• Atkinson, B. W., and M. Zhu, 2006: Coastal effects on radar propagation in atmospheric ducting conditions. Meteor. Appl., 13, 5362.

  • Babin, S. M., 1996: Surface duct height distribution for Wallop Island, Virginia, 1985–1994. J. Appl. Meteor., 35, 8693.

  • Babin, S. M., G. S. Young, and J. A. Carton, 1997: A new model of the oceanic evaporation duct. J. Appl. Meteor., 36, 193204.

  • Bean, B. R., and E. J. Dutton, 1968: Radio Meteorology. Dover, 435 pp.

  • Bebbington, D., S. Rae, J. Bech, B. Codina, and M. Picanyol, 2007: Modelling of weather radar echoes from anomalous propagation using a hybrid parabolic equation method and NWP model data. Nat. Hazards Earth Syst. Sci., 7, 391398.

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

    • Search Google Scholar
    • Export Citation
  • Brooks, I., A. Goroch, and D. Rogers, 1999: Observations of strong surface radar ducts over the Persian Gulf. J. Appl. Meteor., 38, 12931310.

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

    • Search Google Scholar
    • Export Citation
  • Chen, T. C., S. Y. Wang, M. C. Yen, A. J. Clark, and J. D. Tsay, 2010: Sudden surface warming/drying events caused by typhoon passages across Taiwan. J. Appl. Meteor. Climatol., 49, 234252.

    • Search Google Scholar
    • Export Citation
  • Doviak, R. J., and D. S. Zrnic, 1985: Siting of Doppler weather radars to shield ground targets. IEEE Trans. Antennas Propag., 33, 685689.

    • Search Google Scholar
    • Export Citation
  • Doviak, R. J., and D. S. Zrnic, 1993: Doppler Radar and Weather Observations. Academic Press, 562 pp.

  • Drechsel, S., and G. J. Mayr, 2008: Objective forecasting of foehn winds for a subgrid-scale Alpine valley. Wea. Forecasting, 23, 205218.

    • Search Google Scholar
    • Export Citation
  • Fornasiero, A., P. P. Alberoni, and J. Bech, 2006: Statistical analysis and modelling of weather radar beam propagation in the Po Valley (Italy). Nat. Hazards Earth Syst. Sci., 6, 303314.

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

    • Search Google Scholar
    • Export Citation
  • Gaffin, D. M., 2002: Unexpected warming induced by foehn winds in the lee of the Smoky Mountains. Wea. Forecasting, 17, 907915.

  • Gaffin, D. M., 2007: Foehn winds that produced large temperature differences near the southern Appalachian Mountains. Wea. Forecasting, 22, 145159.

    • Search Google Scholar
    • Export Citation
  • Gaffin, D. M., 2009: On high winds and foehn warming associated with mountain-wave events in the western foothills of the southern Appalachian Mountains. Wea. Forecasting, 24, 5375.

    • Search Google Scholar
    • Export Citation
  • Gao, J., K. Brewster, and M. Xue, 2006: A comparison of the radar ray path equations and approximations for use in radar data assimilation. Adv. Atmos. Sci., 32, 190198.

    • Search Google Scholar
    • Export Citation
  • Hubbert, J. C., M. Dixon, and S. M. Ellis, 2009: Weather radar ground clutter. Part II: Real-time identification and filtering. J. Atmos. Oceanic Technol., 26, 11811197.

    • Search Google Scholar
    • Export Citation
  • Jian, G. J., and C. C. Wu, 2008: A numerical study of the track deflection of supertyphoon Haitang (2005) prior to its landfall in Taiwan. Mon. Wea. Rev., 136, 598615.

    • Search Google Scholar
    • Export Citation
  • Lee, W. C., R. E. Carbone, and R. M. Wakimoto, 1992: The evolution and structure of a “bow-echo–microburst” event. Part I: The microburst. Mon. Wea. Rev., 120, 21882210.

    • 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
  • Lopez, P., 2009: A 5-yr 40-km-resolution global climatology of superrefraction for ground-based weather radars. J. Appl. Meteor. Climatol., 48, 89110.

    • Search Google Scholar
    • Export Citation
  • Miller, M. A., J. Verlinde, C. V. Gilbert, G. J. Lehenbauer, J. S. Tongue, and E. E. Clothiaux, 1998: Detection of nonprecipitating clouds with the WSR-88D: A theoretical and experimental survey of capabilities and limitations. Wea. Forecasting, 13, 10461062.

    • Search Google Scholar
    • Export Citation
  • Nkemdirim, L. C., 1986: Chinooks in southern Alberta: Some distinguishing nocturnal features. J. Climatol., 6, 593603.

  • Oard, M. J., 1993: A method for predicting chinook winds east of the Montana Rockies. Wea. Forecasting, 8, 166180.

  • Rogers, R. R., and M. K. Yau, 1989: A Short Course in Cloud Physics. 3rd ed. Pergamon, 293 pp.

  • Seluchi, M. E., F. A. Norte, P. Satyamurty, and S. C. Chou, 2003: Analysis of three situations of the foehn effect over the Andes (zonda wind) using the Eta–CPTEC regional model. Wea. Forecasting, 18, 481501.

    • Search Google Scholar
    • Export Citation
  • 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.

    • Search Google Scholar
    • Export Citation
  • Skolnik, M. L., 2001: Introduction to Radar Systems. 3rd ed. McGraw-Hill, 772 pp.

  • Steiner, M., and J. A. Smith, 2002: Use of three-dimensional reflectivity structure for automated detection removal of nonprecipitating echoes in radar data. J. Atmos. Oceanic Technol., 19, 673686.

    • Search Google Scholar
    • Export Citation
  • Torres, S. M., and D. S. Zrnic, 1999: Ground clutter canceling with a regression filter. J. Atmos. Oceanic Technol., 16, 13641372.

  • Turton, J. D., D. A. Bennets, and S. F. G. Farmer, 1988: An introduction to radio ducting. Meteor. Mag., 117, 245254.

  • Wakimoto, R. M., 1985: Forecasting dry microburst activity over the high plains. Mon. Wea. Rev., 113, 11311143.

  • 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
  • View in gallery
    Fig. 1.

    Distributions of observation stations in Taiwan: filled triangles—radar; filled circles—soundings; open circles—surface stations (presence of a five-digit identification number indicates that the station was used in this study); plus signs—automatic meteorological stations. The rectangle indicates the domain on which temporal variations of surface temperature in Fig. 5 were calculated. The typhoon track determined from radar reflectivity is also indicated (dots and thick solid line).

  • View in gallery
    Fig. 2.

    Distributions of (a) surface pressure and (b) streamlines (dash lines) and isotachs (solid lines) of Typhoon Krosa (2007) at 0700 UTC 6 Oct 2007. (c),(d) As in (a) and (b), but for 1000 UTC.

  • View in gallery
    Fig. 3.

    Distributions of hourly temperature (°C) at (a) 0400, (b) 0600, (c) 0800, and (d) 1000 UTC 6 Oct 2007. The location of the Hualien radar site is denoted by the plus sign. Surface stations are labeled with the station number for reference.

  • View in gallery
    Fig. 4.

    Temporal evolutions of temperature (°C; thin solid line), dewpoint (°C; dashed line), and pressure (hPa; thick solid line) at stations (a) 46749 (Taichung), (b) 46766 (Taitung), (c) 46761 (Chenggong), (d) 46699 (Hualien), and (e) 46786 (Green Island). Panels (a)–(d) show 10-min observations between 0000 and 1200 UTC 6 Oct 2007, and (e) shows hourly observations between 0000 and 0900 UTC. Full-wind barbs correspond to 5 m s−1, and half barbs correspond to 2.5 m s−1. The locations of surface stations are indicated in Fig. 1.

  • View in gallery
    Fig. 5.

    Spatiotemporal variations of temperature in the (a) horizontal and (b) vertical domains as indicated in Fig. 1. The 28°C contour is highlighted with thick lines.

  • View in gallery
    Fig. 6.

    Soundings launched at 0000 UTC 6 Oct 2007 at (a) Hualien and (b) Green Island. Full-wind barbs correspond to 5 m s−1, and half barbs correspond to 2.5 m s−1. (c) As in (a), but for 0500 UTC.

  • View in gallery
    Fig. 7.

    RCHL reflectivities at 0.5° elevation at (a) 0505, (b) 0535, (c) 0605, and (d) 0635 UTC 6 Oct 2007. Range rings of 75 and 150 km centered at RCHL are also indicated. The surface station is labeled with the station number for reference.

  • View in gallery
    Fig. 8.

    Temporal evolution of the RMSE (solid line and squares) and the correlation coefficient (dashed line and times signs) computed from the RCHL reflectivities between 0.5° and 1.4° elevations.

  • View in gallery
    Fig. 9.

    Vertical profiles of temperature (T; thick solid line) and dewpoint (Td; thick dashed line) from Hualien sounding at 0500 UTC 6 Oct 2007. The idealized inversion profiles of T and Td from A to F are used for the simulations of ray paths in Figs. 13 and 14.

  • View in gallery
    Fig. 10.

    The lower-layer modified refractivity index M calculated from Hualien sounding data at 0500 UTC 6 Oct 2007 (Fig. 6c).

  • View in gallery
    Fig. 11.

    (a) Histogram of surface duct heights calculated from Hualien sounding observations at 0000 and 1200 UTC during 2006–09. (b) As in (a), but for the surface duct depths. The duct heights and duct depths are placed in 10-m bins.

  • View in gallery
    Fig. 12.

    The radar ray paths relative to the RCHL radar for elevations from 0° to 0.5° that were derived by using the calculations of refractivity index and ray tracing that are based on the sounding profile at 0500 UTC 6 Oct 2007.

  • View in gallery
    Fig. 13.

    Similar to Fig. 12, but for the radar ray paths at elevation 0.1° from idealized vertical inversion profiles from A to F as indicated in Fig. 9.

  • View in gallery
    Fig. 14.

    As in Fig. 13, but for inversion heights (A–F in Fig. 9) linearly decreased to 0 m from the radar site to the range at (a) 100 and (b) 50 km.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 190 86 7
PDF Downloads 150 83 1

Radar Anomalous Propagation Associated with Foehn Winds Induced by Typhoon Krosa (2007)

Pao-Liang ChangCentral Weather Bureau, Taipei, Taiwan

Search for other papers by Pao-Liang Chang in
Current site
Google Scholar
PubMed
Close
and
Pin-Fang LinCentral Weather Bureau, Taipei, Taiwan

Search for other papers by Pin-Fang Lin in
Current site
Google Scholar
PubMed
Close
Full access

Abstract

In this study, unusual radar anomalous propagation (AP) phenomena associated with foehn winds induced by Typhoon Krosa (2007) were documented by using observations from radar, surface stations, and soundings. The AP echoes embedded within rainband areas and exhibited inward motions toward the radar site within 2–3 h prior to the occurrences of foehn winds at the radar site, which would interfere with the interpretation of radar data and associated downstream applications. As Typhoon Krosa appeared in the vicinity of the northeastern coast of Taiwan, foehn winds with significant subsidence warming and drying generated by downslope winds were observed in southeastern Taiwan. The foehn winds continuously moved northward within confined areas from the southeastern to eastern–central parts of Taiwan. Before the foehn winds penetrated to the surface, the subsidence warming introduced a temperature inversion layer above the surface and caused the ducting of radar beams. Analyses of refractive index and ray tracing suggested that the occurrence and evolution of the AP echoes during Typhoon Krosa were closely related to the varying inversion heights induced by downslope winds.

Corresponding author address: Dr. Pao-Liang Chang, Meteorological Satellite Center, Central Weather Bureau, 64 Gongyuan Road, Taipei 100, Taiwan. E-mail: larkdi@msc.cwb.gov.tw

Abstract

In this study, unusual radar anomalous propagation (AP) phenomena associated with foehn winds induced by Typhoon Krosa (2007) were documented by using observations from radar, surface stations, and soundings. The AP echoes embedded within rainband areas and exhibited inward motions toward the radar site within 2–3 h prior to the occurrences of foehn winds at the radar site, which would interfere with the interpretation of radar data and associated downstream applications. As Typhoon Krosa appeared in the vicinity of the northeastern coast of Taiwan, foehn winds with significant subsidence warming and drying generated by downslope winds were observed in southeastern Taiwan. The foehn winds continuously moved northward within confined areas from the southeastern to eastern–central parts of Taiwan. Before the foehn winds penetrated to the surface, the subsidence warming introduced a temperature inversion layer above the surface and caused the ducting of radar beams. Analyses of refractive index and ray tracing suggested that the occurrence and evolution of the AP echoes during Typhoon Krosa were closely related to the varying inversion heights induced by downslope winds.

Corresponding author address: Dr. Pao-Liang Chang, Meteorological Satellite Center, Central Weather Bureau, 64 Gongyuan Road, Taipei 100, Taiwan. E-mail: larkdi@msc.cwb.gov.tw

1. Introduction

The purpose of this study is to investigate the characteristics of the radar anomalous propagation (AP) phenomenon associated with foehn winds induced by Typhoon Krosa (2007). Foehn winds are typically associated with significant temperature increases and relative humidity decreases due to the adiabatic compression of downslope wind that frequently occurs on the lee sides of mountain ranges worldwide as in, for example, the chinook winds east of the Rocky Mountains (Nkemdirim 1986; Oard 1993), foehn winds near the Appalachian Mountains in the eastern United States (Gaffin 2002, 2009) and in the Central Mountain Range (CMR) and the Snow Mountain Range in Taiwan (Chen et al. 2010), “zonda” downstream of the Andes (Seluchi et al. 2003), and “bora” (or also “bura”) west of the Dinaric Alps (Drechsel and Mayr 2008).

Taiwan, a mountainous island, is characterized geographically by the CMR running across most of the island in a north-northeast–south-southwest orientation (Fig. 1). Because the average elevation stands at more than 2000 m and the highest peak is close to 4000 m in the CMR, typhoons crossing Taiwan often have a profound level of interaction with the island’s complex terrain, which then produces strong winds and heavy rainfalls that lead to substantial losses of property and human lives (e.g., Wu and Kuo 1999; Lin et al. 2002). In addition, as northwestward- or westward-moving typhoons pass across the eastern or northeastern coast of Taiwan, foehn winds frequently occur in the southeastern and northwestern regions of Taiwan because of the strong downslope winds (Shieh et al. 1996; Wu and Kuo 1999) and cause agricultural losses each year. Chen et al. (2010) analyzed 54 surface warming events over Taiwan during typhoon passages from 1961 to 2007 and found that approximately 15% (25%) of typhoon events produced surface warming in central-western (southern) Taiwan. The average warming duration was about 4–5 h. In contrast, the frequency of surface warming in eastern-central Taiwan was only about 5%, predominately because of the higher mountain tops west of this region. One of those warming events in eastern-central and southern Taiwan was induced by Typhoon Krosa.

Fig. 1.
Fig. 1.

Distributions of observation stations in Taiwan: filled triangles—radar; filled circles—soundings; open circles—surface stations (presence of a five-digit identification number indicates that the station was used in this study); plus signs—automatic meteorological stations. The rectangle indicates the domain on which temporal variations of surface temperature in Fig. 5 were calculated. The typhoon track determined from radar reflectivity is also indicated (dots and thick solid line).

Citation: Journal of Applied Meteorology and Climatology 50, 7; 10.1175/2011JAMC2619.1

During the passage of Typhoon Krosa (2007) in Taiwan (Fig. 1), pronounced foehn wind phenomena were found in the eastern-central and southeastern parts of Taiwan. Pronounced radar AP echoes were found prior to the occurrences of foehn winds at the surface in the vicinity of the radar site, where AP echoes are commonly observed in the presence of ducting conditions, the thermodynamic environments of which are often characterized by strong vertical gradients of temperature and humidity. Under ducting conditions, the electromagnetic signal propagates downward, hitting the ground and producing nonmeteorological echoes (Fornasiero et al. 2006). The ducting of even a few percent of the signal power—such as that in side lobes (Doviak and Zrnic 1993)—could result in noticeable AP echoes. Many different methods have been proposed to identify and remove AP echoes, including the signal processing of raw pulse data (e.g., Torres and Zrnic 1999; Hubbert et al. 2009) and image-processing techniques (Steiner and Smith 2002; Berenguer et al. 2006). Although these algorithms have shown success to various extents in mitigating or removing AP echoes, there are still some difficulties in accurately identifying and suppressing AP echoes embedded in precipitation areas (Fulton et al. 1998).

The current study investigates the radar AP echoes associated with the foehn wind environments induced by Typhoon Krosa—a topic that has been seldom addressed in past studies even though the characteristics of foehn winds had been well recognized. Thus far, possibly because of the incomplete radar coverage in the boundary layer adjacent to the foothills of mountain ranges, especially within hundreds of meters of the surface, the impacts of foehn winds on the radar observations have not been widely documented. In this study, the characteristics and temporal evolutions of the sudden warming and drying induced by Typhoon Krosa are analyzed by using surface and sounding data, and the beam path trajectories (Doviak and Zrnic 1993) corresponding to different thermodynamic environments are also examined. In the next section, the data and method used in this study will be described. In section 3, the foehn wind phenomenon and their associated impact on radar observations are explained in detail. Discussions are presented in section 4, followed by conclusions in section 5.

2. Data and methods

a. Data

Data gathered from the Central Weather Bureau and other government agencies in Taiwan concerning the observations collected from more than 200 automatic meteorological stations provided the observations for this study. These surface datasets were then used to analyze the surface variations during the foehn winds period. Radiosonde observations collected from station Hualien (46699) located in eastern-central Taiwan were utilized to analyze thermodynamic conditions in the atmosphere and to calculate vertical refractivity index gradients with respect to the propagation of radar signals. In addition, radiosonde observations from Green Island (46786) operated by the Weather Wing of the Chinese Air Force were also analyzed for thermodynamics of the prefoehn environments off the southeastern coast of Taiwan.

Radar data from the Hualien Doppler radar (RCHL; Fig. 1) were used to illustrate the AP phenomena and associated movements that occurred in the offshore area of southeastern and eastern-central Taiwan during the passage of Typhoon Krosa (2007). The RCHL radar is a Gematronik 1500S system with a wavelength of 10 cm (S band) and a beamwidth of 1° (Table 1). Two scan modes were adopted to fit the Weather Surveillance Radar-1988 Doppler (WSR-88D) scanning strategy, that is, volume coverage pattern 21 (Miller et al. 1998). Because RCHL is very close (about 10 km) to the steep foothills in the eastern side of the CMR, clutter from various interference returns induced directly by the topography was constantly observed. To avoid the interference returns, electromagnetic wave emissions from RCHL were turned off at lower and middle elevations in real-time operation (Chang et al. 2009).

Table 1.

The radar parameters and scanning strategy of the RCHL Gematronik 1500S radar.

Table 1.

b. Refractivity index

The radio refractive index N is commonly used to describe the propagation behavior of electromagnetic radar signals (e.g., Bean and Dutton 1968), which can be approximated by
e1
where n is refractive index of the atmosphere, P is pressure, e is the partial pressure of water vapor (hPa), and T is temperature (K). The refraction of radar beams mainly depends on the vertical gradient of refractive index (Doviak and Zrnic 1993). Because this exact gradient of N is not conveniently seen in Eq. (1), the modified radio refractivity index M (e.g., Babin et al. 1997), which includes the effect of the earth’s curvature, is generally used to diagnose the ducting conditions. Index M is defined as
e2
where a is the earth’s radius (m) and h is the height of the radar beam above sea level (MSL; m). A negative M gradient (dM/dz) will show levels of trapping in the atmosphere and could be easily distinguished as an inversion in a vertical profile, which would possibly result in AP echoes in which the radar beam propagates toward the earth’s surface, that is, ducting. In contrast, a positive M gradient will show levels of electromagnetic waves escaping the atmosphere and no trapping or escaping for zero gradient. The sounding data could be utilized to calculate the radio refractive index, and consequently the atmospheric conditions could be determined.

c. Ray tracing

Based on the equivalent earth model described in Doviak and Zrnic (1993), procedures with stepwise calculations were established to simulate the beam path trajectories according to different thermodynamic environments (Fornasiero et al. 2006; Gao et al. 2006). In this study, the ray tracing will be carried out by directly calculating the integral Eq. (3) as follows (Doviak and Zrnic 1993):
e3
where a and h are the same as in Eq. (2), r(h) is the radar measurable range, h0 is the radar height MSL (m), θe is the antenna elevation angle, and n(h) and n(h0) are the refractivity index of the atmosphere at the altitudes of the beam and radar, respectively. The procedures for the calculations of ray paths are as follows: 1) Calculate the refractivity index with 1-m increments in height starting at h0 by following Eq. (1). 2) The sign will change if the value of the term in the square brackets in the denominator in Eq. (3) is negative, and the sign of the increment of ray height will also be changed, which allows the equation to handle the ray-ducting conditions. 3) The ray path can be simulated by integrating piecewise ranges obtained from consecutive heights in Eq. (3). Previous studies had performed ray tracing using simulated data from high-resolution numerical weather prediction model data (Atkinson and Zhu 2006; Bebbington et al. 2007) and produced nonisotropic distributions of AP echoes comparable to plan position indicator (PPI) radar observations. In the current study, 10–15-m-resolution sounding data were interpolated vertically into 1-m resolution for calculating refractivity index. Because the sounding data are sparse and the focused area in the current study is relatively small, the atmosphere is assumed to be homogeneous horizontally and the ray-tracing algorithm is isotropic. Nevertheless, the results are still representative of the AP process associated with the foehn winds in the local Taiwan area, and the algorithm can be easily expanded if more high-resolution sounding data become available.

3. Foehn winds

a. Overview of Typhoon Krosa

Typhoon Krosa (2007), when located approximately 900 km east of Taiwan (not shown), reached its peak intensity at 0000 UTC 6 October, with an estimated maximum sustained wind of 51 m s−1 (10-min-averaged wind). As Typhoon Krosa approached the coastline of northeastern Taiwan, its eyewall and inner rainbands were captured by the RCHL radar. As revealed by the radar reflectivities, Typhoon Krosa’s eye exhibited a sharp southward turn and slow cyclonic loop off the eastern coast of Taiwan between 0600 and 1400 UTC 6 October (Fig. 1). This loop track, similar to that of Typhoon Haitang (2005), was possibly caused by the terrain-induced channeling effect (Jian and Wu 2008).

Figure 2a shows the distribution of the surface pressure of Typhoon Krosa at 0700 UTC 6 October 2007, at which time Typhoon Krosa’s center was located near the coastline of northeastern Taiwan, with an estimated minimum central pressure of about 945 hPa. The corresponding surface streamlines are shown in Fig. 2b, in which the flow from the western side of the middle of Taiwan could cross over the CMR because of the relatively large Froude number (Wu and Kuo 1999), which might potentially provide favorable conditions for generating the downslope adiabatic warming associated with occurrences of foehn winds in the eastern-central part of Taiwan. Three hours later (at 1000 UTC), the intensity of Typhoon Krosa weakened, with an estimated central pressure of about 958 hPa (Fig. 2c), as it moved along a north–south track off the east coast of Taiwan (Fig. 1). The streamlines show that simultaneously the flow was significantly blocked on the western-central to northern sides of the CMR (Fig. 2d), indicating that the environment is unfavorable for generating downslope winds in the eastern-central part of Taiwan.

Fig. 2.
Fig. 2.

Distributions of (a) surface pressure and (b) streamlines (dash lines) and isotachs (solid lines) of Typhoon Krosa (2007) at 0700 UTC 6 Oct 2007. (c),(d) As in (a) and (b), but for 1000 UTC.

Citation: Journal of Applied Meteorology and Climatology 50, 7; 10.1175/2011JAMC2619.1

b. Surface warming

Figure 3 shows the temperature distributions over Taiwan Island during Typhoon Krosa’s stay in the northeastern coast of Taiwan. At 0400 UTC 6 October 2007, a maximum temperature of approximately 36°C occurred near Taitung (46766), which is located in the southeastern part of Taiwan (Fig. 3a). At subsequent times, areas of temperatures greater than 34°C in southeastern Taiwan extended northward toward eastern-central Taiwan, with a maximum temperature of above 38°C at 0600 UTC (Fig. 3b). In contrast, the temperatures in other areas were of no significant variations except the slight warming with an increase of about 2°C in the Ilan Plain (Fig. 1) in northeastern Taiwan. As Typhoon Krosa moved southwestward toward the eastern coastline of Taiwan (Fig. 1), the warming areas gradually moved northward, with a maximum temperature above 39°C, while the warming mitigated in southeastern Taiwan (Figs. 3c,d). After the typhoon center moved along a north–south track off the east coastline of Taiwan (0600–0900 UTC in Fig. 1), the warming in eastern-central to southern Taiwan weakened gradually, as indicated in the temperature variations (Fig. 3d), because of the unfavorable environmental conditions (Fig. 2d) for the occurrences of foehn winds.

Fig. 3.
Fig. 3.

Distributions of hourly temperature (°C) at (a) 0400, (b) 0600, (c) 0800, and (d) 1000 UTC 6 Oct 2007. The location of the Hualien radar site is denoted by the plus sign. Surface stations are labeled with the station number for reference.

Citation: Journal of Applied Meteorology and Climatology 50, 7; 10.1175/2011JAMC2619.1

In general, warming was observed not only in southeastern Taiwan but also in eastern-central Taiwan. The warming duration was approximately 8 h (between 0300 and 1100 UTC), which differed from the finding of a duration of 4–5 h in Chen et al. (2010). The difference is possibly related to Typhoon Krosa’s loop track in the vicinity of the northeastern coast of Taiwan so that it could allow a longer duration for foehn wind’s occurrence.

c. Temporal variations

A sequence of surface variations observed at five stations in western and eastern Taiwan is shown in Fig. 4. Surface observations collected from the station Taichung (46749), located in the plain area of the western side of the CMR, show the least variations among the four stations in temperature T and dewpoint Td between 0000 and 1200 UTC 6 October 2007 (Fig. 4a), whereas stations on the lee sides of the CMR show noticeable variations. At the station Taitung (46766) (Fig. 4b), sudden warming (about 28°–36°C in T) and drying (about 5°–15°C in TTd) began at about 0300 UTC and lasted for more than 9 h. The degrees of warming and drying observed in Taitung correspond to the criteria necessary for the occurrence of foehn wind as defined in Shieh et al. (1996), which specifies that foehn wind would occur during the typhoon days if T ≥ 28°C and TTd ≥ 6°C. Two hours later (0500 UTC), foehn wind occurred at station Chenggong (46761) (Fig. 4c), located on the southeastern side of the Coastal Range, and lasted for about 6 h. The warming and drying observed in Chenggong were more significant than the ones observed in Taitung, with a maximum T and TTd of approximately 38° and 20°C, respectively, between 0700 and 0800 UTC. The foehn wind phenomenon was also found at the far northern station Hualien (46699) (Fig. 4d), yet the duration of the foehn wind phenomenon observed here lasted for only about 1.5 h—beginning at 0630 UTC and ending at approximately 0800 UTC. The warming and drying tapered off toward the offshore areas of southeastern Taiwan as shown in the observations of the station Green Island (46786) (Fig. 4e), which lasted for 5 h from 0400 to 0800 UTC, with a maximum T of 31°C and TTd of 8°C at 0600 UTC (Fig. 4e), which met the foehn wind criteria defined in Shieh et al. (1996).

Fig. 4.
Fig. 4.

Temporal evolutions of temperature (°C; thin solid line), dewpoint (°C; dashed line), and pressure (hPa; thick solid line) at stations (a) 46749 (Taichung), (b) 46766 (Taitung), (c) 46761 (Chenggong), (d) 46699 (Hualien), and (e) 46786 (Green Island). Panels (a)–(d) show 10-min observations between 0000 and 1200 UTC 6 Oct 2007, and (e) shows hourly observations between 0000 and 0900 UTC. Full-wind barbs correspond to 5 m s−1, and half barbs correspond to 2.5 m s−1. The locations of surface stations are indicated in Fig. 1.

Citation: Journal of Applied Meteorology and Climatology 50, 7; 10.1175/2011JAMC2619.1

The spatiotemporal variations of surface temperature observations are illustrated in Fig. 5 that combine stations in the analyzed domain (Fig. 1) located within 100 km south-southwest of radar site according to different latitudes and elevations. Figure 5a shows the temporal variations of temperature from south to north for station elevations below 200 m (10 stations). The warming (≥28°C) began in southern portion at 0400 UTC and extended toward the north 2 h later (0600 UTC). The warming durations were approximately 7, 5, and 3 h in the southern, middle, and northern portions of the analyzed domain, respectively (Fig. 5a). According to observations from 15 stations at different elevations between 100 and 700 m, the warming occurred at higher altitudes earlier at 0400 UTC and descended to below 200 m approximately 1–2 h later (Fig. 5b). The duration was approximately 5 h for all altitudes. The descending resulted in the lowering of the inversion layer top, which would greatly impact the distributions of AP echoes (which will be described in section 4d). Note that the temporal change of the inversion layer height also reflected horizontal variations, because the vertical structure of the inversion layer might be different within the analyzed domain. Nonetheless, the time–height plot in Fig. 5b generally provided the temporal evolution of the vertical structure of the sudden warming.

Fig. 5.
Fig. 5.

Spatiotemporal variations of temperature in the (a) horizontal and (b) vertical domains as indicated in Fig. 1. The 28°C contour is highlighted with thick lines.

Citation: Journal of Applied Meteorology and Climatology 50, 7; 10.1175/2011JAMC2619.1

d. Environmental conditions

Figure 6a shows the presurface warming sounding profiles launched at Hualien at 0000 UTC 6 October 2007. Significant subsidence warming and drying were found between near the surface layer and 3000 m, with a maximum TTd of approximately 10°C near the inversion top (~900 hPa, or 800 m). During this time, between the inversion top and 300 m, wind directions were characterized as northerly with speeds of 15–20 m s−1, whereas the wind direction below the inversion was dominated by southwesterly winds with speeds of 5–10 m s−1. Meanwhile, the Green Island sounding (46780 in Fig. 1) showed more moist conditions than in Hualien near the surface (Fig. 6b), whereas similar characteristics of subsidence warming and drying were found roughly between 300 and 3000 m. There was a maximum TTd of approximately 10°C near the inversion top (~600 m), indicating an extension of the warming and drying throughout the eastern coast and offshore area. At 0500 UTC (Fig. 6c), 1–2 h prior to the occurrences of foehn winds at Hualien, the sounding shows more remarkable warming as compared with the profiles at 0000 UTC. The TTd and T between the surface and 700 hPa reached a maximum value near the inversion top (~920 hPa, or 600 m) with values of approximately 15° and 31°C, respectively. Below 700 hPa, the wind directions veered clockwise from southerly to northwesterly with increasing height.

Fig. 6.
Fig. 6.

Soundings launched at 0000 UTC 6 Oct 2007 at (a) Hualien and (b) Green Island. Full-wind barbs correspond to 5 m s−1, and half barbs correspond to 2.5 m s−1. (c) As in (a), but for 0500 UTC.

Citation: Journal of Applied Meteorology and Climatology 50, 7; 10.1175/2011JAMC2619.1

e. Radar observations

The radar reflectivities of 0.5° from the RCHL radar during the period of foehn wind occurrence in eastern Taiwan are shown in Fig. 7. No reflectivity was found west of the RCHL radar site because no electromagnetic wave was transmitted at lower elevations, as mentioned in section 2a. A significant beam blockage that was due to tall buildings in the vicinity of the radar site was found to the northeast of the radar. A near-circular typhoon eye, with a radius of approximately 25 km, was apparent at a range of about 100 km to the northeast of the radar site at 0505 UTC (Fig. 7a). Relatively strong and noisy echoes with a maximum value of greater than 55 dBZ (A in Fig. 7a) were found to the south-southeast of the radar site at a range of about 100–150 km and propagated northward at 0535 UTC (A in Fig. 7b). Similar strong and noisy echoes were found to the southeast of the radar site at a range of about 75 km, with a maximum reflectivity of about 65 dBZ (B in Fig. 7b). Corresponding Doppler velocities (not shown) within the areas of these echoes showed values that were very close to zero, indicating that the features were likely AP phenomena. These AP echoes would interfere with the interpretation of weather radar data and associated downstream applications such as quantitative precipitation estimation (Fulton et al. 1998).

Fig. 7.
Fig. 7.

RCHL reflectivities at 0.5° elevation at (a) 0505, (b) 0535, (c) 0605, and (d) 0635 UTC 6 Oct 2007. Range rings of 75 and 150 km centered at RCHL are also indicated. The surface station is labeled with the station number for reference.

Citation: Journal of Applied Meteorology and Climatology 50, 7; 10.1175/2011JAMC2619.1

When Typhoon Krosa’s center moved along a north–south track off the east coast of Taiwan, the noisy echoes—characterized by an arc shape that could be observed on radar PPI observation under a horizontally uniform ducting environment (Fornasiero et al. 2006)—kept moving toward the radar site at ranges between 50 and 75 km (Fig. 7c). At 0635 UTC (Fig. 7d), the noisy echoes exhibited a decrease in both their intensity and the range size.

In tracking the history of the AP echoes, it was found that they initially emerged as a small area to the south of the radar site at a range of ~150 km at about 0355 UTC (not shown). The AP echoes got stronger and moved more northward, and they were located at a range of about 120–150 km at 0435 UTC (not shown). The history of the AP echoes was further investigated by examining the root-mean-square error (RMSE) and correlation coefficients of reflectivities between the lowest two elevations—0.5° and 1.4° in the southeast quadrant of the RCHL radar domain. The time series of RMSE and correlation coefficients are summarized in Fig. 8. The RMSE and correlation coefficients ranged from 1.9 to 5.8 dB and from 0.87 to 0.98, respectively. Between 0525 and 0725 UTC, the noteworthy increase in the RMSE (>3.0 dB) and decrease in the correlation coefficients (<0.95) indicated a lack of coherences between radar echoes on the first and second tilts—a characteristic of AP echoes. This was consistent with the occurrences of most significant AP echoes as shown in Fig. 7.

Fig. 8.
Fig. 8.

Temporal evolution of the RMSE (solid line and squares) and the correlation coefficient (dashed line and times signs) computed from the RCHL reflectivities between 0.5° and 1.4° elevations.

Citation: Journal of Applied Meteorology and Climatology 50, 7; 10.1175/2011JAMC2619.1

4. Discussion

As described in sections 3c and 3d, significant subsidence warming and drying were present below 3000 m (~700 hPa) in altitude and 2–3 h prior to the foehn wind occurrence at the radar site, at which time the radar AP echoes embedded within Typhoon Krosa’s rainband areas were detected and were found to be gradually moving toward the radar site. After foehn winds occurred at the radar site (~0630 UTC), the AP phenomena mitigated significantly. To investigate the environments suitable for inducing the occurrences of AP phenomena, the thermodynamic characteristics of subsidence warming and drying and their possible connections to foehn winds are examined. In addition, for the purpose of describing the possible ray paths under the environments of subsidence warming and drying, a part of this study will be devoted to analyzing the ray tracing by using the sounding observations. Also, the idealized profiles of T and Td that are based on various inversion heights will be assumed to simulate the ducting conditions and the associated characteristics and movements of AP echoes in radar observations.

a. Subsidence warming and drying

The sounding profiles at 0500 UTC 6 October 2007 shown in Fig. 6c indicated a layer of subsidence warming and drying below 3000 m except near the surface layer that was likely caused by downslope winds from westerly to northwesterly between 850 and 700 hPa. The downslope winds did not penetrate into the surface layer at station RCHL because a stable layer near the ground was strong enough to inhibit the penetration (Gaffin 2007) at this time. Figure 9 shows T and Td profiles below 1500 m in altitude at 0500 UTC 6 October 2007, where the lapse rate of T was nearly constant at 7.5°C km−1 between 600 and 3000 m, which is less than the dry-adiabatic lapse rate (9.8°C km−1). The smaller lapse rate was likely due to diabatic processes of evaporations or mixing during the descending of air parcels. When the air parcels descend in the environment, an increasing rate of approximately 1 g kg−1 km−1 in mixing ratio with decreasing height from the Td profile may produce an estimated cooling rate of about 2.2°C km−1 by using the method of a wet-bulb process as described in Rogers and Yau (1989). The net lapse rate of T stands at about 7.6°C km−1, which is approximately the estimated lapse rate of 7.5°C km−1 as estimated from the T profile between 600 and 3000 m, as shown in Fig. 9.

Fig. 9.
Fig. 9.

Vertical profiles of temperature (T; thick solid line) and dewpoint (Td; thick dashed line) from Hualien sounding at 0500 UTC 6 Oct 2007. The idealized inversion profiles of T and Td from A to F are used for the simulations of ray paths in Figs. 13 and 14.

Citation: Journal of Applied Meteorology and Climatology 50, 7; 10.1175/2011JAMC2619.1

The strong subsidence warming and drying associated with downdraft are frequently found in microbursts where the lapse rates approximately follow the dry-adiabatic process with a constant mixing ratio (Wakimoto 1985; Lee et al. 1992). In the current study, on the basis of the lapse rates between 600 and 3000 m (Fig. 9), the expected T and Td on the ground would come to 36° and 19°C, respectively, if a similar thermodynamic process is assumed in the foehn wind environments. The values of T and Td were close to the actual observations at station Hualien (Fig. 4d)—36.7°C in T and 19.2°C in Td—that were taken when the most significant foehn winds occurred at 0730 UTC. Although horizontal thermal advection also might contribute to the warming, the weak surface wind (5–10 m s−1), observed at stations Taitung (Fig. 4b) and Chenggong (Fig. 4c) in southeastern Taiwan was unlikely to bring the sudden warming within 1–2 h as shown in Figs. 3a,b. Therefore, on the basis of the above discussion, the adiabatic warming from downslope flows can be concluded to be the primary cause leading to the occurrences of warming/drying on the lee sides of the CMR during Typhoon Krosa’s stay in Taiwan.

b. Ducting conditions

The vertical profile of modified refractivity index composed by using the Hualien sounding at 0500 UTC 6 October 2007 (Fig. 6c) is shown in Fig. 10. A pronounced ducting layer (dM/dz < 0) was found between 120 and 200 m in altitude; this ducting feature was similar to the elevated ducts documented in Babin (1996) and Brooks et al. (1999). Both studies include descriptions of various types of surface duct, including evaporation, surface based, and elevated ducts. The incident angle between the radar beams and the ducting layer also contributes to the ducting effect, however—a steeper radar elevation angle will lead to a smaller ducting layer (Skolnik 2001).

Fig. 10.
Fig. 10.

The lower-layer modified refractivity index M calculated from Hualien sounding data at 0500 UTC 6 Oct 2007 (Fig. 6c).

Citation: Journal of Applied Meteorology and Climatology 50, 7; 10.1175/2011JAMC2619.1

As revealed in past studies (e.g., Babin 1996; Brooks et al. 1999; Steiner and Smith 2002), knowing the climatological characteristics of surface ducts was helpful in determining the frequency and possible effects of AP phenomena, especially for coastal radars where evaporation-induced ducts were common. Because RCHL radar faces the Pacific Ocean to the east, the impact of surface ducts is a concern. By using an approach similar to Babin (1996), occurrences of various duct heights were calculated from 4 yr of sounding data from 2006 to 2009 (0000 and 1200 UTC) for 10-m bins in altitude. Surface duct occurrences were found to occur about 30% of the time, which was similar to the previous findings (Babin 1996; Steiner and Smith 2002). The duct heights were predominately less than 50 m and accounted for about 50% of all surface ducts (Fig. 11a). A second and very weak maximum duct height occurrence was found at 100 m, with a value of less than 3%. The duct depth histogram showed a maximum occurrence of about 62% in the 10–20-m bin and no occurrences above 100 m (Fig. 11b). The RCHL radar is located at 63 m MSL, higher than the most common surface duct heights. Furthermore, the duct depths in the area seemed to be too shallow for the occurrences of ducting (Turton et al. 1988). Therefore AP phenomena should less likely occur around the RCHL site. Multiyear observations from RCHL radar showed low frequencies of AP occurrences, confirming the unfavorable conditions for ducting at the radar site.

Fig. 11.
Fig. 11.

(a) Histogram of surface duct heights calculated from Hualien sounding observations at 0000 and 1200 UTC during 2006–09. (b) As in (a), but for the surface duct depths. The duct heights and duct depths are placed in 10-m bins.

Citation: Journal of Applied Meteorology and Climatology 50, 7; 10.1175/2011JAMC2619.1

c. Simulations of ray paths

Figure 12 shows the ray paths simulated using the equivalent earth model (Doviak and Zrnic 1993) and the sounding profile at Hualien at 0500 UTC 6 October 2007 (Fig. 6c). Given that the 3-dB radar beamwidth is 0.5°, simulations were performed for elevation angles from 0° to 0.5° assuming a main lobe directed at 0.5° (the lowest elevation angle for RCHL). For elevation angles lower than 0.2°, the ray paths propagated downward toward the surface and the ducting features of these elevation angles were characterized by approximately 200 km in wavelength and approximately 60 m in amplitude. In contrast, for elevation angles greater than or equal to 0.2°, the ray paths propagated upward smoothly but the beam heights with increasing range were still lower than these under normal propagation conditions (not shown). The beam-splitting feature between 0.1° and 0.2°, known as the “radio hole” as described in Fornasiero et al. (2006), would distort drastically the integration volume assumed under the normal propagation condition and would result in interpretation errors of weather echoes both in intensities and in altitudes.

Fig. 12.
Fig. 12.

The radar ray paths relative to the RCHL radar for elevations from 0° to 0.5° that were derived by using the calculations of refractivity index and ray tracing that are based on the sounding profile at 0500 UTC 6 Oct 2007.

Citation: Journal of Applied Meteorology and Climatology 50, 7; 10.1175/2011JAMC2619.1

As shown in Figs. 3 and 4, the onset of foehn winds in the southeastern part of Taiwan was about 2 h earlier than in the eastern-central part of Taiwan where the differences of thermodynamic environments might contribute to variations in the theoretical ray paths. The ducting features of ray paths below 0.2°, however, generally agreed with the occurrences of AP echoes at ranges between 100 and 150 km at 0505 UTC (Fig. 7a) near the southern direction of the RCHL radar.

d. Temporal variations of inversion heights

1) Constant inversion heights along ray paths

As shown in radar reflectivities (Fig. 7), the AP echoes continued to move toward the radar 2–3 h prior to the occurrence of foehn winds (Fig. 4d) around the radar. From the successive sounding observations at station Hualien, it is found that the subsidence warming was enhanced and the inversion height was decreased by strengthening downslope winds from 0000 to 0500 UTC (Figs. 6a,c). To investigate the moving characteristics of AP echoes over time, the T and Td profiles below 600 m (Fig. 9), which were based on assuming various inversion heights (from 300 to 50 m) with an interval of 50 m (A–F in Fig. 9) from the Hualien sounding observation at 0500 UTC 6 October 2007, were modified to examine the ducting effects at a fixed elevation angle of 0.1°.

Figure 13 shows the pronounced features of the ducting phenomena, which only presented at inversion heights between the ranges of 100 and 250 m. As the inversion heights decreased from 250 to 100 m, the simulated ducting amplitudes decreased from around 50 to 25 m and the wavelengths shortened from about 200 to 40 km. Because the inversion height was assumed to be at 50 m, the ray path went upward dramatically because the inversion height was lower than the RCHL radar height (63 m) and would induce the subrefraction phenomenon for the radar electromagnetic waves. As documented in Lopez (2009), increasing the installation height of ground-based radars may reduce the frequency of ducting conditions because the steepest refractivity gradients are generally confined within tens of meters of the surface (e.g., Babin et al. 1997; Brooks et al. 1999). The ground clutter induced by the side lobes is likely to intensify as the radar height increases, however (e.g., Doviak and Zrnic 1985).

Fig. 13.
Fig. 13.

Similar to Fig. 12, but for the radar ray paths at elevation 0.1° from idealized vertical inversion profiles from A to F as indicated in Fig. 9.

Citation: Journal of Applied Meteorology and Climatology 50, 7; 10.1175/2011JAMC2619.1

2) Various inversion heights along ray paths

The impact of spatial variations of the inversion height was also studied by assuming a changing inversion height with range along a ray path. Four experiment sets of inversion profiles were examined: 1) the various inversion heights (A–F in Fig. 9) at the radar site linearly decrease to 0 m for a range of 0–100 km, 2) as in experiment set 1 but for a range of 0–50 km, and 3) and 4) as in experiment sets 1 and 2, respectively, but the inversion heights increase from 0 m to a given height.

The ducting features were most remarkable in experiment 1: there was no ducting when the beginning inversion height was 50 m (case F, Fig. 14a) at the radar site. For cases A, B, C, D, and E, the first ducting positions occurred at the ranges of 61, 58, 40, 29, and 20 km, respectively. The ray path in case E escaped the ducting and went monotonically up after 20-km range. Beyond 85 km, all of the ray paths were free of ducting except for case D, in which the ray was trapped at a height near 42 m beyond 50 km. In experiment 2 in which the inversion height decreased more rapidly with a range 0–50 km, the ray paths escaped the ducting at shorter ranges (Fig. 14b) than in experiment 1 (Fig. 14a). The first ducting positions occurred at the ranges of 44, 35, 26, and 19 km for cases A, C, D, and E, respectively. The ray trapping occurred for case B at a height of 78 m in experiment 2 instead of D in experiment 1. There was no ducting in experiments 3 and 4 (not shown), for which all ray paths were very similar to case F in Fig. 14b even though significant inversions occurred at far ranges. These sensitivity experiments indicate that the inversion height near the radar site plays a more critical role in determining the ducting conditions than do the inversions at far ranges.

Fig. 14.
Fig. 14.

As in Fig. 13, but for inversion heights (A–F in Fig. 9) linearly decreased to 0 m from the radar site to the range at (a) 100 and (b) 50 km.

Citation: Journal of Applied Meteorology and Climatology 50, 7; 10.1175/2011JAMC2619.1

5. Conclusions

This study uses radar, surface stations, and sounding data to document the unusual radar anomalous propagation phenomena that accompanied the occurrence of foehn winds that were induced by Typhoon Krosa (2007). As Typhoon Krosa moved along a north–south track off the east coast of Taiwan on 6 October, foehn winds with significant subsidence warming (a maximum T of 38°C) and drying (a maximum TTd of 20°C) generated by downslope winds were observed at station Chenggong (46761) (Fig. 4c), and foehn winds continuously moved northward within confined areas from the southeast to eastern-central parts of Taiwan. Two to three hours prior to the occurrence of foehn wind at the radar site, the radar AP echoes were found to be embedded within Typhoon Krosa’s rainband areas and exhibited an inward motion toward the radar site.

Sounding profiles also showed remarkable warming and drying, with a maximum TTd of approximately 15°C near the inversion top of approximately 600 m, while the wind directions were characterized by veering clockwise from southerly to northwesterly (Fig. 6c). The ducting layer was found between 120 and 200 m in altitude (Fig. 10)—just below the inversion top whose location was calculated by using the modified refractive index. In this study, ray paths calculated by applying ray-tracing techniques showed that the electromagnetic signal propagated downward to the surface and showed the occurrence of beam splitting, which probably distort the radar observations (Fig. 12). These beam-splitting paths were characterized by a wavelength of approximately 200 km and amplitudes of approximately 60 m for elevation angles lower than 0.2°. The ducting features of ray paths generally agreed with the occurrences of AP echoes recorded in the actual radar observations.

By using the varying inversion heights from 300 to 50 m, the ray paths were further examined to diagnose the ducting conditions. When the inversion heights decreased from 250 to 100 m, the wavelengths under ducting conditions shortened from about 200 to 40 km (Fig. 13). At the inversion height of 50 m, however, the ducting conditions disappeared and were replaced by subrefraction conditions because the inversion height in this case was lower than the radar height. These results generally agree with the occurrences and inward motions of AP echoes during the passage of Typhoon Krosa.

It is suggested that the prefoehn wind environments induced by a typhoon could potentially provide favorable conditions for the occurrences of radar ducting, which might lead to contaminations in precipitation echoes and errors in range–height calculations. With sounding observations, the occurrences of AP could potentially be assessed in real time by using the proposed algorithm and misinterpretations of weather radar data could be minimized. Because of limited high-resolution temperature and humidity sounding data, the current study only explored ray paths along the radial, and the AP phenomena were investigated at a relatively large scale. The scheme can be straightforwardly expanded to two dimensions if proper sounding data become available, however. It is suggested that the coupling of a radar propagation model and mesoscale model data in the future would potentially provide a better understanding of the occurrences and evolutions of AP phenomena under foehn wind environments over the Taiwan area.

Acknowledgments

The authors thank the Central Weather Bureau for providing the analyzed data and computer resources. Dr. Jian Zhang and three anonymous reviewers provided valuable comments that greatly improved the manuscript. This research is supported by the National Science Council of Taiwan, Republic of China, under Grants 98-2625-M-052-005- and 99-2625-M-052-004-MY3.

REFERENCES

  • Atkinson, B. W., and M. Zhu, 2006: Coastal effects on radar propagation in atmospheric ducting conditions. Meteor. Appl., 13, 5362.

  • Babin, S. M., 1996: Surface duct height distribution for Wallop Island, Virginia, 1985–1994. J. Appl. Meteor., 35, 8693.

  • Babin, S. M., G. S. Young, and J. A. Carton, 1997: A new model of the oceanic evaporation duct. J. Appl. Meteor., 36, 193204.

  • Bean, B. R., and E. J. Dutton, 1968: Radio Meteorology. Dover, 435 pp.

  • Bebbington, D., S. Rae, J. Bech, B. Codina, and M. Picanyol, 2007: Modelling of weather radar echoes from anomalous propagation using a hybrid parabolic equation method and NWP model data. Nat. Hazards Earth Syst. Sci., 7, 391398.

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

    • Search Google Scholar
    • Export Citation
  • Brooks, I., A. Goroch, and D. Rogers, 1999: Observations of strong surface radar ducts over the Persian Gulf. J. Appl. Meteor., 38, 12931310.

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

    • Search Google Scholar
    • Export Citation
  • Chen, T. C., S. Y. Wang, M. C. Yen, A. J. Clark, and J. D. Tsay, 2010: Sudden surface warming/drying events caused by typhoon passages across Taiwan. J. Appl. Meteor. Climatol., 49, 234252.

    • Search Google Scholar
    • Export Citation
  • Doviak, R. J., and D. S. Zrnic, 1985: Siting of Doppler weather radars to shield ground targets. IEEE Trans. Antennas Propag., 33, 685689.

    • Search Google Scholar
    • Export Citation
  • Doviak, R. J., and D. S. Zrnic, 1993: Doppler Radar and Weather Observations. Academic Press, 562 pp.

  • Drechsel, S., and G. J. Mayr, 2008: Objective forecasting of foehn winds for a subgrid-scale Alpine valley. Wea. Forecasting, 23, 205218.

    • Search Google Scholar
    • Export Citation
  • Fornasiero, A., P. P. Alberoni, and J. Bech, 2006: Statistical analysis and modelling of weather radar beam propagation in the Po Valley (Italy). Nat. Hazards Earth Syst. Sci., 6, 303314.

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

    • Search Google Scholar
    • Export Citation
  • Gaffin, D. M., 2002: Unexpected warming induced by foehn winds in the lee of the Smoky Mountains. Wea. Forecasting, 17, 907915.

  • Gaffin, D. M., 2007: Foehn winds that produced large temperature differences near the southern Appalachian Mountains. Wea. Forecasting, 22, 145159.

    • Search Google Scholar
    • Export Citation
  • Gaffin, D. M., 2009: On high winds and foehn warming associated with mountain-wave events in the western foothills of the southern Appalachian Mountains. Wea. Forecasting, 24, 5375.

    • Search Google Scholar
    • Export Citation
  • Gao, J., K. Brewster, and M. Xue, 2006: A comparison of the radar ray path equations and approximations for use in radar data assimilation. Adv. Atmos. Sci., 32, 190198.

    • Search Google Scholar
    • Export Citation
  • Hubbert, J. C., M. Dixon, and S. M. Ellis, 2009: Weather radar ground clutter. Part II: Real-time identification and filtering. J. Atmos. Oceanic Technol., 26, 11811197.

    • Search Google Scholar
    • Export Citation
  • Jian, G. J., and C. C. Wu, 2008: A numerical study of the track deflection of supertyphoon Haitang (2005) prior to its landfall in Taiwan. Mon. Wea. Rev., 136, 598615.

    • Search Google Scholar
    • Export Citation
  • Lee, W. C., R. E. Carbone, and R. M. Wakimoto, 1992: The evolution and structure of a “bow-echo–microburst” event. Part I: The microburst. Mon. Wea. Rev., 120, 21882210.

    • 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
  • Lopez, P., 2009: A 5-yr 40-km-resolution global climatology of superrefraction for ground-based weather radars. J. Appl. Meteor. Climatol., 48, 89110.

    • Search Google Scholar
    • Export Citation
  • Miller, M. A., J. Verlinde, C. V. Gilbert, G. J. Lehenbauer, J. S. Tongue, and E. E. Clothiaux, 1998: Detection of nonprecipitating clouds with the WSR-88D: A theoretical and experimental survey of capabilities and limitations. Wea. Forecasting, 13, 10461062.

    • Search Google Scholar
    • Export Citation
  • Nkemdirim, L. C., 1986: Chinooks in southern Alberta: Some distinguishing nocturnal features. J. Climatol., 6, 593603.

  • Oard, M. J., 1993: A method for predicting chinook winds east of the Montana Rockies. Wea. Forecasting, 8, 166180.

  • Rogers, R. R., and M. K. Yau, 1989: A Short Course in Cloud Physics. 3rd ed. Pergamon, 293 pp.

  • Seluchi, M. E., F. A. Norte, P. Satyamurty, and S. C. Chou, 2003: Analysis of three situations of the foehn effect over the Andes (zonda wind) using the Eta–CPTEC regional model. Wea. Forecasting, 18, 481501.

    • Search Google Scholar
    • Export Citation
  • 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.

    • Search Google Scholar
    • Export Citation
  • Skolnik, M. L., 2001: Introduction to Radar Systems. 3rd ed. McGraw-Hill, 772 pp.

  • Steiner, M., and J. A. Smith, 2002: Use of three-dimensional reflectivity structure for automated detection removal of nonprecipitating echoes in radar data. J. Atmos. Oceanic Technol., 19, 673686.

    • Search Google Scholar
    • Export Citation
  • Torres, S. M., and D. S. Zrnic, 1999: Ground clutter canceling with a regression filter. J. Atmos. Oceanic Technol., 16, 13641372.

  • Turton, J. D., D. A. Bennets, and S. F. G. Farmer, 1988: An introduction to radio ducting. Meteor. Mag., 117, 245254.

  • Wakimoto, R. M., 1985: Forecasting dry microburst activity over the high plains. Mon. Wea. Rev., 113, 11311143.

  • 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
Save