis of the heavy rain event observed in the Lago Maggiore region during MAP IOP 2b. Quart. J. Roy. Meteor. Soc., 129 , 495522.

    • Search Google Scholar
    • Export Citation
  • Hamuro, M., and Coauthors, 1969: Precipitation bands of Typhoon Vera in 1959 (Part 1). J. Meteor. Soc. Japan, 47 , 298308.

  • Hope, J. R., 1975: Atlantic hurricane season of 1974. Weatherwise, 28 , 1218.

  • Houze Jr, R. A., 1993: Cloud Dynamics. Academic Press, 573 pp.

  • Jiang, Q., , and R. B. Smith, 2003: Cloud timescales and orographic precipitation. J. Atmos. Sci., 60 , 15431559.

  • Lee, C-S., , and D-L. Tsai, 1995: An analysis of rainbands associated with four landfalling typhoons using CAA Doppler radar data (in Chinese with English abstract). Atmos. Sci., 23 , 209235.

    • Search Google Scholar
    • Export Citation
  • Lee, C-S., , L-R. Huang, , H-S. Shen, , and S-T. Wang, 2006: A climatology model for forecasting typhoon rainfall in Taiwan. Nat. Hazards, 37 , 87105.

    • Search Google Scholar
    • Export Citation
  • Lin, Y-L., , S. Chiao, , T-A. Wang, , M. L. Kaplan, , and R. P. Weglarz, 2001: Some common ingredients for heavy orographic rainfall. Wea. Forecasting, 16 , 633660.

    • Search Google Scholar
    • Export Citation
  • Lin, Y-L., , D. B. Ensley, , S. Chiao, , and C-Y. Huang, 2002: Orographic influences on rainfall and track deflection associated with the passage of a tropical cyclone. Mon. Wea. Rev., 130 , 29292950.

    • Search Google Scholar
    • Export Citation
  • Lin, Y-L., , S-Y. Chen, , C. M. Hill, , and C-Y. Huang, 2005a: Control parameters for the influence of a mesoscale mountain range on cyclone track continuity and deflection. J. Atmos. Sci., 62 , 18491866.

    • Search Google Scholar
    • Export Citation
  • Lin, Y-L., , H. D. Reeves, , S-Y. Chen, , and S. Chiao, 2005b: Formation mechanisms for convection over the Ligurian Sea during MAP IOP-8. Mon. Wea. Rev., 133 , 22272245.

    • Search Google Scholar
    • Export Citation
  • Mass, C. F., , and G. K. Ferber, 1990: Surface pressure perturbations produced by an isolated mesoscale topographic barrier. Part I: General characteristics and dynamics. Mon. Wea. Rev., 118 , 25792596.

    • Search Google Scholar
    • Export Citation
  • Mohr, C. G., , and L. J. Miller, 1983: CEDRIC—A software package for Cartesian space editing, synthesis, and display of radar fields under interactive control. Preprints, 21st Conf. on Radar Meteorology, Edmonton, AB, Canada, Amer. Meteor. Soc., 569–574.

  • Neiman, P. J., , F. M. Ralph, , A. B. White, , D. E. Kingsmill, , and P. O. G. Persson, 2002: The statistical relationship between upslope flow and rainfall in California’s coastal mountains: Observations during CALJET. Mon. Wea. Rev., 130 , 14681492.

    • Search Google Scholar
    • Export Citation
  • Neiman, P. J., , P. O. G. Persson, , F. M. Ralph, , D. P. Jorgensen, , A. B. White, , and D. E. Kingsmill, 2004: Modification of fronts and precipitation by coastal blocking during an intense landfalling winter storm in southern California: Observations during CALJET. Mon. Wea. Rev., 132 , 242273.

    • Search Google Scholar
    • Export Citation
  • Nettleton, L., , S. Daud, , R. Neitzel, , C. Burghart, , W. C. Lee, , and P. Hildebrand, 1993: SOLO: A program to peruse and edit radar data. Preprints, 26th Int. Conf. on Radar Meteorology, Norman, OK, Amer. Meteor. Soc., 338–339.

  • Oye, R., , C. Mueller, , and S. Smith, 1995: Software for radar translation, visualization, editing, and interpolation. Preprints, 27th Conf. on Radar Meteorology, Vail, CO, Amer. Meteor. Soc., 359–361.

  • Parrish, J. R., , R. W. Burpee, , F. D. Marks Jr., , and R. Grebe, 1982: Rainfall patterns observed by digitized radar during the landfall of Hurricane Frederic (1979). Mon. Wea. Rev., 110 , 19331944.

    • Search Google Scholar
    • Export Citation
  • Powell, M. D., 1982: The transition of the Hurricane Frederic boundary-layer wind field from the open Gulf of Mexico to landfall. Mon. Wea. Rev., 110 , 19121932.

    • Search Google Scholar
    • Export Citation
  • Ray, P. S., , C. L. Ziegler, , W. Bumgarner, , and R. J. Serafin, 1980: Single- and multiple-Doppler radar observations of tornadic storms. Mon. Wea. Rev., 108 , 16071625.

    • Search Google Scholar
    • Export Citation
  • Roe, G. H., , and M. B. Baker, 2006: Microphysical and geometrical controls on the pattern of orographic precipitation. J. Atmos. Sci., 63 , 861880.

    • Search Google Scholar
    • Export Citation
  • Roeloffzen, J. C., , W. D. Van Den Berg, , and J. Oerlemans, 1986: Frictional convergence at coastlines. Tellus, 38A , 397411.

  • Shieh, S-L., , S-T. Wang, , M-D. Cheng, , and T-C. Yeh, 1996: User’s guide (1) for typhoon forecasting in the Taiwan area (in Chinese). Res. Rep. CWB84-1M-01, 356 pp.

  • Sinclair, M. R., , D. S. Wratt, , R. D. Henderson, , and W. R. Gray, 1997: Factors affecting the distribution and spillover of precipitation in the Southern Alps of New Zealand—A case study. J. Appl. Meteor., 36 , 428442.

    • Search Google Scholar
    • Export Citation
  • Smith, R. B., 1979: The influence of mountains on the atmosphere. Advances in Geophysics, Vol. 21, Academic Press, 87–230.

  • Tuleya, R. E., , M. A. Bender, , and Y. Kurihara, 1984: A simulation study of the landfall of tropical cyclones using a movable nested-mesh model. Mon. Wea. Rev., 112 , 124136.

    • Search Google Scholar
    • Export Citation
  • Wang, S-T., 1980: Prediction of the behavior and intensity of typhoons in Taiwan and its vicinity (in Chinese). Research Rep. 018, NSC-67M-0202-01(01), National Science Council of Taiwan, 100 pp. [Available from S.-T. Wang, Central Weather Bureau, 64, Kung-Yuan Rd., Taipei 100, Taiwan.].

  • Wang, S-T., 1989: Track, intensity, structure, wind and precipitation characteristics of typhoons affecting Taiwan (in Chinese). Disaster Mitigation Research Rep. 80-73, NSC 80-04140-P052-02B, National Science Council of Taiwan, 285 pp. [Available from S.-T. Wang, Central Weather Bureau, 64, Kung-Yuan Rd., Taipei 100, Taiwan.].

  • Willoughby, H. E., , and P. G. Black, 1996: Hurricane Andrew in Florida: Dynamics of a disaster. Bull. Amer. Meteor. Soc., 77 , 543549.

  • Wu, C-C., , and Y-H. Kuo, 1999: Typhoons affecting Taiwan: Current understanding and future challenges. Bull. Amer. Meteor. Soc., 80 , 6780.

    • Search Google Scholar
    • Export Citation
  • Wu, C-C., , T-H. Yen, , Y-H. Kuo, , and W. Wang, 2002: Rainfall simulation associated with Typhoon Herb (1996) near Taiwan. Part I: The topographic effect. Wea. Forecasting, 17 , 10011015.

    • Search Google Scholar
    • Export Citation
  • Yu, C-K., , and B. F. Smull, 2000: Airborne Doppler observations of a landfalling cold front upstream of steep coastal orography. Mon. Wea. Rev., 128 , 15771603.

    • Search Google Scholar
    • Export Citation
  • Yu, C-K., , and N. A. Bond, 2002: Airborne Doppler observations of a cold front in the vicinity of Vancouver Island. Mon. Wea. Rev., 130 , 26922708.

    • Search Google Scholar
    • Export Citation
  • Yu, C-K., , D. P. Jorgensen, , and F. Roux, 2007: Multiple precipitation mechanisms over mountains observed by airborne Doppler radar during MAP IOP5. Mon. Wea. Rev., 135 , 955984.

    • Search Google Scholar
    • Export Citation
Fig. 1.
Fig. 1.

Topographic features of northern Taiwan. Terrain height (m MSL) is indicated by shading. Locations of Snow Mountain Range, DT, NKR, and YTR are marked. Locations of the WFS Doppler radar and the CAA Doppler radar at Taoyuan International Airport are denoted by triangles. Locations of surface observing stations, automatic meteorological observing stations, automatic rain gauge, and sounding station are denoted by squares, hollow circles, solid circles, and asterisk, respectively. The surface observing station at Keelung (KE) and the sounding station at Banciao (BA) are also marked. The inset box immediately off the northern coast of Taiwan denotes the dual-Doppler synthesis domain (45 × 15 km2) adopted in this study.

Citation: Monthly Weather Review 136, 2; 10.1175/2007MWR2129.1

Fig. 2.
Fig. 2.

Track of Typhoon Xangsane (2000). Typhoon center is indicated every 6 h from 0000 UTC 26 Oct to 1800 UTC 1 Nov 2000. Filled (open) typhoon symbols indicate that the maximum wind speed of the typhoon is greater (smaller) than ∼33 m s−1 (adapted from the Central Weather Bureau of Taiwan).

Citation: Monthly Weather Review 136, 2; 10.1175/2007MWR2129.1

Fig. 3.
Fig. 3.

The Central Weather Bureau surface analysis at 0000 UTC 1 Nov 2000. Full wind barbs correspond to 5 m s−1; half barbs correspond to 2.5 m s−1.

Citation: Monthly Weather Review 136, 2; 10.1175/2007MWR2129.1

Fig. 4.
Fig. 4.

Low-level PPI scan (1.4° elevation) from the WFS radar (location marked by +) at 0329 UTC 1 Nov 2000. (a) Radar reflectivity (dBZ, shading); (b) radial velocity (m s−1) with a contour interval of 10 m s−1. Terrain height in (a) is indicated by contours at 300, 600, and 1800 m while terrain height in (b) is indicated by shading (key at top of panel). In (a), two inset boxes (A and B) encompassing the DT and the northern portion of SMR indicate the area for calculating the mean radar reflectivities shown in Fig. 6.

Citation: Monthly Weather Review 136, 2; 10.1175/2007MWR2129.1

Fig. 5.
Fig. 5.

Time–height cross section of Doppler-derived winds averaged over the dual-Doppler synthesis domain (box in Fig. 1) during the influence of Xangsane (from 1500 UTC 31 Oct to 0600 UTC 1 Nov 2000). Wind flags correspond to 25 m s−1; full wind barbs correspond to 5 m s−1; half barbs correspond to 2.5 m s−1.

Citation: Monthly Weather Review 136, 2; 10.1175/2007MWR2129.1

Fig. 6.
Fig. 6.

Temporal variation of precipitation in the vicinity of northern Taiwan during 1500 UTC 31 Oct to 0600 UTC 1 Nov 2000. The low-level PPI scan (1.4° elevation) of radar reflectivity from the WFS radar within the elongated boxes (shown in Fig. 4) was averaged in a direction normal to the orientation of the box and plotted as a function of time and alongbox distance. Results for box A and B are shown in (a) and (b), respectively. The right panel of each figure shows the corresponding mean terrain height along the box. Low-level mean Doppler-derived winds (averaged below 1 km over the dual-Doppler synthesis domain indicated in Fig. 1) are also indicated (full wind barbs correspond to 5 m s−1; half barbs correspond to 2.5 m s−1).

Citation: Monthly Weather Review 136, 2; 10.1175/2007MWR2129.1

Fig. 7.
Fig. 7.

Ten-hour accumulated rainfall (mm) observed by rain gauge over northern Taiwan from 2000 UTC 31 Oct to 0600 UTC 1 Nov 2000. Terrain height is indicated by contours with an interval of 300 m (MSL).

Citation: Monthly Weather Review 136, 2; 10.1175/2007MWR2129.1

Fig. 8.
Fig. 8.

Skew T–logp plot. (a) The Banciao sounding (location in Fig. 1) taken at 0000 UTC 1 Nov 2000 and (b) the corresponding vertical profiles of potential temperature and equivalent potential temperature. In (a), full wind barbs correspond to 5 m s−1; half barbs correspond to 2.5 m s−1.

Citation: Monthly Weather Review 136, 2; 10.1175/2007MWR2129.1

Fig. 9.
Fig. 9.

(a) Horizontal distribution of cumulative reflectivity (contours with 150-dBZ interval) derived from the low-level PPI scan (1.4° elevation) of the WFS radar during the 10-h period from 2000 UTC 31 Oct to 0600 UTC 1 Nov 2000. Terrain height (m MSL) is indicated by shading. (b) As in (a) except showing the frequency distribution of heavy precipitation (>40 dBZ) with contour interval of 10%. Line segments D1, D2, and N mark location of vertical cross sections shown in Figs. 11 –13.

Citation: Monthly Weather Review 136, 2; 10.1175/2007MWR2129.1

Fig. 10.
Fig. 10.

(a) Frequency distribution of topographically forced vertical velocities (>1 m s−1) during the 10-h period from 2000 UTC 31 Oct to 0600 UTC 1 Nov 2000, with a contour interval of 30%. Shading indicates terrain height (m MSL). (b) As in (a), except that shading indicates the frequency distribution of heavy precipitation (>40 dBZ).

Citation: Monthly Weather Review 136, 2; 10.1175/2007MWR2129.1

Fig. 11.
Fig. 11.

Mean vertical structures of reflectivity (dBZ, shading) along D1 in Fig. 9b, obtained from different intervals of low-level oncoming flow component along the section. Shown is the oncoming flow component at (a) 10–15, (b) 15–20, (c) 20–25, (d) 25–30, and (e) 30–35 m s−1. For clarity, regions of reflectivity greater than 40 dBZ are also contoured with a 1-dBZ interval. Shading and arrow in lower portion of each panel indicate topography and mountain peak along the section, respectively. The windward (i.e., northern) side is on the left of each panel.

Citation: Monthly Weather Review 136, 2; 10.1175/2007MWR2129.1

Fig. 12.
Fig. 12.

As in Fig. 11, but along D2 in Fig. 9b.

Citation: Monthly Weather Review 136, 2; 10.1175/2007MWR2129.1

Fig. 13.
Fig. 13.

As in Fig. 11, but along N in Fig. 9b, showing the oncoming flow component at (a) 5–10, (b) 10–15, (c) 15–20, (d) 20–25, (e) 25–30, and (f) 30–35 m s−1. Note that the blank region near the right end of each panel (from X = ≈14 to X = ≈20 km) was largely influenced by the interactions between topography over the southern slopes of the northern portion of SMR (cf. Fig. 9b) and landfalling typhoon rainbands coming from the eastern coast of Taiwan. Because radar echoes in this region were much less relevant to the orographic effects associated with NKR discussed in the text, for clear illustration they have been precluded in the analysis.

Citation: Monthly Weather Review 136, 2; 10.1175/2007MWR2129.1

Fig. 14.
Fig. 14.

Display of the lowest-level (750 m MSL) reflectivity (dBZ, shading) of the WFS radar as a function of oncoming flow component along (a) D1, (b) D2, and (c) N in Fig. 9b. Thick line in each panel denotes the location of strongest radar reflectivity found at the given oncoming flow components. Shading and arrow in the lower portion of each panel indicates topography and mountain peak along the section, respectively.

Citation: Monthly Weather Review 136, 2; 10.1175/2007MWR2129.1

Fig. 15.
Fig. 15.

Distance (Dp, km) between the location of the lowest-level (750 m MSL) reflectivity maximum and the mountain peak along the vertical cross section (D1, D2, and N in Fig. 9b) as a function of the low-level oncoming flow component. The negative (positive) value of Dp denotes distance upstream (downstream) of the mountain peak. Vertical line marks location of mountain peak.

Citation: Monthly Weather Review 136, 2; 10.1175/2007MWR2129.1

Fig. 16.
Fig. 16.

Schematic diagram illustrating the downstream shift of hydrometeors over (a) DT and (b) NKR due to changes in upstream oncoming flow. Shading denotes the main region of heavy precipitation with darker shading representing stronger precipitation intensity. Solid (dashed) arrows indicate the trajectory of hydrometeors in the weak (strong) oncoming flow condition. Open arrows denote airflow patterns over mountains. The hD and hN represent the altitude for hydrometeors starting their descent to the ground over DT and NKR, respectively. The “ΔxD” and “ΔxN” represent the distance of downstream shift of hydrometeors over DT and NKR, respectively.

Citation: Monthly Weather Review 136, 2; 10.1175/2007MWR2129.1

Fig. 17.
Fig. 17.

(a) The location of four boxes (A–D) used to calculate their differential reflectivity (see details in text). Shading indicates topographic features over northern Taiwan. (b) The differences in mean reflectivity (dBZ; contours) calculated over A and B (B − A) as a function of height and low-level oncoming flow. (c) The differences in mean reflectivity (dBZ, contours) calculated over C and D (D − C) as a function of height and low-level oncoming flow. In (b) and (c), regions of positive differential reflectivity are shaded.

Citation: Monthly Weather Review 136, 2; 10.1175/2007MWR2129.1

Fig. 18.
Fig. 18.

As in Figs. 17b,c but showing the differences in mean reflectivity factor (Z; mm6 m−3): (a) over A and B, and (b) over C and D.

Citation: Monthly Weather Review 136, 2; 10.1175/2007MWR2129.1

Fig. 19.
Fig. 19.

Differential reflectivity factors (Z × 103 mm6 m−3) averaged in the lowest 1 km (MSL) as a function of low-level oncoming flow. Dashed (solid) curve denotes results calculated over A and B (C and D) shown in Fig. 17a.

Citation: Monthly Weather Review 136, 2; 10.1175/2007MWR2129.1

Table 1.

WFS and CAA radar characteristics.

Table 1.
1

The reason why such orientation is chosen herein is that, for an approximately 2D barrier, the direction perpendicular to the orientation of the mountain barrier is more representative of dynamical importance.

2

The upper-level [i.e., above 1 km (MSL)] oncoming flow appears to have a minor relationship to Dp, as suggested by a prominently decreasing correlation coefficient with height. Hence, our discussions are focused only on the low-level oncoming flow.

3

This observational aspect will be discussed further in section 6.

Save