• Amitai, E., 2000: Systematic variation of observed radar reflectivity–rainfall rate relations in the tropics. J. Appl. Meteor., 39 , 21982208.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Babin, S. M., 1996: Surface duct height distributions for Wallops Island, Virginia, 1985–1994. J. Appl. Meteor., 35 , 8693.

  • Bellon, A., Lee G. W. , and Zawadzki I. , 2005: Error statistics of VPR corrections in stratiform precipitation. J. Appl. Meteor., 44 , 9981015.

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

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chen, C. S., and Chen Y. L. , 2003: The rainfall characteristics of Taiwan. Mon. Wea. Rev., 131 , 13231341.

  • Doviak, R., and Zrnić D. , 1993: Doppler Radar and Weather Observations. Academic Press, 562 pp.

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

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

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Joe, N. C., Rinderknecht D. M. , and Hamilton R. S. , 1995: WSR-88D clutter suppression and its impact on meteorological data interpretation. WSR-88D Operational Support Facility Operations Training Branch, Norman, OK, 14 pp.

    • Search Google Scholar
    • Export Citation
  • Kessinger, C., Ellis S. , and Andel J. V. , 2003: The radar echo classifier: A fuzzy logic algorithm for the WSR-88D. Preprints, 19th Int. Conf. on Interactive Information Processing Systems (IIPS) for Meteorology, Oceanography, and Hydrology, Long Beach, CA, Amer. Meteor. Soc., P1.6. [Available online at http://ams.confex.com/ams/pdfpapers/54946.pdf].

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

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

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lakshmanan, V., Fritz A. , Smith T. , Hondl K. , and Stumpf G. , 2007: An automated technique to quality control radar reflectivity data. J. Appl. Meteor. Climatol., 46 , 288305.

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

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

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Morin, E., and Gabella M. , 2007: Radar-based quantitative precipitation estimation over Mediterranean and dry climate regimes. J. Geophys. Res., 112 , D20108. doi:10.1029/2006JD008206.

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

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

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

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ulbrich, C. W., and Lee L. G. , 1999: Rainfall measurement error by WSR-88D radars due to variations in Z–R law parameters and the radar constant. J. Atmos. Oceanic Technol., 16 , 10171024.

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

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zhang, J., Wang S. , and Clarke B. , 2004: WSR-88D reflectivity quality control using horizontal and vertical reflectivity structure. Preprints, 11th Conf. on Aviation, Range, and Aerospace Meteorology, Hyannis, MA, Amer. Meteor. Soc., P5.4. [Available online at http://ams.confex.com/ams/pdfpapers/81858.pdf].

    • Search Google Scholar
    • Export Citation
  • View in gallery
    Fig. 1.

    The weather radar network in Taiwan. Gray shades represent the terrain heights. Radar sites are labeled with their abbreviations. Range rings of 230 km centered at each radar site are also indicated.

  • View in gallery
    Fig. 2.

    FOR (%) of reflectivities ≥0 dBZ at 0.5° elevation for (a) RCWF, (b) RCHL, (c) RCCG, and (d) RCKT. The coverage range is 460 km for all radar sites.

  • View in gallery
    Fig. 3.

    FOR (%) of reflectivities ≥0 dBZ at 0.5° elevation for RCWF in (a) summer and (b) winter. The coverage range is 460 km for all radar sites.

  • View in gallery
    Fig. 4.

    Total rainfall accumulations for (a) spring, (b) summer, (c) fall, and (d) winter during 2005–07. The gauge stations are labeled with plus signs (+).

  • View in gallery
    Fig. 5.

    Same as in Fig. 4, but for rainfall frequency of ≥0.5 mm at 10-min intervals.

  • View in gallery
    Fig. 6.

    Hybrid scans (a),(c) before and (b),(d) after the application of the reflectivity climatology for the (a),(b) RCWF and (c),(d) RCCG radars.

  • View in gallery
    Fig. 7.

    Mosaicked hybrid scan heights (a) before and (b) after the application of the reflectivity climatology. White colors represent the hybrid scan areas higher than the fifth tilt.

  • View in gallery
    Fig. 8.

    Hybrid scan FORs (≥20 dBZ) (a),(c) before and (b),(d) after the application of the reflectivity climatology for (a),(b) spring and (c),(d) summer.

  • View in gallery
    Fig. 9.

    Same as in Fig. 8, but for rainfall accumulations.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 333 153 11
PDF Downloads 260 123 10

An Application of Reflectivity Climatology in Constructing Radar Hybrid Scans over Complex Terrain

Pao-Liang ChangCentral Weather Bureau, Taipei, Taiwan

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

Search for other papers by Pin-Fang Lin in
Current site
Google Scholar
PubMed
Close
,
Ben Jong-Dao JouDepartment of Atmospheric Sciences, National Taiwan University, Taipei, Taiwan

Search for other papers by Ben Jong-Dao Jou in
Current site
Google Scholar
PubMed
Close
, and
Jian ZhangNOAA/OAR/National Severe Storms Laboratory, Norman, Oklahoma

Search for other papers by Jian Zhang in
Current site
Google Scholar
PubMed
Close
Full access

Abstract

Three years’ worth of radar reflectivity data from four radars in an area of complex terrain (Taiwan) from 2005 to 2007 were analyzed and a reflectivity climatology was developed. The climatology was applied in the construction of new hybrid scans to minimize the impacts of ground clutter and beam blockages. The reflectivity climatology showed significant seasonal variations and captured distributions of ground/sea clutters, beam blockages, and anomalous propagations in addition to precipitation systems in the radar domains.

By comparing the reflectivity climatology with gauge observations, it was found that 15 (20) dBZ was a good approximation for rain/no-rain segregation during cool (warm) seasons. Comparisons between the standard (i.e., based on terrain and scan strategies only with the assumption of standard propagations) and nonstandard (i.e., standard plus the clutter and blockage mitigation using the reflectivity climatology) hybrid scans showed that the former did not accurately reflect the clutter and blockage distributions in the real atmosphere. The application of the reflectivity climatology was shown to significantly reduce the impacts of clutter and blockages and provided improved radar quantitative precipitation estimates (QPEs) in the complex terrain.

Corresponding author address: Pao-Liang Chang, Meteorological Satellite Center, Central Weather Bureau, No. 64, Gongyuan Road, Taipei, Taiwan. Email: larkdi@msc.cwb.gov.tw

Abstract

Three years’ worth of radar reflectivity data from four radars in an area of complex terrain (Taiwan) from 2005 to 2007 were analyzed and a reflectivity climatology was developed. The climatology was applied in the construction of new hybrid scans to minimize the impacts of ground clutter and beam blockages. The reflectivity climatology showed significant seasonal variations and captured distributions of ground/sea clutters, beam blockages, and anomalous propagations in addition to precipitation systems in the radar domains.

By comparing the reflectivity climatology with gauge observations, it was found that 15 (20) dBZ was a good approximation for rain/no-rain segregation during cool (warm) seasons. Comparisons between the standard (i.e., based on terrain and scan strategies only with the assumption of standard propagations) and nonstandard (i.e., standard plus the clutter and blockage mitigation using the reflectivity climatology) hybrid scans showed that the former did not accurately reflect the clutter and blockage distributions in the real atmosphere. The application of the reflectivity climatology was shown to significantly reduce the impacts of clutter and blockages and provided improved radar quantitative precipitation estimates (QPEs) in the complex terrain.

Corresponding author address: Pao-Liang Chang, Meteorological Satellite Center, Central Weather Bureau, No. 64, Gongyuan Road, Taipei, Taiwan. Email: larkdi@msc.cwb.gov.tw

1. Introduction

Weather radar has been one of the main remote sensing techniques used for measuring precipitation around the world. Radar precipitation estimates are usually computed from reflectivity observations on the so-called hybrid scans. The hybrid scan of a radar includes the lowest radar bins that do not have significant blockages or clutter contaminations (O’Bannon 1997; Maddox et al. 2002). These radar bins, possibly from different tilts at different locations, consist of a 2D polar grid. These standard “terrain based” hybrid scans have been shown to effectively mask the significant blockages and have been widely used in operational radars (Fulton et al. 1998) and also for improving quantitative precipitation estimation in complex orography (Morin and Gabella 2007).

However, the beam blockage information in these algorithms was based on standard beam propagations and only accounted for the main lobes. The actual blockages and clutter distributions can deviate significantly from these assumptions due to anomalous propagations, clutter contaminations from sidelobes, as well as clutter from trees and man-made objects (e.g., buildings, telecommunication towers, or wind farms) that are not included in standard terrain data.

Reflectivity climatology has been used to study spatial distributions of precipitation in different geography regions (e.g., Kuo and Orville 1973; Steenburgh et al. 2000; Heinselman and Schultz 2006). This method has also been used to build rainfall detection maps (e.g., Krajewski and Vignal 2001) to account for beam blockages that were not represented under standard propagation conditions. Further, the reflectivity climate maps can show clutter occurrences in real atmospheric environments, whether under standard or nonstandard refraction conditions. These studies suggest that the reflectivity climatology can be useful over complex terrain where radar-based quantitative precipitation estimates (QPEs) are often problematic due to ground clutter and blockages.

Taiwan experiences on average three to four typhoons each year (Wu and Kuo 1999); these storms bring torrential rains to the island and can trigger floods and landslides, especially in mountainous catchments. To improve our capability of monitoring and predicting flash floods, debris flows, and severe storms over the complex terrain of Taiwan, four Doppler weather radars were deployed during 1996–2001 by the Central Weather Bureau (CWB) of Taiwan. The base-level radar data were archived at the CWB and provided a great opportunity to build a radar reflectivity climatology for Taiwan.

In the current study, 3 years of radar data from 2005 to 2007 were analyzed to investigate characteristics of meteorological and nonmeteorological echoes as well as the beam blockage distributions for the four radars in Taiwan. New hybrid scans based on the climatology and gauge observations were developed to delineate areas with high probabilities of clutter and blockages. The next section describes the data and methodology for deriving the reflectivity climatology. Section 3 presents the application of the climatology in constructing radar hybrid scans and the impacts of the new hybrid scans on radar QPE. A brief summary is given in section 4.

2. Radar reflectivity climatology

a. Radar data

There are two types of operational Doppler radar systems in the Taiwan weather radar network. One is the Weather Surveillance Radar-1988 Doppler (WSR-88D), located at Wu-Fan San (RCWF). In addition, there are Gematronik 1500S Doppler radars located at Hual-Lien (RCHL), Chi-Gu (RCCG), and Ken-Ting (RCKT) (Fig. 1). Both radar systems have a wavelength of 10 cm (S band) and a beamwidth of 1°. Two scan modes were adopted for the Gematronik radar to fit the WSR-88D scanning strategy, that is, volume coverage pattern (VCP) 21 (Miller et al. 1998). The first mode is for surveillance, which uses a low pulse repetition frequency (PRF) to obtain long unambiguous range (460 km) in the lowest two tilts. The second mode, with nine scans, uses high PRF to obtain large unambiguous velocities, and there is a 1-min time gap between the two scan modes. Each composite radar volume contains two scan modes with 11 plan position indicator (PPI; Doviak and Zrnić 1993) scans with nine elevation angles, and takes 6–7 min to complete for the Gematronik radar. The volume update cycle for RCKT radar is 8 min and it is 10 min for the RCCG and RCHL radars.

Due to the complex terrain and radio environment in and around Taiwan, some special scan strategies were used in the CWB’s radar operations. There are two major mountain ranges in Taiwan: one is the Snow Mountain Range in northern Taiwan, which stretches southwest–northeast, and the other is the Central Mountain Range, which stretches south–north across most of Taiwan (Fig. 1). The Central Mountain Range has an average elevation of more than 2000 m, with the highest peak close to 4000 m above mean sea level. As a result, ground clutter and beam blockage problems are acute for all radars. To avoid interfered returns directly from the topography, electromagnetic wave emissions are turned off for some sectors of the RCHL and RCKT radars (Table 1). Most of the ground clutter and heavy electronic interference from military radar were removed in the signal processing of RCCG at the lowest two tilts, yet they still exist in the third and fourth tilts. These complex scan strategies can limit the effectiveness of some existing radar data quality control (QC) algorithms (e.g., Kessinger et al. 2003; Zhang et al. 2004; Lakshmanan et al. 2007) that depend on vertical reflectivity gradients. The reflectivity climatology, on the other hand, can capture persistent clutter and electronic interference features and also help the QC algorithms.

b. Frequency of occurrences of reflectivity

The radar reflectivity climatology in this paper is defined as the frequency of occurrence of reflectivity (FOR) that is higher than a certain threshold, and it is calculated on a pixel-by-pixel basis in the radar (spherical) coordinates (similar to Krajewski and Vignal 2001). The values of FOR range from 0 to 1, with 1 representing echoes that are detected all the time (e.g., power returns from a tall building or mountain) and 0 indicating that no echoes higher than the threshold are ever detected (e.g., behind an object that completely blocks the radar beam). The FOR fields for the lowest five elevation angles (i.e., 0.5°, 1.5°, 2.4°, 3.4°, and 4.3°) and 16 thresholds (i.e., from −10 to 65 dBZ at 5-dBZ intervals) are calculated for the four radars using real-time reflectivity data starting in 2005. Table 2 summarizes the data used in the current radar climatology study. Reflectivity data were generally available every 6–10 min, and about 150 000 volume scans from each radar site (Table 2) were obtained in the 3 years (2005–07) used for this study. This large and complete dataset is comparable to previous studies of radar climatology such as those of Krajewski and Vignal (2001) and Berenguer et al. (2006).

Figure 2 shows FOR (≥0 dBZ) maps of the four radars on the first tilt. Sea clutters are evident for all radars with high FORs in the ocean sectors. For RCWF, sea clutter rings centered near 60-km range are apparent from the southeast to northwest, with FORs greater than 15% (Fig. 2a). A large sector to the south and southwest of the radar was blocked by the Snow Mountain Range because of its high peaks. Ground clutters were found near the radar as well as along the ridge of the Snow Mountain (Fig. 2a). Some scattered high FORs were found in the southeast coast of mainland China about 300 km from RCWF (Fig. 2a) and are highly correlated with the terrain in the region (Fig. 1). These high FORs occur much more often in the warm seasons than in winter (Fig. 3). The features are possibly related to anomalous propagations associated with temperature inversions and moisture gradients in the boundary layer (Steiner and Smith 2002), and the seasonal trend is consistent with the climatological studies of sounding data performed by Babin (1996) and Steiner and Smith (2002). The sea clutter features exist for all seasons with the most severe occurring in winter (Fig. 3b). This may be related to the strong winds associated with the northeastern monsoon in this season (Chen and Chen 2003).

Figure 2b shows the FOR distribution for RCHL. The most significant feature is the sea clutter to the east of the radar site, which exhibits an elliptical fan shape, with the long axis parallel to the coastline. A narrow beam blockage wedge to the northeast of the radar is a result of a tall building near the radar. According to the scan strategy (Table 1), the FOR should be void on the mountainside west of the radar. But low FOR values still exist because of some radio interferences from other radars with the same wavelength and some power reflected by the terrain. A large sector of low FOR values to the east of RCCG (Fig. 2c) was a result of the clutter suppression algorithm (Joe et al. 1995) employed in the signal processing of the radar for the first two tilts. The suppression is turned off at 2.4°, resulting in ground clutter echoes around the radar site and in high-elevation areas of the Central Mountain Range. The clutter area associated with the mountains is quite large and can be shown in tilts up to 4.3° (fifth tilt; not shown). A sector of silence (Table 1) is also implemented for the RCKT radar to avoid radiation hazards impacting people living to the north of the radar. The sector is shown in the reflectivity climatology as a low FOR area in Fig. 2d. Significant sea clutter is also observed by RCKT (Fig. 2d) due to its low altitude and proximity to the seashore.

3. Construction of hybrid scans using reflectivity climatology

The reflectivity climatology (FOR) at any given location is related to whether there is rain or clutter at the location. If all the echoes detected at the location are rain, then the expected range of FOR values should be close to those of the frequency of occurrence of gauge rainfall (FOG) (Steenburgh et al. 2000). The FOG in the current study is computed from observations of a dense rain gauge network of 370+ stations at a 10-min time interval. If the FOR at any point is lower (higher) than the homogeneous FOG in a statistical sense, then the radar bin at the point is considered to be blocked (clutter). Due to the large seasonal variabilities of rainfall inside the island of Taiwan (Chen and Chen 2003), maps of FORs and FOGs are compared for different seasons. Clutter and blockage distributions are analyzed based on FOR–FOG comparisons and the information is then used in radar hybrid scans.

a. Frequency of occurrences of gauge rainfall

Figure 4 shows the rainfall accumulation and FOG maps in Taiwan during the period of 2005–07. The maps were obtained by using an inverse distance weighting function with a 30-km radius of influence. The rainfall threshold for the FOG analysis was 0.5 mm, which was the minimum detectable rainfall amount for all gauges.

During winter, most of the rain fell in a narrow area at the northern tip of Taiwan (Fig. 4d), and there was little in the southwest plains. The wintertime FOG (Fig. 5d) showed similar patterns with a maximum of 15%. In spring, the rainfall area covered most of central and northwestern Taiwan with the maximum of ∼5000 mm in the central region (Fig. 4a). Similar patterns were found in the FOG (Fig. 5a) with a maximum of ∼12%. In summer, heavy rainfall of 5000 mm or more covered most of the mountain areas, especially at the central and southern Central Mountain Range where the 3-yr total reached 8000 mm (Fig. 4b). Similar patterns were found in the FOG (Fig. 5b) with a maximum of about 15%. The heavy rainfalls are mainly from mei-yu fronts (early summer), typhoons, and local rain showers during summertime. There was less rainfall in the fall season, with the high rainfall amounts and FOGs shifted toward northeastern Taiwan. The highest rainfall is about 2000 mm (Fig. 4c) and FOG is about 13% (Fig. 5c). These characteristics of the seasonal rainfall distributions are consistent with the study of Chen and Chen (2003). They documented that, due to the presence of the Central Mountain Range, the regional rainfall climate over the island was strongly dependent on the direction of low-level prevailing flows with much higher rainfall on the windward side.

b. Nonstandard hybrid scan

The FOG analysis is compared with the FORs (at 5-dBZ intervals) for the lowest five tilts of each radar. Areas with high frequencies of clutter and beam blockages are identified. The clutter and blockage climatology is then used to refine the standard hybrid scans, which is based on scan strategies and the terrain, only assuming standard propagations (O’Bannon 1997). The new hybrid scan will be referred to as a nonstandard hybrid scan herein.

If the FOR at a radar bin is much lower (higher) than the corresponding FOG, then the bin is considered to be potentially blocked (clutter contaminated) and the radar bin on the next higher tilt should be used in the hybrid scan. An initial threshold for the difference between the FOG and FOR is set to 10%. Any radar bin with |FOR − FOG| ≥ 10% is marked as a “bad” bin and the corresponding radar data should be subject to removal when constructing the new hybrid scan.

For areas with |FOR − FOG| < 10%, two additional variables are examined to further identify areas of clutter contamination. One variable is the texture of FOR (TFOR) and the other is the vertical gradient of FOR (VFOR). Similar variables were used for reflectivity fields in previous studies, such as Kessinger et al. (2003) and Zhang et al. (2004), for identifying clutter echoes. The TFOR is the mean-squared difference between adjacent FOR values along a radial; that is,
i1520-0426-26-7-1315-e1
where i and j are indices of radar bins in the azimuthal and range directions, respectively; P represents FOR values; and N is the number of points for calculation. The default size for the computational box is five rays by five gates. All radar gates with TFOR values greater than a threshold (default value is 0.01) are considered to be contaminated by clutter echoes and should be excluded when constructing the new hybrid scan.
The VFOR field is calculated as follows:
i1520-0426-26-7-1315-e2
where P is the FOR value under examination; Pu is the FOR value on the next higher tilt; and H and Hu are the heights associated with P and Pu, respectively. All radar bins with VFOR values larger than a threshold (default value is −0.05 km−1) are considered to be significantly blocked and/or contaminated by clutter. The radar bins at the next higher tilt should be used in the hybrid scan instead.

Figure 6 shows the hybrid scans before and after the reflectivity climatology is applied. Based on the standard propagation, the third tilt of RCWF is completely free of blockage. Also, the first tilt clears the ground near the radar (Fig. 6a). However, the nonstandard hybrid scan goes up to the fourth and even fifth tilts near and to the southwest of the radar (Fig. 6b). For the RCCG radar, the standard hybrid scan is fully under the fourth tilt (Fig. 6c), while the reflectivity climatology indicates that it is not until the fourth and fifth tilts where the radar is free of significant ground clutter and blockages near the Central Mountain Range. In addition, the nonstandard hybrid scan reaches the third to fifth tilts within tens of kilometers of the radar (Fig. 6d) to avoid heavy electronic interference. These interference signals were captured by the reflectivity climatology but are not reflected in the standard hybrid scan.

Hybrid scan heights from the four radars are mosaicked to assess the radar coverage in the Taiwan area. At any given location, the lowest hybrid scan height among all radars is chosen. The mosaicked standard hybrid scan height field (Fig. 7a) is within 1 km above ground level (AGL) in northern and southwestern Taiwan. In eastern part of Taiwan, the hybrid scan height is about 3–6 km AGL, mostly in the valley area south of the RCHL radar. After the reflectivity climatology was applied (Fig. 7b), the hybrid scan height in central-western Taiwan increased from 1 km to about 3–4 km AGL with some areas reaching 5–6 km. In eastern Taiwan, the hybrid scan height is mostly 3–5 km AGL with a maximum of 6–7 km in the valley south of the RCHL radar. It has long been recognized that the nonuniform vertical profile of reflectivity (VPR) is a major error source in radar QPEs (e.g., Bellon et al. 2005). The results in Fig. 7b suggest that the current weather radar network in Taiwan is not adequate for QPEs in the central and eastern areas. Additional radars or other observational data are needed to fill in the gaps in these regions.

c. Evaluations

To evaluate the impact of the reflectivity climatology, FOR fields are created on nonstandard hybrid scans with different reflectivity thresholds ranging from 5 to 25 dBZ. Root-mean-square errors (RMSEs) and coefficients of determination (CODs) between the FOGs and the corresponding FORs of different reflectivity thresholds are calculated for different seasons, and the results are summarized in Table 3. The RMSEs between the standard hybrid scan FORs and FOGs are significantly larger than those of the nonstandard hybrid scan (Table 3), showing some of the advantages of using the reflectivity climatology in reducing the impact of nonstandard blockages and ground clutter.

Further analysis indicates that the explained variance between the FOG and the FOR is only 15% in fall for a 5-dBZ threshold. As the threshold increases, the COD improves and the RMSE decreases until the threshold reaches 25 dBZ. The 10–15-dBZ thresholds gave the least error between the hybrid scan FORs and FOG in winter, and the 20–25-dBZ thresholds resulted in better correlations and smaller errors in summer and fall. In the summer, a 20-dBZ threshold gave the minimum RMSE (1.39%), while 25 dBZ gave the best explained variance (76%). This implies that the rain/no-rain threshold of reflectivity is higher in warm seasons and lower in cool seasons. The warm season precipitation system may contain more large raindrops than does the cool season system because of stronger updrafts. The larger drops are associated with higher reflectivities and, therefore, a better correlation between the FOG and FORs at higher reflectivity thresholds in warm seasons than in cool seasons. Based on the minimum RMSEs in Table 3, a reflectivity threshold of 15 dBZ could be used as a guide for rain/no-rain segregation in winter and 20 dBZ in spring to fall.

Figure 8 shows mosaicked FOR fields (≥20 dBZ) on the standard and nonstandard hybrid scans in the spring and summer seasons. The standard hybrid scan FORs (Figs. 8a and 8c) showed many discontinuities and spurious high values on the tops of the Snow Mountain Range and the southern part of the Central Mountain Range that are not found in the FOGs (Figs. 5a and 5b). Erroneous low FORs are found in the southwestern plain and part of central Taiwan as a result of beam blockages and clutter suppressions employed at RCCG. Similar discontinuities are also found in other seasons (not shown). The FORs from the nonstandard hybrid scans (Figs. 8b and 8d) showed a much more continuous set of distributions than did the standard hybrid scan FORs (Figs. 8a and 8c). In addition, the nonstandard hybrid scan FOR patterns are consistent with the FOG patterns (Figs. 5a and 5b), with two well-defined maximum areas in summer (Fig. 8d).

Figure 9 shows 3-yr rainfall accumulations derived from the archived FORs on the standard and nonstandard hybrid scans. The FORs were segregated for different reflectivity categories ranging from −10 to 65 dBZ at 5-dBZ interval. Rain rates were calculated from the middle value of each of the reflectivity categories using a relationship of Z = 300R1.4. The rates were then accumulated for the 3-yr period based on FOR values at each point on the hybrid scans. Similar to the FOR analysis in Fig. 8, the rainfall fields from the nonstandard hybrid scan (Figs. 9b and 9d) showed a much more physically realistic distribution than that of the standard hybrid scan (Figs. 9a and 9c). The maximum rainfall was about 3000 mm in central Taiwan in spring (Fig. 9b) and 7000 mm in the southwestern part of the Central Mountain Range in summer (Fig. 9d). In general, the maximum rainfall amounts from the radar QPEs are systematically lower than those of the gauge observations (Figs. 4a and 4b). The underestimations are mainly due to the limitations of the radar observations, such as the higher beam height (Fig. 6b) with nonuniform VPRs (Bellon et al. 2005) and the improper ZR relationship (Ulbrich and Lee 1999; Amitai 2000).

4. Conclusions

Three years’ worth of radar reflectivity data from four radars in an area of complex terrain (Taiwan) during 2005–07 were analyzed and a reflectivity climatology was developed. The climatology was applied in the construction of new hybrid scans to minimize the impacts of ground clutter and beam blockages. The reflectivity climatology showed significant seasonal variations and captured distributions of ground/sea clutters, beam blockages, and anomalous propagations in addition to precipitation systems in the radar domains.

By comparing the reflectivity climatology with gauge observations, it was found that 15 dBZ was a good approximation for rain/no-rain segregation in cool seasons and 20 dBZ worked well in warm seasons. Comparisons between the standard (i.e., based on terrain and scan strategies only with the assumption of standard propagations) and nonstandard (i.e., standard plus the clutter and blockage mitigation using the reflectivity climatology) hybrid scan FORs showed that the former did not accurately reflect the clutter and blockage distributions in the real atmosphere. The application of the reflectivity climatology was shown to significantly reduce the impacts of clutter and blockages and provided improved radar QPEs over the complex terrain.

Acknowledgments

The authors thank the Central Weather Bureau for providing the radar data and computer resources. The authors are also grateful to Dr. J. Marshall Shepherd and Dr. Peng-Fei Zhang for their valuable comments. We also thank three anonymous reviewers for their helpful comments on the manuscript. This research is supported by the National Science Council of Taiwan, Republic of China, under Grant 96-2625-Z-052-005.

REFERENCES

  • Amitai, E., 2000: Systematic variation of observed radar reflectivity–rainfall rate relations in the tropics. J. Appl. Meteor., 39 , 21982208.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Babin, S. M., 1996: Surface duct height distributions for Wallops Island, Virginia, 1985–1994. J. Appl. Meteor., 35 , 8693.

  • Bellon, A., Lee G. W. , and Zawadzki I. , 2005: Error statistics of VPR corrections in stratiform precipitation. J. Appl. Meteor., 44 , 9981015.

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

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chen, C. S., and Chen Y. L. , 2003: The rainfall characteristics of Taiwan. Mon. Wea. Rev., 131 , 13231341.

  • Doviak, R., and Zrnić D. , 1993: Doppler Radar and Weather Observations. Academic Press, 562 pp.

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

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

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Joe, N. C., Rinderknecht D. M. , and Hamilton R. S. , 1995: WSR-88D clutter suppression and its impact on meteorological data interpretation. WSR-88D Operational Support Facility Operations Training Branch, Norman, OK, 14 pp.

    • Search Google Scholar
    • Export Citation
  • Kessinger, C., Ellis S. , and Andel J. V. , 2003: The radar echo classifier: A fuzzy logic algorithm for the WSR-88D. Preprints, 19th Int. Conf. on Interactive Information Processing Systems (IIPS) for Meteorology, Oceanography, and Hydrology, Long Beach, CA, Amer. Meteor. Soc., P1.6. [Available online at http://ams.confex.com/ams/pdfpapers/54946.pdf].

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

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

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lakshmanan, V., Fritz A. , Smith T. , Hondl K. , and Stumpf G. , 2007: An automated technique to quality control radar reflectivity data. J. Appl. Meteor. Climatol., 46 , 288305.

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

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

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Morin, E., and Gabella M. , 2007: Radar-based quantitative precipitation estimation over Mediterranean and dry climate regimes. J. Geophys. Res., 112 , D20108. doi:10.1029/2006JD008206.

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

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

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

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ulbrich, C. W., and Lee L. G. , 1999: Rainfall measurement error by WSR-88D radars due to variations in Z–R law parameters and the radar constant. J. Atmos. Oceanic Technol., 16 , 10171024.

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

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zhang, J., Wang S. , and Clarke B. , 2004: WSR-88D reflectivity quality control using horizontal and vertical reflectivity structure. Preprints, 11th Conf. on Aviation, Range, and Aerospace Meteorology, Hyannis, MA, Amer. Meteor. Soc., P5.4. [Available online at http://ams.confex.com/ams/pdfpapers/81858.pdf].

    • Search Google Scholar
    • Export Citation

Fig. 1.
Fig. 1.

The weather radar network in Taiwan. Gray shades represent the terrain heights. Radar sites are labeled with their abbreviations. Range rings of 230 km centered at each radar site are also indicated.

Citation: Journal of Atmospheric and Oceanic Technology 26, 7; 10.1175/2009JTECHA1162.1

Fig. 2.
Fig. 2.

FOR (%) of reflectivities ≥0 dBZ at 0.5° elevation for (a) RCWF, (b) RCHL, (c) RCCG, and (d) RCKT. The coverage range is 460 km for all radar sites.

Citation: Journal of Atmospheric and Oceanic Technology 26, 7; 10.1175/2009JTECHA1162.1

Fig. 3.
Fig. 3.

FOR (%) of reflectivities ≥0 dBZ at 0.5° elevation for RCWF in (a) summer and (b) winter. The coverage range is 460 km for all radar sites.

Citation: Journal of Atmospheric and Oceanic Technology 26, 7; 10.1175/2009JTECHA1162.1

Fig. 4.
Fig. 4.

Total rainfall accumulations for (a) spring, (b) summer, (c) fall, and (d) winter during 2005–07. The gauge stations are labeled with plus signs (+).

Citation: Journal of Atmospheric and Oceanic Technology 26, 7; 10.1175/2009JTECHA1162.1

Fig. 5.
Fig. 5.

Same as in Fig. 4, but for rainfall frequency of ≥0.5 mm at 10-min intervals.

Citation: Journal of Atmospheric and Oceanic Technology 26, 7; 10.1175/2009JTECHA1162.1

Fig. 6.
Fig. 6.

Hybrid scans (a),(c) before and (b),(d) after the application of the reflectivity climatology for the (a),(b) RCWF and (c),(d) RCCG radars.

Citation: Journal of Atmospheric and Oceanic Technology 26, 7; 10.1175/2009JTECHA1162.1

Fig. 7.
Fig. 7.

Mosaicked hybrid scan heights (a) before and (b) after the application of the reflectivity climatology. White colors represent the hybrid scan areas higher than the fifth tilt.

Citation: Journal of Atmospheric and Oceanic Technology 26, 7; 10.1175/2009JTECHA1162.1

Fig. 8.
Fig. 8.

Hybrid scan FORs (≥20 dBZ) (a),(c) before and (b),(d) after the application of the reflectivity climatology for (a),(b) spring and (c),(d) summer.

Citation: Journal of Atmospheric and Oceanic Technology 26, 7; 10.1175/2009JTECHA1162.1

Fig. 9.
Fig. 9.

Same as in Fig. 8, but for rainfall accumulations.

Citation: Journal of Atmospheric and Oceanic Technology 26, 7; 10.1175/2009JTECHA1162.1

Table 1.

The radar parameters and scanning strategies of the WSR-88D and Gematronik 1500S Doppler radars.

Table 1.
Table 2.

Summary of data used in the reflectivity climatology from 2005 to 2007.

Table 2.
Table 3.

CODs and RMSEs between FOG and FORs with (without) clutter removal for various reflectivity thresholds and in different seasons. Minimum RMSE for each season is in bold.

Table 3.
Save