1. Introduction
In coastal regions, multiple interactive factors modify the moisture distribution and consequently change the precipitation patterns and dispersion of pollution. The moisture distribution in coastal regions is primarily modulated by the local diurnal land–sea circulations. Horizontal temperature gradients caused by differential solar heating and longwave radiative cooling result in pressure gradients that drive the daytime sea breeze and nighttime land breeze. In addition, the diurnal land–sea circulation interacts with a myriad of natural and anthropogenic processes, such as the prevailing large-scale flow (Arritt 1993; Chen et al. 2016, 2017; Wang and Sobel 2017), local circulations induced by topography or rivers (Zhong et al. 1991; Laird et al. 1995; Boybeyi and Raman 1992; Wang and Kirshbaum 2015), storm cold pools (Wilson and Megenhardt 1997; Soderholm et al. 2016), heterogeneous land use (Grant and van den Heever 2014; Park et al. 2020), and urban heat island effects (Baker et al. 2001; Kusaka et al. 2001; Ryu et al. 2016). These intertwined factors affect the intensity, orientation and propagation of the land–sea-breeze fronts and further result in complex moisture distributions, which most conventional observing network are unable to resolve in turn. The lack of frequent and dense surface observations limits the ability to represent the moisture variability that retards the short-term forecasting skills (Banacos and Schultz 2005; Madaus and Hakim 2016).
Under weak synoptic-scale forcing, the moisture flux convergence by the daytime sea breeze is one of the critical factors for the development and growth of inland afternoon thunderstorms (Johnson and Bresch 1991; Banacos and Schultz 2005; Chen and Nash 1994; Chen et al. 2014). For example, in Taiwan, the onset of the sea breeze usually occurs between 0900 and 1100 local solar time (LST). Moisture convergence induced by the daytime sea breeze usually occurs at the foothills around noon. Average hourly rainfall amounts frequently peak in the midafternoon along the lower slopes of the mountains. Warmer (0.5°–1.5°C in temperature) and more moist (0.5°–2°C in dewpoint temperature) surface air conditions are observed before convection initiation occurred on afternoon thunderstorms days compared to the no thunderstorm days (Lin et al. 2011). Diurnal rainfall is a prominent feature in coastal areas dominated by land–sea circulations (Atkins et al. 1995; Miller et al. 2003; Chen et al. 2018). Even though the statistical diurnal precipitation mechanisms have been widely studied, forecasting the diurnal precipitation in coastal areas remains an outstanding challenge. The location and timing of convection initiation and rainfall intensity in coastal areas are still poorly represented in weather and climate models (Stephens et al. 2010; Nguyen et al. 2015; Bergemann and Jakob 2016). Thus, investigating the complex interactions between the low-level moisture distribution and the local diurnal circulations are important to improve forecasts of diurnal precipitation and air pollution in coastal environments.
High-spatiotemporal-resolution observations are necessary to sample highly variable moisture anomalies. Current low-level moisture data are obtained in forms of point (e.g., surface stations, aircraft), profile (e.g., radiosonde, radiometers, space-based GPS receivers), and areal (e.g., satellites) observations. Satellite observations can provide the low-level vertically integrated moisture from synoptic to meso-γ scales with broad spatial coverage. For example, split window difference by Geostationary Operational Environmental Satellite-16 (GOES-16) is applied to detect boundaries of low-level integrated water vapor from the surface to 750 hPa prior to convection initiation (Lindsey et al. 2018). Yet, this technique has limited capabilities in environments with clouds and precipitation. Surface-based observing systems, such as surface mesonets and thermodynamic profilers (NRC report; National Research Council 2009; Wulfmeyer et al. 2015, Brotzge et al. 2020) are able to depict the low-level moisture distribution at meso-α to meso-β scales under all weather conditions, but they are currently deployed only in limited regions and field experiments. An additional rare but promising approach to bridge the low-level moisture observation gap is the refractivity fields estimated by ground-based weather radars (Fabry et al. 1997; Fabry 2004). Refractivity can provide higher-resolution observations of moisture distribution than the typical mesonet spacing of 30–50 km.
In addition to the conventional reflectivity, radial velocity and polarimetric variables, radars measure the refractivity distribution of the air approximately 20 m above the surface within a range of ~40 km from the radar. The radar-retrieved refractivity field can well represent the near-surface moisture distribution. During well-mixed boundary layer conditions in the afternoon, the refractivity can further represent the atmospheric conditions up to 200 m height above the ground (Weckwerth et al. 2005). Radar refractivity has been used for studying the evolving low-level moisture distribution at the meso-γ scale associated with convection initiation and boundary layer processes (Fabry 2006; Roberts et al. 2008; Buban et al. 2007; Koch et al. 2008; Bodine et al. 2010; Wakimoto and Murphey 2010; Besson et al. 2016). Radar refractivity has also been applied to adjust the initialization of moisture in numerical weather prediction models to improve quantitative precipitation forecasts through data assimilation (Montmerle et al. 2002; Sun 2005; Gasperoni et al. 2013; Seko et al. 2017). However, most of these previous studies focus on midlatitude continental storm environments, and no studies of refractivity have been applied in tropical coastal regions.
In this study, the S-band polarimetric Doppler radar of the National Center for Atmospheric Research (NCAR S-Pol; Hubbert et al. 2018) is used to retrieve the refractivity of near-surface air during the Terrain-influenced Monsoon Rainfall Experiment/Southwest Monsoon Experiment (TiMREX/SoWMEX) experiment in southwestern Taiwan in 2008. The scientific goal of this experiment is to investigate the physical mechanisms of heavy rainfall associated with southwesterly monsoons in order to advance quantitative precipitation forecasts (Davis and Lee 2012). This dataset provides us an opportunity to study the near-surface moisture distribution in a complex moisture-rich tropical coastal environment. The complex local physiography lends to a variety of flow features and corresponding moisture behavior, which are not fully understood yet.
This research aims to characterize the spatiotemporal variability of near-surface moisture in a summer tropical coastal region using radar-retrieved refractivity data. We will apply the high-resolution of refractivity analyses to address scientific questions: 1) What are the characteristics of the refractivity (moisture) fields in a moisture rich environment across various temporal scales? 2) How do these moisture fields evolve with the atmospheric processes, terrain, and heterogeneous land use? Section 2 introduces the data, geographical environment, and relates the radar-retrieved refractivity fields to the moisture distribution in the tropics. Section 3 presents the empirical spatiotemporal characteristics of the refractivity during four weeks of SoWMEX/TiMREX and discusses the diurnal evolution of the refractivity distribution under strong and weak synoptic forcing weather regimes. Section 4 shows three cases studies pinpointing the different diurnal refractivity characteristics under weak synoptic forcing conditions. Section 5 compares the similarities and differences of the diurnal refractivity variation between different weak synoptic forcing regimes. Conclusions and future applications of refractivity are summarized in section 6.
2. Data and methodology
The deployment of the instruments and the topography are shown in Fig. 1a. The S-Pol radar was located on the plains of southwestern Taiwan between the ocean and the meridional-oriented steep mountains exceeding 2000 m. The refractivity is retrieved from the azimuth-averaged in-phase and quadrant (I/Q) components obtained from the 0.5°-elevation scans every 7.5 min from 1 June to 28 June 2008 (UCAR/NCAR–Earth Observing Laboratory 2011). The refractivity retrieval is based on the methodology described Fabry (2004) and Fabry and Pettet (2002). For refractivity calibration, the relatively homogeneous reference phase is calculated between 1500 and 1530 LST 4 June 2008 under stratiform precipitation; the ground target selection is based on the data between 0830 and 0931 LST 9 June 2008. Weather information (e.g., temperature, relative humidity, wind, precipitation) obtained from the surface stations (Fig. 1b) with 1-min temporal resolution are applied to demonstrate the variation of the atmospheric states from coast to inland, as well as to evaluate the data quality of the radar-retrieved refractivity. Kaohsiung and ISS stations collect continuous data through June, but the other stations only operated after 5 June.
(a) The topography and Central Weather Bureau (CWB) conventional surface stations (cyan squares) in Taiwan. The terrain height (color shaded) is in meters above sea level. (b) The deployment of the NCAR S-Pol radar (yellow star), radiosonde (red triangle), and the surface stations (green dots) during the SoWMEX/TiMREX experiment [see red box in (a)]. Rivers are shown as black lines. (c) Land-use types.
Citation: Monthly Weather Review 149, 9; 10.1175/MWR-D-20-0425.1
The coverage of refractivity fields on the plains encompasses a maximum range of ~40 km northeast of the radar. Refractivity retrieval is not available over the ocean or on the high terrain of the eastern mountains due to the lack of reliable ground targets. The land-use map (Fig. 1c) retrieved from satellites (Cheng et al. 2013) indicates diverse land-use across the refractivity coverage region from the coast to the foothills, mainly including urban, rural dryland and irrigated cropland. Three rivers flowing from the mountains through the plains to the ocean also add complexity to the land surface.
a. The relationship between refractivity and moisture
To understand the causes of the temporal change of refractivity in the tropical coastal area, we analyze the first order of the refractivity change from Taylor series expansion of (1) based on the surface station data. Figure 2 shows that more than 70% of the hourly N change resulted from hourly changes of e, which is much greater than the contributions from T (15%–20%) and P (<5%). The evolving moisture dominates the temporal change of N at all stations. Similar results were also found in the summertime midlatitude continental regions (Creese 1999). In Fig. 2, a slightly higher T effect on N change is observed at stations inland compared to the coastal stations. This may be related to the larger diurnal temperature change inland compared with the relatively stable temperature condition at the coast.
The normalized contribution (%) of hourly changes in water vapor (blue), temperature (red), and pressure (black) to hourly change of refractivity in June 2008. This analysis is calculated from the observations at surface stations from coast to inland: Kaohsiung, Shiwei, Fenhua, and ISS, respectively.
Citation: Monthly Weather Review 149, 9; 10.1175/MWR-D-20-0425.1
The refractivity distribution elucidates the spatial gradient of water vapor and can denote the location and intensity of moisture boundaries. Since the horizontal temperature and pressure fields are generally homogeneous in the tropics compared to midlatitudes (Yano et al. 2013), the effect of water vapor variation on refractivity is more important in tropics. Under a weak temperature gradient, the refractivity field can be used to represent the water vapor distribution. Nevertheless, some strong temperature gradients occur during the passage of interesting weather phenomena, such as fronts, squall lines and cold pools. In these situations, uncertainties in the estimation of moisture may occur while directly relating the refractivity field to the moisture field. Quantitatively, moisture can be recovered with an error in TD on the order of 0.2°C given an average spatial T difference of ~1°C over the refractivity coverage region. In this case, based on Eq. (1), the accurate moisture distribution may be retrieved from the refractivity field once the temperature field is provided from either a dense surface station network, reanalysis data or numerical model forecasts through objective analysis or local-scale data assimilation.
b. Data quality of refractivity
To evaluate the data quality of the radar-retrieved refractivity, it is compared with the refractivity calculated from the surface observations based on Eq. (1). The N obtained from the radar-retrieval method and the surface observations at 2 m height above the ground are highly temporally consistent. Their temporal correlation coefficients over four weeks are high, namely, 0.82, 0.94, 0.90, and 0.89 at Kaohsiung, Shiwei, Fenhua, and ISS, respectively. The mean absolute difference of the N between the radar-retrieval and the surface measurements is ~4.1 N-unit, i.e., 0.8 g kg−1 in q (Hsu 2019). This bias is within the requirement of the accuracy of water vapor necessary for operational forecast applications, which is 1 g kg−1 in q (Hardesty et al. 2012). The radar-retrieved N at ~20 m above the surface is usually smaller than N at the surface, because the amount of moisture decreases with height in the boundary layer. The height difference between the surface measurement and the radar retrieval interplay with the evolving vertical gradient of moisture associated with the boundary layer mixing processes, resulting in N biases when compared with surface stations (Fabry 2004; Bodine et al. 2011; Feng and Fabry 2018).
3. The spatiotemporal characteristics of the near-surface refractivity fields
a. The average spatial distribution of refractivity
The four-week average refractivity field shown in Fig. 3 illustrates a regional near-surface moisture contrast associated with the general land–sea geography in a summertime coastal environment of southwestern Taiwan. The average refractivity field shows the moisture gradient perpendicular to the coastline. The value of refractivity is higher along the coast, gradually decreases inland, and reaches a minimum at the foothills. Moreover, high N extends further inland in the northern urban area in contrast with the southern coastal rural area. The lowest average N area at the foothills is parallel to the meridional oriented mountains and is bounded by small hills along 120°25′E and rivers. The magnitude and orientation of the refractivity gradient denote the finer spatial distribution of moisture related to land-use types, rivers, and the inland topography. The refractivity field further depicts detailed spatial variability of moisture between the conventional surface stations that have spacings of ~20 km on the horizontal scale.
The mean radar-derived refractivity distribution with the mean temperature (superscript next to the surface station, °C) and the mean dewpoint temperature (subscript, °C) from surface stations averaged from 1 to 28 Jun 2008. The black contours are the topographic heights at 50, 100, and 500 m above the sea level.
Citation: Monthly Weather Review 149, 9; 10.1175/MWR-D-20-0425.1
A spatial gradient of the refractivity presents a ~10 N-unit difference across 30 km in distance. The decreasing trend of moisture from the coast to the foothills is consistent with the in situ surface observations. The average TD at the coastal urban site Kaohsiung is higher than the inland site Fenhua by ~1°C. This spatial TD difference mainly results in a ~10 N-unit difference, whereas the spatial average T field was a nearly homogeneous ~27.5°C. Temperature at Kaohsiung is ~1.1°C higher than at Fenhua, but this spatial T difference only causes a −1.6 N-unit change.
b. The temporal variation of refractivity
The time series of the probability density function (PDF) of radar-derived refractivity shown in Fig. 4 displays the temporal variation of refractivity from 1 to 28 June 2008. The refractivity varies from 350 to 405 N-unit depending on weather conditions across scales from the local circulations to the synoptic weather systems. We can simply put the spatiotemporal characteristics of the refractivity in Fig. 4 into two categories based on the duration of the rainfall. The duration of the rainfall distinguishes the scales of the weather systems and the corresponding spatiotemporal characteristics of the refractivity fields. Longer-rainfall-duration events lasting more than 6 h are dominated by synoptic-scale forcing, and these events are usually identified as intensive observation periods (IOP) in the TiMREX/SoWMEX experiment. Three IOP periods are defined in this study, including 1–5 June (mei-yu front and mesoscale convective system), 14–16 June (mesoscale convective system), and 24–28 June (outer circulation of Typhoon FengSeng). Fair weather or shorter rainfall periods less than 6 h are categorized as non-IOP and these events are influenced by weak synoptic forcing.
The time series of radar-derived refractivity during the TiMREX/SoWMEX experiment. The reddish color shading shows the probability density function (PDF, %) of refractivity at each given radar scan at 0.5° elevation. The temporal resolution is about 7 min. The green histogram is the average hourly rainfall rate (mm h−1) from the surface stations except Juru shown in Fig. 1. The long duration of the rainfall events, lasting more than 6 h, indicates the synoptic forcing weather events, as labeled in gray shading. The local time equals UTC + 8 h.
Citation: Monthly Weather Review 149, 9; 10.1175/MWR-D-20-0425.1
Figure 4 presents different temporal variation characteristics of the refractivity between the IOP and non-IOP events. In the IOP cases, consistent and relatively uniform high values of refractivity (380–400 N-unit) indicate the persistent high water vapor. Larger-scale forcing systems, e.g., synoptic fronts, mesoscale convective systems and typhoons, continuously advect abundant water vapor favorable to sustain heavy precipitation over many hours. The different amount of moisture during IOP cases also depends on the sources of moisture that are modified by physical processes. Higher N values (390–400 N-unit) are observed during the impact of the outer circulation of Typhoon FengShen coming from the south at the end of June, compared with the lower N (380–390 N-unit) associated with the mei-yu frontal system from the north in the beginning of June. In the non-IOP cases, a periodic diurnal cycle of larger refractivity variations (30–40 N-unit) is frequently observed. The local circulations, such as the land–sea breeze and mountain–valley breeze, associated with diurnal heat contrasts and land–atmosphere interactions cause the significant diurnal peaks and valleys and modify the spatial moisture distribution.
The spread of the PDF in Fig. 4 indicates the spatial homogeneity of the refractivity fields. A higher percentage of the PDF suggests spatially uniform distribution of the N, whereas a lower percentage of PDF denotes a more heterogeneous distribution. During the IOP events, a generally high PDF is found during successive precipitation periods, suggesting a spatially homogeneous moisture distribution within the widespread precipitation system. During the non-IOP events, a smaller PDF indicates a heterogeneous distribution of the N field, which may be caused by the evolving moisture gradients across land–sea-breeze fronts or across different land-use types. During IOP events, some abrupt temporal changes and heterogeneous distributions of refractivity occurs over a few hours along with a significant change of rainfall rate (Fig. 4). For instance, the sharp increments of refractivity during the early hours on 2 June indicates a quick passage of the mei-yu front that moisten the relatively dry prestorm environment within an hour. These discontinuous spatiotemporal characteristics of refractivity during the IOP events suggest the arrival or progression of the thermodynamic boundaries, such as fronts, cold pools, and leading edges of precipitation systems.
c. Diurnal characteristics of refractivity of IOP and non-IOP events
We compare the diurnal spatiotemporal characteristics of radar-retrieved refractivity between the IOP and non-IOP events (Figs. 5 and 6). The average sunset is 1845 LST and the sunrise is 0500 LST. In Fig. 5, the average refractivity is persistently higher during the IOP events compared with the non-IOP events. No diurnal signal of the average N is found during the IOP cases, since the large-scale forcing strongly dominates the low-level moisture and overwhelms the effect of diurnal circulations. Meanwhile, a small standard deviation of N during the day and night suggests the relatively homogeneous spatial distribution of moisture. For the non-IOP events, an average diurnal change of N is ~10 N-unit. The average diurnal maximum refractivity occurs in the evening before midnight (2000–2300 LST) while the average diurnal minimum occurs in the morning (0900–1200 LST). The diurnal cycle of the N found here in a coastal environment is opposite from the continental environment due to different dominant physical processes. In the U.S. Great Plains environment under weak synoptic forcing conditions, the diurnal N peaks at dawn and lowers to its minimum in the early evening (Fabry 2006; Weckwerth et al. 2005). The land-atmospheric interaction drives the continental diurnal N variation (Haugland and Crawford 2005; Bodine et al. 2011). The maximum N and higher near-surface moisture at dawn is due to increment of the evapotranspiration process, evaporation of morning dew and the stable surface layer to limits vertical mixing from the surface. The minimum N in the later afternoon is caused by the dry-air entrainment from the upper convective boundary layer, which is strongly offsets the evapotranspiration process. However, in coastal environments, diurnal land–sea and mountain–valley circulations provide an extra source and sink of moisture in addition to the land–atmospheric interaction and boundary layer processes to modify the moisture diurnal change.
The mean and standard deviation of the diurnal cycle of the radar-estimated refractivity calculated from IOP (blue) and non-IOP (green) periods. The IOP periods, defined based on the long-lasting precipitation more than 6 h, include 1–5, 14–16, and 24–28 Jun as identified in Fig. 4. The non-IOP periods include 7–11 and 18–22 Jun. Dates before and after the IOP are not considered in either IOP or non-IOP cases due to the complicated transition pattern affected by both circulations at all scales. Gray shading indicates the nighttime from sunset (1845 LST) until sunrise (0500 LST).
Citation: Monthly Weather Review 149, 9; 10.1175/MWR-D-20-0425.1
The anomaly of refractivity fields with respect to the mean state in (a) the morning (0800–1100 LST), (b) evening (2000–2300 LST), and (c) late night before dawn (0200–0500 LST) during the IOP cases. (d)–(f) As in (a)–(c), but for the non-IOP cases. The black contours are the topographic heights as in Fig. 3.
Citation: Monthly Weather Review 149, 9; 10.1175/MWR-D-20-0425.1
There is also a significant diurnal variation of the standard deviation of the refractivity during the non-IOP events. Smaller standard deviation of the N, representing a spatially homogeneous distribution of moisture, is found in the evening before midnight when the daily maximum refractivity occurs. On the contrast, the most heterogeneous refractivity distribution with larger standard deviation happens during the onset of the sea breezes in the morning.
Figure 6 presents the normalized distribution of the refractivity during the morning, evening and predawn hours. For the IOP cases, a gentle gradient of refractivity is persistently observed during daytime and nighttime (Figs. 6a–c). The barely visible diurnal variation of the N field suggests that the near surface moisture is less affected by the local diurnal circulation during the long-duration precipitation events. The persistent N gradient denotes the nonnegligible impact of the land–sea geographical environment on the near-surface moisture distribution even under synoptic forcing. The weak diurnal signal and the small spatial difference of the N are consistent with the characteristics shown in Fig. 5.
On the other hand, during the non-IOP events, the average gradient of the refractivity changes its orientation from perpendicular to the coastline in the morning (Fig. 6d) to perpendicular to the meridionally oriented mountains during nighttime (Fig. 6f). A weaker spatial gradient of refractivity occurred in the evening (Fig. 6e) showing the transition period between the distinct daytime and predawn patterns. These changes of the moisture gradient (Figs. 6d,f) are related to the diurnal circulations. In the morning, the strong moisture gradient is a result of the high moisture along the sea-breeze front at the coast and the decreasing moisture inland caused by active daytime boundary layer mixing processes. During nighttime, the downslope wind from the meridionally oriented mountains dries the air at the foothills and yields the patch of low refractivity parallel to the mountains. To advance the understanding of the diurnal moisture distribution, more detailed case studies are examined in section 4.
4. Case studies of diurnal evolution refractivity
Under weak synoptic forcing conditions, the spatial distribution of moisture varies diurnally due to the local circulations in the coastal regions. The complex coastal environment with bays, mountains, and different land use results in a myriad of interactions between the local land–sea breezes, mountain–valley–plains flows, and the prevailing large-scale flow. Three types of diurnal refractivity patterns presented here are associated with different local wind conditions under weak synoptic forcing conditions (Fig. 7). On 8 June, Taiwan is on the ridge of the east–west-oriented North Pacific high pressure region (Fig. 7a). The synoptic surface wind in southern Taiwan is south-southeasterly. On 20 June, the North Pacific high pressure moves toward the northeast and the synoptic surface wind in southern Taiwan becomes easterly (Fig. 7b). An afternoon thunderstorm occurs in this case after the sea-breeze front arrived at the foothills. In the third case, 22 June, the synoptic surface wind over our interest area is still due to the low pressure system on the ocean south of Taiwan (Fig. 7c).
The surface weather maps with pressure patterns in black contours (hPa) at 0800 LST (a) 8, (b) 20, and (c) 22 Jun revealed from the ERA-Interim data. The red triangle indicates the refractivity analysis area.
Citation: Monthly Weather Review 149, 9; 10.1175/MWR-D-20-0425.1
a. 8 June case study
Figure 8 displays the diurnal spatial distribution of 3-h mean refractivity fields under the prevailing southerly winds influenced by the North Pacific high pressure system. During the daytime (0800–1700 LST, Figs. 8a–c), high refractivity along the coast and low refractivity inland are persistently observed from morning until early evening. The distribution of the refractivity suggests that the water vapor decreased from the coastline toward the inland regions. The ~20 N-unit difference in N between the coastal and the inland regions indicates ~2.5°C spatial difference in TD given the nearly spatially homogeneous T field. This orientation of N gradient continuously lasts beyond sunset, but the leading edge of the high N gradually penetrates inland in the late afternoon (Fig. 8d). During early evening to midnight (1700–2300 LST, Figs. 8d,e), the values of N continuously increase and reaches their maximum before midnight. Meanwhile, the spatial distribution of N becomes more homogeneous and the spatial difference of the N is reduced to ~10 N-unit. During nighttime until dawn (2300–0500 LST, Figs. 8f,g), the values of N begin decreasing from the foothills and the low moisture air slowly moves westward due to the weak dry downsloping easterly wind. Nighttime refractivity fields further depict a more complicated spatial moisture distribution potentially attributed to the geography and land–atmospheric interaction (Fig. 8g). Some inland patches of higher N are found overlaying the irrigated cropland areas and rivers (Fig. 1c). Relatively higher moisture over these irrigated areas and rivers are associated with the higher evapotranspiration rate that compensates the dry downslope wind compared with the neighboring dryland areas. In the early morning after sunrise (0500–0800 LST, Fig. 8h), the refractivity decreases to the minimum diurnal values and its distribution became more homogeneous after the dry land-breeze and the downslope winds extends toward the coastline.
The diurnal variation of the radar-retrieved refractivity field on 8 Jun with measurements of temperature (superscript next to the surface station, °C), dewpoint temperature (subscript, °C), and wind vectors from surface stations. Each panel shows the 3-h average starting from (a) 0800–1100 to (h) 0500–0800 LST.
Citation: Monthly Weather Review 149, 9; 10.1175/MWR-D-20-0425.1
The surface observations shown in Fig. 9 help to relate the changes of the atmospheric state and physical processes to the refractivity characteristics (Fig. 8). Three prominent refractivity features on 8 June at different phases are discussed below.
The diurnal observation of temperature (red lines), water vapor mixing ratio (blue lines), and wind (reference wind vector of 3 m s−1 at the bottom) on 8 Jun at surface stations from coast to inland: (a) Kaohsiung, (b) Shiwei, and (c) Fenhua. Refractivity (black lines) was calculated from the surface observations. Gray shading denotes the nighttime period.
Citation: Monthly Weather Review 149, 9; 10.1175/MWR-D-20-0425.1
1) The persistent daytime refractivity gradient
The daytime refractivity gradient (Figs. 9a–c) is established between 0800 and 0900 LST through the sudden wind transition from the nighttime weak easterly and northeasterly to moderate southerly winds (Figs. 9a–c). A sharp rise of q of ~3 g kg−1 associated with the wind transition is observed first at the coastal station (Fig. 9a) and at the inland stations less than 1 h later (Figs. 9b,c). This fast transition of wind and thermodynamic fields are not caused by the typical sea-breeze front observed in Taiwan that usually slowly progresses inland starting after 0900 LST and arrives at the foothills around noon (Chen et al. 2014). Instead, the persistent daytime strong southerly winds at all stations, and additionally at night in Kaohsiung, suggest the dominance of the synoptic-scale high pressure system over the sea-breeze circulation induced by the local land–sea heat contrast.
One hypothesis to explain the intriguing daytime refractivity gradient (Figs. 8a–c) is the local kinematic convergence. The southerly wind at Kaohsiung and the southeasterly wind at Shiwei form a consistently narrow convergence zone, which potentially leads to the persistent higher daytime moisture along the coast (Figs. 8, 9). Conversely, the low refractivity inland is caused by the daytime active mixing processes with the dry air above in the convective boundary layer and by the divergence zone depicted from the southeasterly flow at Shiwei and the southwesterly winds at Fenhua. The surface wind at Shiwei plays a critical role to impact the meso-γ-scale convergence and divergence that redistribute the orientation of the N gradient. The local surface winds observed from the surface stations are slightly different from the synoptic winds and are likely due to the complex interaction between the coastline and terrain. Another hypothesis is that additional anthropological sources of water vapor emitted by gasoline factories along the coastline may continuously enhance water vapor within ~10 km from the coast. More data in the boundary layer and numerical simulations are needed to investigate this hypothesis.
2) The evening homogeneous and maximum refractivity field
The rising refractivity after sunset is primarily influenced by the increase of the water vapor mixing ratio. A greater increase in q is observed inland compared to along the coast (Fig. 9). For example, a rise of q by ~2 g kg−1 is observed at Fenhua 3 h after sunset. The increasing moisture is related to the land–atmospheric interaction in addition to the continuous moisture advection from the southerly flow. Latent heat flux caused by evapotranspiration, i.e., evaporation from soil moisture and transpiration from the canopy, may continue even after sunset in the high vegetation and abundant soil moisture availability areas (Pielke and Niyogi 2010). Meanwhile, with the end of the solar heating after sunset, the decreasing surface temperature leads to increase the surface layer stability and the weakened vertical mixing processes inhibit the dry-air entrainment from the upper convective boundary layer. Without the dry-air entrainment to offset the other moisture sources, the surface water vapor continuously increases and reaches a maximum within a few hours after sunset. The increase in surface TD in the evening is also observed in the second peak of TD in summertime central United States due to similar processes (Haugland and Crawford 2005).
3) The nighttime refractivity gradient
From midnight to early morning (Figs. 8f–h), a meridionally oriented low refractivity pattern gradually moves out from the foothills toward the coast. At night, the mountain slopes are cooled by longwave radiation and chilled air sinks downslope due to the negative buoyancy, resulting in a cold dry downslope wind. Surface observations inland (Fig. 9) show that the reduction of the water vapor mixing ratio is affected by the easterly dry downslope wind from the mountains, observed at ~2300 LST at Fenhua and ~0200 LST at Shiwei. A large decrease in q ~ 6 g kg−1 through the night at Fenhua closer to the foothills may be impacted both by stronger downslope wind and the drier soil moisture.
b. 20 June case study
The refractivity fields on 20 June display a typical surface moisture evolution associated with the sea-breeze front followed by an afternoon thunderstorm on the upslope of the mountains. In the morning, the high refractivity boundary moving from the coastline toward the inland region suggests the passage of the moist sea-breeze front (Figs. 10a,b). A diffuse reflectivity fine line, as a classical way to identify the sea-breeze front, is coincident with the refractivity gradient (Figs. 11a,b). Moreover, the N gradient is much easier to be identified than the reflectivity fine lines in complex terrain environments with lots of ground clutter. Around noon, N becomes more homogeneously distributed (Fig. 10b) with the arrival of the sea-breeze front at the foothills in contrast with the persistent large refractivity gradient shown on 8 June (Fig. 8b). Meanwhile, the surface winds indicate a low-level convergence zone close to the foothills at 22°40′N, 120°35′E (Figs. 10b, 11d). This moisture convergence near the foothills may favor the formation of moist convection. Later in the early afternoon at 1345 LST, the radar detects convection initiation with a small area of ~40 dBZ around the foothills (Figs. 11c, d). In an hour, this convective cell develops into a mature storm with an outflow and a high refractivity boundary propagates radially westward out from the storm (Figs. 11e–h). The convective storm lasts for a few hours and dissipates by 1900 LST. A similar N distribution related to the storm activities is also shown in the 3-h average field in Figs. 10c,d. In the evening through midnight (Figs. 10d–f), a meridionally oriented high refractivity pattern forms parallel to the mountains south of the location of the previous storms and then propagates to the coast by the weak easterly flow. After midnight until the morning (Figs. 10f–g), the nighttime spatial variability of refractivity is similar to 8 June (Figs. 8f–g), due to the same physical processes of dry downslope winds and the land-use distribution. The nocturnal N retrieval at Fenhua does not suggest low-enough TD. This may be related to the nighttime inversion and the height representations between two measurements (Fabry 2004; Bodine et al. 2011).
As in Fig. 8, but for 20 Jun.
Citation: Monthly Weather Review 149, 9; 10.1175/MWR-D-20-0425.1
(left) Reflectivity (dBZ) and (right) refractivity observed at (top to bottom) 1100, 1345, 1445 and 1600 LST 20 Jun 2008.
Citation: Monthly Weather Review 149, 9; 10.1175/MWR-D-20-0425.1
1) The evolving moisture gradient associated with the sea-breeze front
In Fig. 12, the refractivity difference within 30-min intervals (ΔN30min) is calculated to investigate the position and the magnitude of the moisture gradient evolving with the sea-breeze front and storm environments (Figs. 12a–d). In the morning (Figs. 12a–c), a band of positive ΔN30min moving from the coast to inland regions denotes the progressing leading edge of the moist sea-breeze front. The sea-breeze front modulates the environment with lower temperature and higher moisture resulting in the positive ΔN30min. It is interesting to note a time lag of the formation of the sea-breeze fronts in the rural and urban coasts. The sea-breeze front formed about 1 h earlier in the southern rural coastal area compared to the northern urban coast. The northwest–southeast-oriented sea-breeze front initiates at the southern coastline (shown by the southern arrow in Fig. 12a), propagates northward and joins together with the later initiated sea breeze in the northern urban area (the northern arrow in Fig. 12a). Then, the joint sea-breeze orients parallel to the coast and moves inland (Figs. 12b,c). The time difference of the formation of the sea-breeze front in urban and rural areas is consistent with previous numerical modeling studies (Kusaka et al. 2001; Keeler and Kristovich 2012). The rapid increase of temperature in the rural area during early morning causes higher temperature than the urban areas (Runnalls and Oke 2000, Yang et al. 2017). The stronger thermal land–sea contrast thus creates larger pressure gradients to drive the earlier initiation of the sea-breeze fronts in rural areas. Larger buildings in the urban areas exert stronger drag over the wind and may also retard the development and movement of the sea-breeze fronts (Yoshikado 1992; Ohashi and Kida 2002).
The 30-min refractivity difference (ΔN30min) on 20 Jun. (a)–(i) The morning and afternoon ΔN30min from 1000–0930 to 1600–1530 LST with 1-h intervals. (j)–(l) The evening ΔN30min for 1900–1830, 2100–2030, and 0000–2300 LST, respectively. The bold cyan contours indicate the location of afternoon thunderstorms with reflectivity > 30 dBZ. The green arrows highlight the propagation of the intensifying refractivity gradients of the (a)–(c) morning sea-breeze front and (j)–(l) the evening obscure moisture gradient.
Citation: Monthly Weather Review 149, 9; 10.1175/MWR-D-20-0425.1
The magnitude of the positive ΔN30min along the sea-breeze front intensifies as it moves inland. It suggests a larger contrast of moisture between the sea-breeze front and the environment close to the foothills (Figs. 12b,c). The negative ΔN30min inland ahead of the sea-breeze front also denotes a drier condition due to the intensifying daytime solar radiation and the boundary layer mixing processes. With the arrival of the front, around 1230 LST at Fenhua (Fig. 13c), a decrease in T ~ 1°C and an increase in q of ~2.5 g kg−1 results in a significant rise of the refractivity ~20 N-unit. The increasing width of ΔN30min band further indicates the increasing propagation speed of the sea-breeze front as it responds to the larger thermodynamic contrast inland around noon (Figs. 12b–d). There are some smaller-scale refractivity lines and patches, which are associated with randomly distributed formation of clouds after the sea-breeze front passage. The ΔN30min fields provide the magnitude of the moisture gradient across the sea-breeze front to complement the fine line shown on the N field that only detects the location of the boundary (Fig. 11a).
As in Fig. 9, but for 20 Jun.
Citation: Monthly Weather Review 149, 9; 10.1175/MWR-D-20-0425.1
2) The refractivity retrieval in the storm environment
A patch of positive ΔN30min is observed around 22°40′N, 120°35′E along the foothills (Figs. 12e–i) near Fenhua in Figs. 12d,e, suggesting increasing moisture accumulated at the foothills about 1 h ahead of convection initiation (CI, Fig. 11c) due to the combination of the arrival of the sea-breeze front and the low-level convergence. When the sea breeze approaches the foothills, the ΔN30min at the foothills increases rapidly up to 20 N-unit (30 min)−1 (Figs. 12c,d). When the sea breeze arrives at the foothills (Fig. 12e), the area of positive ΔN30min is not as large as the previous time, which suggests the environment has been modified by the sea breeze and the moisture tendency at the foothills increases slowly. A higher moisture anomaly inland and at the foothills has been usually detected 1–3 h prior to the CI in Taiwan (Lin et al. 2011; Chen et al. 2014). The lag time between the positive moisture anomaly and consequent afternoon thunderstorm formation suggests that the N and the ΔN30min fields inland near the foothills may be used as one of the precursors for nowcasting CI, along with wind and temperature fields. Besides the higher moisture accumulated inland in this case, CI occurs accompanied with the low-level wind convergence (Fig. 11d), relative moister boundary layer (Fig. 14), the large atmospheric instability (convective available potential energy, CAPE, ~4694 J kg−1), and terrain lifting. The high moisture along the foothills south of the CI in Fig. 12d may not have had other favorable kinematic conditions, such as topographic trigger and mass convergence, for development.
Upper-air soundings at 1400 LST (0600 UTC) 8 Jun (in red) and 20 Jun (in dark blue) at Pingtung near the S-Pol radar. Plotted are temperature (solid line, °C), dewpoint temperature (dashed line, °C), and wind profiles (kt).
Citation: Monthly Weather Review 149, 9; 10.1175/MWR-D-20-0425.1
For the mature storm stage (Figs. 12g–i), the cold outflow propagating outward toward the west is detected as a positive ΔN30min circular shaped region surrounding the storm edge. The Fenhua surface station shows the N increasing ~10 N-unit about 1430 LST, primarily due to the large ~6°C decrease in T instead of the small increase of 0.5 g kg−1 in q (Fig. 13c). The storm outflow boundary shown as the positive ΔN30min band propagates westward instead of southward (Fig. 12i). The land-use contributing to drier soil moisture of the dryland and urban areas toward the west in contrast with the moist soil of the southern irrigated cropland (Fig. 1c) may be one of the reasons that contributed to the different propagation characteristics of the storm cold pool. Idealized simulations show that regions with dry soil moisture lead to a deeper drier subcloud layer in the daytime that enhances the evaporation cooling of precipitation and yields stronger and longer-lasting cold pools (Drager et al. 2020).
3) The obscure evening propagating moisture gradient
A positive ΔN30min band slowly propagates from the foothills westward starting ~1830 LST (Figs. 12j–l) and reached the coastline around midnight (arrows in Figs. 12j–l). A weak easterly wind observed from the radar radial wind (not shown) associated with the land breeze and down-slope wind may push the boundary westward. The surface observations suggest that the positive ΔN30min was caused by the increase of q. It is unclear what causes the higher q at the foothills in addition to the evapotranspiration processes. Profiling observations and numerical weather prediction model simulations are further required for explaining the formation of the increasing moisture.
c. 22 June case study
Unlike the previous two cases, the daytime refractivity field on 22 June (Figs. 15a,b) displayed a predominantly meridionally oriented distribution similar to the typical nighttime refractivity pattern (Fig. 6f). The northerly and northwesterly winds associated with the outer circulation of the low pressure system over the ocean south of Taiwan transport drier air from the northern hills into the region and yields the unusual daytime low N values. Without the cool moist sea breeze to modulate the inland environment in the afternoon, the T at Shiwei and Fenhua reaches ~35°C around 1330 and 1530 LST, respectively (not shown). A 4°C T difference exists between the inland and coastal stations in the afternoon, which is the greatest among all of the weak synoptic forcing cases. This thermal contrast further induces local surface flows from the coast to the foothills. The combination of local thermally induced flow and the low pressure system moving toward the north in the afternoon causes the inland surface wind to gradually change from northerly toward southwesterly (1400–1700 LST, Fig. 15c). Moisture is thus slowly transported inland through the westerlies in the late afternoon. The evolution of the nighttime N distribution (Figs. 15d–h) is similar to the previous cases on 8 and 20 June. There is a homogeneous N distribution in the evening transition phase (Figs. 15d,e) and a westward progression of the N gradient perpendicular to the mountains at night (Figs. 15f,g). But, the low N boundary due to the dry downslope wind is stalled by the hills around 120°30′E (Figs. 15g,h).
As in Fig. 8, but for 22 Jun.
Citation: Monthly Weather Review 149, 9; 10.1175/MWR-D-20-0425.1
5. Discussion of the diurnal variation of refractivity
A Hovmöller diagram of the refractivity fields along an east–west cross section, from the coast to inland, presents the diurnal moisture variation in three distinct phases within a day: daytime, evening transition, and nighttime until early morning (Fig. 16). The most different spatiotemporal characteristics among the cases occur during the daytime, but they are similar during the evening transition and nighttime periods. The daytime N in Fig. 16 shows the difference of how far from the coast the water vapor penetrated inland among the three cases. Only on 20 June, the high N is transported to the foothills with a propagation speed of ~17 km h−1 by the southwesterly sea-breeze front (red arrow in Fig. 16b). On 8 and 22 June, high moisture is accumulated along the coastline and not advected to the foothills before noon. The complicated local surface winds affected by the synoptic and local circulations plays a critical role in modulating the moisture distribution. Under the prevailing moderate southerly wind on 8 June, the moisture distribution occurs with a large spatial contrast of N ~ 20 N-unit (~5 g kg−1 in q) that stalls ~10 km from the coastline. Under the northerly and northwesterly flows on 22 June, the N contrast increases with time due to the continuous dry-air advection and daytime boundary layer mixing processes. The moisture advection is delayed and finally propagates inland when the wind direction changes to southwesterly in late afternoon (red arrow in Fig. 16c).
Hovmöller diagram of the relative refractivity change with respect to the daily mean of each given day on (a) 8, (b) 20, and (c) 22 Jun of the east–west cross section about 6 km north of the S-Pol radar (over the Shiwei surface station). The x axis is the east–west distance with respect to the radar. The arrow on the y axis indicates the time sequence. The dark red and dark blue arrows in the plot denote the sea breeze and downslope winds, respectively. The gray dashed arrow shows the intriguing moist advection from the foothills toward the coast.
Citation: Monthly Weather Review 149, 9; 10.1175/MWR-D-20-0425.1
The distance that high moisture advects inland may affect the likelihood of afternoon CI. High moisture content is a necessary condition for afternoon thunderstorms, in addition to topographic lifting near the foothills and instability. The surface q at the foothills (Fenhua) is ~2 g kg−1 higher in the early afternoon of 20 June than it is on 8 June, while their T is similar at 32°C (Figs. 9c and 13c). Meanwhile, the radar-retrieved refractivity during 1300–1400 LST (before CI) also shows ~10–15 N-unit greater on 20 June compared to 8 June. Crook (1996) shows that 1 g kg−1 variation in q (~4.2 N-unit) at the surface could make a difference between deep convection and no convection. Yet, from the sounding site between the S-Pol and Fenhua (Fig. 1b) at 1400 LST 8 and 20 June (Fig. 14), the CAPE (>4000 J kg−1) and CIN (~0 J kg−1) are both favorable for developing convection. The N fields provide a valuable qualitative tool to examine the magnitude and the spatial homogeneity of near-surface moisture. Integrating an observational thermodynamic profiling network along with the N field would aid in advancing our understanding of the roles of the surface, low-level and midlevel moisture distribution in storm environments and their impact on CI through adjusting the lifted condensation levels and instabilities.
During the evening transition and nighttime periods, the N patterns are generally similar among the cases (Fig. 16). The evening enhancement of moisture is contributed by the remnant evapotranspiration processes from the land–atmospheric interaction, weakened boundary layer mixing processes, and by the moisture advection from late-afternoon sea breeze. Some relatively lower N shown on 20 June (Fig. 16b) is affected by the afternoon storm outflows. Furthermore, a similar intriguing feature is repeatedly shown in all cases, a high N slowly propagates by weak easterly winds from the foothills starting ~2200 LST and reaches the coastline before dawn (gray dashed arrows in Fig. 16). After midnight until early morning, a consistent decreasing trend of N from the foothills to the coast (blue arrows in Fig. 16) is caused by the dry downslope wind. Several other days with weak synoptic forcing present similar diurnal variations of the N fields as the case studies shown in section 4. Thus, we summarized all of these N variations into three types based on the daytime surface winds and the synoptic conditions. First, the southerly wind category (S) includes 7–10 June, while Taiwan is on the ridge of the Pacific high pressure system. Second, the southwesterly category (SW) includes 18–20 June while the Pacific high pressure system is located further northeast. Third, the northerly and northwesterly category (NW) includes 6, 11–12, and 21–22 June due to the outer circulation of the low pressure system surrounding Taiwan or dry surface fronts.
In Fig. 17, similar diurnal signals are observed including the maximum refractivity at night and minimum during the day. However, the extreme values of the diurnal variations are different among these categories. The diurnal extremes of N in the NW category occur later than the S and SW cases, where their maximum and minimum N values are found in the evening and early morning, respectively. In the SW category, afternoon storms occur in the foothills and their morning N is higher and more homogeneous with smaller standard deviation than the other two categories. Higher moisture set up a favorable condition for the afternoon moist convection.
Mean and standard deviation of the diurnal variation of refractivity under three types of non-IOP categories depending on the daytime local wind direction: southerly cases (S, green), southwesterly cases (SW, red), and northerly/northwesterly (NW, yellow).
Citation: Monthly Weather Review 149, 9; 10.1175/MWR-D-20-0425.1
The consistent nocturnal refractivity patterns among the non-IOP cases are averaged (Fig. 18) to show how the spatial variability of lower N values evolves with topography and land use. Around midnight with the onset of the dry downslope winds, the initial decrease in N at the foothills is elongated along the river valleys embedded within the eastern mountains (arrows in Fig. 18b). Later, during 0200–0800 LST (Figs. 18c,d), the N values became lower over the dry land compared with it over the irrigated cropland. The statistical PDF of N and mean N further showed lower values in dryland than the irrigated croplands starting after 0200 LST (Figs. 19a,b).
The nocturnal mean refractivity fields average from all non-IOP cases, namely, (a) 2000–2300, (b) 2300–0200, (c) 0200–0500, and (d) 0500–0800 LST. The gray circles indicate the current automated surface weather observing stations.
Citation: Monthly Weather Review 149, 9; 10.1175/MWR-D-20-0425.1
The PDF of nocturnal refractivity over different land use, dryland (a) and irrigated cropland (b), calculated from the inland region (red box in Fig. 18) within different time interval shown in different colors. The dashed lines are the mean values of each PDF function.
Citation: Monthly Weather Review 149, 9; 10.1175/MWR-D-20-0425.1
6. Conclusions
The radar-retrieved refractivity data over 4 weeks in southern Taiwan are analyzed to depict the morphology of the diurnal near-surface moisture distribution in the summertime tropical coastal areas. This is the first dataset of radar-refractivity retrieval technique applied in the tropics to explore refractivity variations associated with coastal processes, complex terrain, and land use. On average, the moisture is persistently lower at the foothills and higher along the coast. A refractivity gradient is found perpendicular to the coastline showing a 1°C difference in dewpoint temperature over 30 km in distance. A similar spatial moisture gradient is also observed during the long-lasting rainfall events all day long, suggesting that the impact of the land–sea geographic environment on the low-level moisture distribution is strong even in the presence of the synoptic-scale forcing. Under weak synoptic forcing with no precipitation, the refractivity field quantitatively denotes the gradient of the moisture distribution which evolves with the recurring diurnal land–sea circulations. The maximum refractivity occurs in the evening and the minimum occurs during the daytime. The diurnal cycle of the refractivity in the coastal environment is nearly opposite from that in the semiarid continental environment shown in previous studies (Fabry 2006; Weckwerth et al. 2005; Bodine et al. 2010).
The diurnal cycle of the radar refractivity retrieval includes three distinct phases: daytime, evening transition, and nighttime. The daytime refractivity fields vary with the local surface wind, dictated by the combination of the synoptic prevailing wind and the local sea-breeze circulation. Under local southwesterly winds, the high moisture is distributed spatially uniformly after the moisture is transported to the foothills by the sea-breeze front around noon. High moisture accumulates at the foothills, thereby reducing the thermodynamic inhibition and favoring the formation of moist convection in the afternoon along the slopes of the foothills. On the mornings with no sea-breeze fronts, the daytime refractivity fields are heterogeneously distributed with persistent mixing ratio contrasts of >3 g kg−1 caused by the local southerly or northerly winds. The daytime uneven moisture fields may result in a spatially inhomogeneous distribution of cloud formation and atmospheric instability, which consequently affects the formation and the location of afternoon moist convection. During the evening transition period, the refractivity field typically has a homogeneous distribution caused by the weakening boundary layer mixing and land-atmosphere evapotranspiration processes. During the nighttime, the dry downslope winds often cause a decrease in the refractivity values slowly from the foothills toward the coast, forming a refractivity gradient perpendicular to the meridionally oriented mountains.
The meso-γ-scale moisture variability revealed from the radar-derived refractivity strongly complements the limited number of surface station observations (Fig. 18). Between the surface station measurements, the radar-retrieved refractivity depicts detailed uneven moisture distributions related to the land-use, such as the lagged initiation and propagations of the daytime sea-breeze moisture fronts between the rural and urban coasts, and the evolving nighttime heterogeneous refractivity over river valleys, irrigated cropland and the drylands. Even though refractivity retrievals allow for limited areal coverage, this study demonstrates that the refractivity field is valuable in complex environments over heterogeneous land-use types, such as highly populated coastal urban areas. In addition, the evolving refractivity distributions in convective environments and under drought conditions are worthy of further investigation. The operational WSR-88D network should implement this advanced refractivity measurement technique to obtain the detailed mesoscale thermodynamic fields (Bluestein et al. 2014). Applying the refractivity technique to operational radars would allow forecasters to better monitor and understand day-to-day large- and small-scale variation in local moisture. The refractivity from the operational radars would help to further explore the potential value of assimilating refractivity retrievals on different cases in various environments, as suggested by WMO (2021).
Acknowledgments
The material is based upon work supported by the National Center for Atmospheric Research, which is a major facility sponsored by the National Science Foundation under Cooperative Agreement No. 1852977. The authors appreciate the insightful comments from the editor David Bodine and two anonymous reviewers. We also thank Jim Wilson and John Hubbert for the pre-submission reviews. The contribution of Hsu and Lin are supported by the Ministry of Science and Technology (MOST) of Taiwan under MOST 107-2111-M-008-038, MOST 108-2111-M-008-028, and MOST 104-2923-M-008-003-MY5. The contributions of Liou and Wang are funded by the under MOST 108-2111-M-008-040 and MOST 109-2111-M-008-012.
Data availability statement
The data presented in this manuscript are available at NCAR data repository at https://doi.org/10.5065/D6WH2N9T (for S-Pol refractivity), https://doi.org/10.5065/D6RJ4GN3 (for other S-Pol observations), https://doi.org/10.26023/VD76-2S0X-E10V (for surface stations), and https://data.eol.ucar.edu/dataset/119.013 (for soundings).
REFERENCES
Arritt, R. W., 1993: Effects of the large-scale flow on characteristics features of the sea breeze. J. Appl. Meteor. Climatol., 32, 116–125, https://doi.org/10.1175/1520-0450(1993)032<0116:EOTLSF>2.0.CO;2.
Atkins, N. T., R. M. Wakimoto, and T. M. Weckwerth, 1995: Observations of the sea breeze front during CaPE. Part II: Dual-Doppler and aircraft analysis. Mon. Wea. Rev., 123, 944–969, https://doi.org/10.1175/1520-0493(1995)123<0944:OOTSBF>2.0.CO;2.
Baker, R. D., B. H. Lynn, A. Boone, W. Tao, and J. Simpson, 2001: The influence of soil moisture, coastline curvature, and land-breeze circulations on sea-breeze-initiated precipitation. J. Hydrometeor., 2, 193–211, https://doi.org/10.1175/1525-7541(2001)002<0193:TIOSMC>2.0.CO;2.
Banacos, P. C., and D. M. Schultz, 2005: The use of moisture flux convergence in forecasting convective initiation: Historical and operational perspectives. Wea. Forecasting, 20, 351–366, https://doi.org/10.1175/WAF858.1.
Bean, B. R., and E. J. Dutton, 1966: Radio Meteorology. National Bureau of Standards Monogr., No. 92, U.S. Government Printing Office, 435 pp.
Bergemann, M., and C. Jakob, 2016: How important is tropospheric humidity for coastal rainfall in the tropics? Geophys. Res. Lett., 43, 5860–5868, https://doi.org/10.1002/2016GL069255.
Besson, L., and Coauthors, 2016: Comparison of real-time refractivity measurements by radar with automatic weather stations, AROME-WMED and WRF forecast simulations during SOP1 of the HyMeX campaign. Quart. J. Roy. Meteor. Soc., 142, 138–152, https://doi.org/10.1002/qj.2799.
Bluestein, H. B., and Coauthors, 2014: Radar in atmospheric sciences and related research: Current systems, emerging technology, and future needs. Bull. Amer. Meteor. Soc., 95, 1850–1861, https://doi.org/10.1175/BAMS-D-13-00079.1.
Bodine, D., P. L. Heinselman, B. L. Cheong, R. D. Palmer, and D. Michaud, 2010: A case study on the impact of moisture variability on convection initiation using radar refractivity retrievals. J. Appl. Meteor. Climatol., 49, 1766–1778, https://doi.org/10.1175/2010JAMC2360.1.
Bodine, D., and Coauthors, 2011: Understanding radar refractivity: Sources of uncertainty. refractivity retrievals. J. Appl. Meteor. Climatol., 50, 2543–2560, https://doi.org/10.1175/2011JAMC2648.1.
Boybeyi, Z., and S. Raman, 1992: A three-dimensional numerical sensitivity study of convection over the Florida peninsula. Bound.-Layer Meteor., 60, 325–359, https://doi.org/10.1007/BF00155201.
Brotzge, J. A., and Coauthors, 2020: A technical overview of the New York State Mesonet standard network. J. Atmos. Oceanic Technol., 37, 1827–1845, https://doi.org/10.1175/JTECH-D-19-0220.1.
Buban, M. S., C. L. Ziegler, E. N. Rasmussen, and Y. P. Richardson, 2007: The dryline on 22 May 2002 during IHOP: Ground-radar and in situ data analyses of the dryline and boundary layer evolution. Mon. Wea. Rev., 135, 2473–2505, https://doi.org/10.1175/MWR3453.1.
Chen, G., R. Lan, W. Zeng, H. Pan, and W. Li, 2018: Diurnal variations of rainfall in surface and satellite observations at the monsoon coast (South China). J. Climate, 31, 1703–1724, https://doi.org/10.1175/JCLI-D-17-0373.1.
Chen, T., M. Yen, J. Tsay, C. Liao, and E. S. Takle, 2014: Impact of afternoon thunderstorms on the land–sea breeze in the Taipei basin during summer: An experiment. J. Appl. Meteor. Climatol., 53, 1714–1738, https://doi.org/10.1175/JAMC-D-13-098.1.
Chen, X., F. Zhang, and K. Zhao, 2016: Diurnal variations of the land–sea breeze and its related precipitation over South China. J. Atmos. Sci., 73, 4793–4815, https://doi.org/10.1175/JAS-D-16-0106.1.
Chen, X., F. Zhang, and K. Zhao, 2017: Influence of monsoonal wind speed and moisture content on intensity and diurnal variations of the mei-yu season coastal rainfall over South China. J. Atmos. Sci., 74, 2835–2856, https://doi.org/10.1175/JAS-D-17-0081.1.
Chen, Y., and A. J. Nash, 1994: Diurnal variation of surface airflow and rainfall frequencies on the island of Hawaii. Mon. Wea. Rev., 122, 34–56, https://doi.org/10.1175/1520-0493(1994)122<0034:DVOSAA>2.0.CO;2.
Cheng, F., Y. Hsu, P. Lin, and T. Lin, 2013: Investigation of the effects of different land use and land cover patterns on mesoscale meteorological simulations in the Taiwan area. J. Appl. Meteor. Climatol., 52, 570–587, https://doi.org/10.1175/JAMC-D-12-0109.1.
Creese, C. W., 1999: Meteorological interpretation of near-surface refractivity measurements. M.Sc. thesis, Department of Atmospheric and Oceanic Sciences, McGill University, 83 pp.
Crook, N. A., 1996: Sensitivity of moist convection forced by boundary layer processes to low-level thermodynamic fields. Mon. Wea. Rev., 124, 1767–1785, https://doi.org/10.1175/1520-0493(1996)124<1767:SOMCFB>2.0.CO;2.
Davis, C. A., and W. Lee, 2012: Mesoscale analysis of heavy rainfall episodes from SoWMEX/TiMREX. J. Atmos. Sci., 69, 521–537, https://doi.org/10.1175/JAS-D-11-0120.1.
Drager, A. J., L. D. Grant, and S. C. van den Heever, 2020: Cold pool responses to changes in soil moisture. J. Adv. Model. Earth Syst., 12, e2019MS001922, https://doi.org/10.1029/2019MS001922.
Fabry, F., 2004: Meteorological value of ground target measurements by radar. J. Atmos. Oceanic Technol., 21, 560–573, https://doi.org/10.1175/1520-0426(2004)021<0560:MVOGTM>2.0.CO;2.
Fabry, F., 2006: The spatial variability of moisture in the boundary layer and its effect on convection initiation: Project-long characterization. Mon. Wea. Rev., 134, 79–91, https://doi.org/10.1175/MWR3055.1.
Fabry, F., and C. Pettet, 2002: A primer to the interpretation of refractivity imagery during IHOP 2002. IHOP 2002 Refractivity Manual, 11 pp.
Fabry, F., C. Frush, I. Zawadzki, and A. Kilambi, 1997: On the extraction of near-surface index of refraction using radar phase measurements from ground targets. J. Atmos. Oceanic Technol., 14, 978–987, https://doi.org/10.1175/1520-0426(1997)014<0978:OTEONS>2.0.CO;2.
Feng, Y.-C., and F. Fabry, 2018: Quantifying the information of radar-estimated refractivity by dual-polarimetric and multiple elevation data. J. Atmos. Oceanic Technol., 35, 1897–1911, https://doi.org/10.1175/JTECH-D-18-0008.1.
Gasperoni, N. A., M. Xue, R. D. Palmer, and J. Gao, 2013: Sensitivity of convective initiation prediction to near-surface moisture when assimilating radar refractivity: Impact tests using OSSEs. J. Atmos. Oceanic Technol., 30, 2281–2302, https://doi.org/10.1175/JTECH-D-12-00038.1.
Grant, L. D., and S. C. van den Heever, 2014: Aerosol-cloud-land surface interactions within tropical sea breeze convection. J. Geophys. Res. Atmos., 119, 8340–8361, https://doi.org/10.1002/2014JD021912.
Hardesty, R., and Coauthors, 2012: Thermodynamic profiling technologies workshop report to the National Science Foundation and the National Weather Service. NCAR Tech. Note NCAR/TN-488+STR, 80 pp., https://doi.org/10.5065/D6SQ8XCF.
Haugland, M. J., and K. C. Crawford, 2005: The diurnal cycle of land–atmosphere interactions across Oklahoma’s winter wheat belt. Mon. Wea. Rev., 133, 120–130, https://doi.org/10.1175/MWR-2842.1.
Hsu, H.-W., 2019: The characteristics and application of radar refractivity during 2008 SoWMEX (in Chinese). M.Sc. thesis, National Central University, 123 pp., https://etd.lib.nctu.edu.tw/cgi-bin/gs32/ncugsweb.cgi?o=dncucdr&s=id=%22GC106621004%22.&searchmode=basic.
Hubbert, J. C., J. W. Wilson, T. M. Weckwerth, S. M. Ellis, M. Dixon, and E. Loew, 2018: S-Pol’s polarimetric data reveal detailed storm features (and insect behavior). Bull. Amer. Meteor. Soc., 99, 2045–2060, https://doi.org/10.1175/BAMS-D-17-0317.1.
Johnson, R. H., and J. F. Bresch, 1991: Diagnosed characteristics of precipitation systems over Taiwan during the May–June 1987 TAMEX. Mon. Wea. Rev., 119, 2540–2557, https://doi.org/10.1175/1520-0493(1991)119<2540:DCOPSO>2.0.CO;2.
Keeler, J. M., and D. A. R. Kristovich, 2012: Observations of urban heat island influence on lake-breeze frontal movement. J. Appl. Meteor. Climatol., 51, 702–710, https://doi.org/10.1175/JAMC-D-11-0166.1.
Koch, S. E., W. Feltz, F. Fabry, M. Pagowski, B. Geerts, K. M. Bedka, D. O. Miller, and J. W. Wilson, 2008: Turbulent mixing processes in atmospheric bores and solitary waves deduced from profiling systems and numerical simulation. Mon. Wea. Rev., 136, 1373–1400, https://doi.org/10.1175/2007MWR2252.1.
Kusaka, H., and Coauthors, 2001: A simple single-layer urban canopy model for atmospheric models: Comparison with multi-layer and slab models. Bound.-Layer Meteor., 101, 329–358, https://doi.org/10.1023/A:1019207923078.
Laird, N. F., D. A. Kristovich, R. M. Rauber, H. T. Ochs, and L. J. Miller, 1995: The Cape Canaveral sea and river breezes: Kinematic structure and convective initiation. Mon. Wea. Rev., 123, 2942–2956, https://doi.org/10.1175/1520-0493(1995)123<2942:TCCSAR>2.0.CO;2.
Lin, P.-F., P.-L. Chang, B. J.-D. Jou, J. W. Wilson, and R. D. Roberts, 2011: Warm season afternoon thunderstorm characteristics under weak synoptic-scale forcing over Taiwan Island. Wea. Forecasting, 26, 44–60, https://doi.org/10.1175/2010WAF2222386.1.
Lindsey, D. T., D. Bikos, and L. Grasso, 2018: Using the GOES-16 split window difference to detect a boundary prior to cloud formation. Bull. Amer. Meteor. Soc., 99, 1541–1544, https://doi.org/10.1175/BAMS-D-17-0141.1.
Madaus, L. E., and G. J. Hakim, 2016: Observable surface anomalies preceding simulated isolated convective initiation. Mon. Wea. Rev., 144, 2265–2284, https://doi.org/10.1175/MWR-D-y15-0332.1.
Miller, S. T. K., B. D. Keim, R. W. Talbot, and H. Mao, 2003: Sea breeze: Structure, forecasting, and impacts. Rev. Geophys., 41, 1011, https://doi.org/10.1029/2003RG000124.
Montmerle, T., A. Caya, and I. Zawadzki, 2002: Short-term numerical forecasting of a shallow storms complex using bistatic and single-Doppler radar data. Wea. Forecasting, 17, 1211–1225, https://doi.org/10.1175/1520-0434(2002)017<1211:STNFOA>2.0.CO;2.
National Research Council, 2009: Observing Weather and Climate from the Ground Up: A Nationwide Network of Networks. The National Academies Press, https://doi.org/10.17226/12540.
Nguyen, H., A. Protat, V. Kumar, S. Rauniyar, M. Whimpey, and L. Rikus, 2015: A regional forecast model evaluation of statistical rainfall properties using the CPOL radar observations in different precipitation regimes over Darwin, Australia. Quart. J. Roy. Meteor. Soc., 141, 2337–2349, https://doi.org/10.1002/qj.2525.
Ohashi, Y., and H. Kida, 2002: Local circulations developed in the vicinity of both coastal and inland urban areas: A numerical study with a mesoscale atmospheric model. J. Appl. Meteor., 41, 30–45, https://doi.org/10.1175/1520-0450(2002)041<0030:LCDITV>2.0.CO;2.
Park, J. M., and Coauthors, 2020: Environmental controls on tropical sea breeze convection and resulting aerosol redistribution. J. Geophys. Res. Atmos., 125, e2019JD031699, https://doi.org/10.1029/2019JD031699.
Pielke, R. A., Sr., and D. Niyogi, 2010: The role of landscape processes within the climate system. Landform—Structure, Evolution, Process Control: Proc. of the Int. Symp. on Landforms Organized by the Research Training Group 437, J. C. Otto and R. Dikaum, Eds., Lecture Notes in Earth Sciences. Vol. 115, Springer, 67–85.
Roberts, R. D., and Coauthors, 2008: REFRACTT 2006: Real-time retrieval of high-resolution, low-level moisture fields from operational NEXRAD and research radars. Bull. Amer. Meteor. Soc., 89, 1535–1548, https://doi.org/10.1175/2008BAMS2412.1.
Runnalls, K. E., and T. R. Oke, 2000: Dynamics and controls of the near-surface heat island of Vancouver, British Columbia. Phys. Geogr., 21, 283–304, https://doi.org/10.1080/02723646.2000.10642711.
Ryu, Y., J. A. Smith, E. Bou-Zeid, and M. L. Baeck, 2016: The influence of land surface heterogeneities on heavy convective rainfall in the Baltimore–Washington metropolitan area. Mon. Wea. Rev., 144, 553–573, https://doi.org/10.1175/MWR-D-15-0192.1.
Seko, H., E.-I. Sato, H. Yamauchi, and T. Tsuda, 2017: Data assimilation experiments of refractivity observed by JMA operational radar. Data Assimilation for Atmospheric, Oceanic and Hydrologic Applications, S. K. Park and L. Xu, Eds., Vol. III. Springer, 327–336.
Soderholm, J., H. McGowan, H. Richter, K. Walsh, T. Weckwerth, and M. Coleman, 2016: The Coastal Convective Interactions Experiment (CCIE): Understanding the role of sea breezes for hailstorm hotspots in eastern Australia. Bull. Amer. Meteor. Soc., 97, 1687–1698, https://doi.org/10.1175/BAMS-D-14-00212.1.
Stephens, G. L., and Coauthors, 2010: Dreary state of precipitation in global models. J. Geophys. Res., 115, D24211, https://doi.org/10.1029/2010JD014532.
Sun, J., 2005: Convective-scale assimilation of radar data: Progress and challenges. Quart. J. Roy. Meteor. Soc., 131, 3439–3463, https://doi.org/10.1256/qj.05.149.
UCAR/NCAR–Earth Observing Laboratory, 2011: S-Pol radar, derived particle ID, DORADE format, Version 1.0. UCAR/NCAR–Earth Observing Laboratory, accessed 6 December 2020, https://doi.org/10.5065/D6CC0XWZ.
Wakimoto, R. M., and H. V. Murphey, 2010: Frontal and radar refractivity analyses of the dryline on 11 June 2002 during IHOP. Mon. Wea. Rev., 138, 228–241, https://doi.org/10.1175/2009MWR2991.1.
Wang, C.-C., and D. J. Kirshbaum, 2015: Thermally forced convection over a mountainous tropical island. J. Atmos. Sci., 72, 2484–2506, https://doi.org/10.1175/JAS-D-14-0325.1.
Wang, S., and A. H. Sobel, 2017: Factors controlling rain on small tropical islands: Diurnal cycle, large-scale wind speed, and topography. J. Atmos. Sci., 74, 3515–3532, https://doi.org/10.1175/JAS-D-16-0344.1.
Weckwerth, T. M., C. R. Pettet, F. Fabry, S. J. Park, M. A. LeMone, and J. W. Wilson, 2005: Radar refractivity retrieval: Validation and application to short-term forecasting. J. Appl. Meteor., 44, 285–300, https://doi.org/10.1175/JAM-2204.1.
Wilson, J. W., and D. L. Megenhardt, 1997: Thunderstorm initiation, organization, and lifetime associated with Florida boundary layer convergence lines. Mon. Wea. Rev., 125, 1507–1525, https://doi.org/10.1175/1520-0493(1997)125<1507:TIOALA>2.0.CO;2.
WMO, 2021: Requirements defined for nowcasting and very short range forecasting (VSRF). WMO, accessed 30 July 2021, https://space.oscar.wmo.int/applicationareas/view/nowcasting_vsrf.
Wulfmeyer, V., and Coauthors, 2015: A review of the remote sensing of lower tropospheric thermodynamic profiles and its indispensable role for the understanding and the simulation of water and energy cycles. Rev. Geophys., 53, 819–895, https://doi.org/10.1002/2014RG000476.
Yang, X., and Coauthors, 2017: Contribution of urbanization to the increase of extreme heat events in an urban agglomeration in east China. Geophys. Res. Lett., 44, 6940–6950, https://doi.org/10.1002/2017GL074084.
Yano, J.-I., M. Bister, Ž. Fuchs, L. Gerard, V. T. J. Phillips, S. Barkidija, and J.-M. Piriou, 2013: Phenomenology of convection-parameterization closure. Atmos. Chem. Phys., 13, 4111–4131, https://doi.org/10.5194/acp-13-4111-2013.
Yoshikado, 1992: Numerical study of the daytime urban effect and its interaction with the sea breeze. J. Appl. Meteor., 31, 1146–1164, https://doi.org/10.1175/1520-0450(1992)031<1146:NSOTDU>2.0.CO;2.
Zhong, S., J. M. Leone, and E. S. Takle, 1991: Interaction of the sea breeze with a river breeze in an area of complex coastal heating. Bound.-Layer Meteor., 56, 101–139, https://doi.org/10.1007/BF00119964.