1. Introduction
In 2008 humanity reached a landmark point in history. For the first time, more than half of the world’s population, over 3.3 billion people, inhabited urban areas. This trend is expected to continue so that by 2030 the number of people living in urban areas could reach 5 billion (UNPF 2007). Humankind’s ever expanding urban footprint has broad-ranging impacts on many aspects of the environment including surface energetics, water and carbon cycle processes, weather, climate, and ecosystem development.
Concurrent with this increase in urbanization has been an increase in coastal population density. Almost 53% of the population of the United States lives in coastal counties that make up only 17% of its land area (Crossett et al. 2004). Coastlines, like cities, tend to be dominant forcing mechanisms on local weather and climate. Thus, the places where these coastal and urban atmospheric processes overlap are becoming an increasingly important area of research for mesoscale weather phenomena and dispersion and transport of atmospheric pollutants and contaminants (Banta et al. 2005).
The issues involved in forecasting urban–coastal interactions are complex, and perhaps nowhere are their impacts more evident than in Houston, Texas (Fig. 1). Houston is one of the most rapidly growing coastal cities in the United States and is also a major center of the international petrochemical industry. Houston’s position in relation to the Gulf of Mexico and the unique morphology of Galveston Bay cause complex coastal circulation patterns to evolve near the city. Coupled with the circulations generated by the urban heat island (UHI), this complex flow regime often results in the advection of atmospheric pollutants, which are largely byproducts of the petrochemical refinement process, into the more densely populated metropolitan area (Banta et al. 2005). Studies have addressed elements of this problem from the perspective of turbulence, land cover, radiation, and other factors (Bao et al. 2005; Zhong et al. 2007; Taha 2008; Cheng and Byun 2008; Shepherd et al. 2010). Further, urban–sea-breeze mesocirculations alter dispersion and transport patterns of chemical, biological, and radiological agents and are of interest because of their implications for homeland and national security concerns. Section 2 describes the background of the phenomena addressed in the study. Section 3 provides the context and motivation for the study. Section 4 presents data and methodology. Sections 5 and 6 present results and concluding statements, respectively.
Representation of the Houston study area comprising the fourth domain of the WRF simulations used herein. Each pixel represents one model grid cell. The land cover types represented are urban (gray), evergreen needleleaf (dark green), dryland–cropland pasture (yellow), deciduous broadleaf (white), crop–grass mosaic (light green), mixed dry-irrigated pasture (lavender), and water (blue).
Citation: Journal of Atmospheric and Oceanic Technology 29, 3; 10.1175/2011JTECHA1524.1
2. Background
a. The sea breeze
Glickman (2000) defines the sea breeze as “a coastal local wind that blows from sea to land, caused by the temperature difference when the sea surface is colder than the adjacent land.” Theoretical understanding of the sea breeze goes back as far as Haurwitz (1947). Figure 2 illustrates how the sea breeze is initiated and maintained.
Illustration of the basic model of sea-breeze formation (from Holton 2004). Variables are defined in the text. The parameters are discussed in the text.
Citation: Journal of Atmospheric and Oceanic Technology 29, 3; 10.1175/2011JTECHA1524.1
Limiting factors on the sea-breeze circulation include surface friction and weakening of the coastal temperature gradient due to thermal advection by the sea breeze itself. The strength of the sea breeze is directly related to latitude in that the sea breeze is strongest when the diurnal forcing mechanisms are synchronized with the Coriolis force (Banta et al. 2005). Houston, located just equatorward of 30°, has a particularly strong sea-breeze influence. The sea breeze is most prevalent on synoptically quiescent days, but the signal, a perturbed wind field, can still be present under stronger flow regimes even as the sea breeze is nonexistent. Further, the sea-breeze signal can be mitigated or washed out completely by the dominant synoptic flow.
b. The sea-breeze front and Houston
The sea-breeze front is the leading edge of cooler, moist marine air associated with the sea breeze. Under the proper conditions, the mesoscale boundary begins to develop in the morning as the land–sea temperature contrast sharpens and the sea-breeze circulation sets up. Depending on a number of factors, this boundary can progress inland less than 100 km or more than 300 km at the peak of its diurnal cycle (Simpson 1994). The sea-breeze front typically follows the general contours of the coastline, except when interrupted by an obstacle such as an urban area, and can generate patterns of convergence and divergence that affect the mesoscale environment. McPherson (1970) found a region of enhanced convergence in the convex regions around Galveston Bay. Complicating this simple model of convergence and divergence zones, however, is the city of Houston itself, which has its own effects on the local environment largely attributed to the UHI (Orville et al. 2001).
c. Urban heat island dynamics
Urban climatologies have shown that the urban rural temperature contrast can be as much as 10°C (Bornstein and Oke 1981; Oke 1981; Bornstein 1987; Ohashi and Kida 2002). Souch and Grimmond (2006) summarize much of the previous research on the characteristics of the urban heat island, stating that it reaches its diurnal peak during the nighttime hours, that it can be limited by increased wind speed and cloud cover, that its intensity varies by season with the minimum signal occurring in summer, and that it is strongly related to surface–building geometry, land use, vegetation, and patterns of anthropogenic heat release. The thermal and dynamic effects of the UHI have been linked to many weather phenomena, but of particular interest to this study are the enhanced convergence zones that exist around the periphery of UHIs (Hjelmfelt 1982; Bornstein and Lin 2000; Nielsen-Gammon 2000; Orville et al. 2001; Rozoff et al. 2003; Shem and Shepherd 2009; Miao et al. 2009; Shepherd et al. 2010).
d. Previous modeling work on the urban–coastal system
Efforts to model complex urban–coastal systems have been successfully completed (Yoshikado 1992, 1994; Kitada et al. 1998; Kusaka et al. 2000; Ohashi and Kida 2002; Lo et al. 2007; Shepherd et al. 2010; Dandou et al. 2009). The primary findings of these studies are that 1) urban–coastal circulations can be successfully simulated using a coupled atmosphere–urban canopy model (UCM) (Kusaka et al. 2000); 2) large cities such as Hong Kong contribute to an enhanced coastal temperature gradient; 3) large urbanized areas (>10 km) create thermal circulations in the mesoscale leading to enhanced vertical motions where the UHI and sea-breeze front converge; 4) the sea-breeze front penetration can be delayed by frictional retardation; and 5) if an inland urban area exists, the combined urban–coastal circulation is stronger and lasts longer than a typical sea breeze (Ohashi and Kida 2002). These studies primarily focused on cities outside of the United States. Also, while these studies have focused on sea-breeze–UHI interactions, very few have incorporated urban canopy parameter (UCP) data derived from direct measurements of the urban environment.
3. Motivation
a. Simulation of the coastal–urban mesoscale circulation and its evolution
A primary motivation is to add to the growing body of research on the evolution of complex coastal–urban circulations. Our primary interests are the four-dimensional evolution of the mesoscale boundaries and circulations under different urban land cover representations.
b. Diagnosing shallow coastal–urban mesoscale circulations
Complex coastal–urban circulations can be challenging to diagnose or simulate because they are relatively shallow systems. One difficulty is related to the depth. Nielsen-Gammon et al. (2008) found that the depth of the Houston sea breeze is typically less than or equal to 500 m. Local circulations like the sea breeze and urban wind perturbations are typically shallow and can be difficult to diagnose. Herein, we evaluate the use of the well-known bulk Richardson shear as a method to diagnose shallow mesocirculations.
c. Sensitivity of coastal–urban mesoscale circulations to UCPs based on degraded resolution lidar data
Accurate representation of the urban environment within atmospheric models is dependent on a series of UCPs. These parameters account for the aspects of the urbanized environment that have an effect on atmospheric circulation, such as the increased mechanical turbulence from tall buildings, the albedo of paved surfaces, and the urban canyon effect (Brown 2000). Therefore, in order to produce the most realistic simulation possible of urban-induced circulations, it is necessary to include as accurate a depiction as possible of the urban landscape (Jeyachandran et al. 2010). Holt and Pullen (2007) specifically cite “improving our understanding of the sensitivity of urban parameterizations to the specification of the urban morphology within the model” (p. 1926) as a future research need. By using the Weather Research and Forecasting (WRF) urban canopy model (Kusaka and Kimura 2004) along with the Noah land surface model (LSM; Ek et al. 2003), this study has been able to incorporate the degraded resolution UCP dataset into the Advanced Research WRF (ARW-WRF) simulations. Computing expense limits the resolution of our model simulations, so the primary goal was to determine whether 1-m UCP data that are degraded to approximately 1 km change the mesocirculation evolution compared to a two-dimensional urban feature.
d. Research objectives
The first objective of this study is to establish that a realistic representation of a sea-breeze event can be simulated using a coupled modeling system. An actual sea-breeze event is used as the basis for the study but the goal of these simulations was not to exactly reproduce the specific characteristics of this event in detail. Rather, the purpose of these simulations is to determine whether the modeling system can reproduce a theoretical sea-breeze response that can serve as a control for the UCP sensitivity analysis.
The second objective is to demonstrate the possible application of the bulk Richardson shear parameter as a diagnostic tool for local mesoscale circulations within a numerical modeling framework. This application of BRN is novel and could lead to other relevant applications in both observational and modeling studies.
The third objective is to evaluate, from a theoretical perspective, whether degradation of high-resolution, lidar-derived UCP data in the coupled WRF–Noah–urban modeling system alters the simulation of the coastal–urban circulation system in comparison to two-dimensional urban landscapes. This objective is motivated by an emerging availability of lidar products and more potential applications in models. It is not trivial to properly integrate such parameters, and we acknowledge that our approach is just one possible approach to consider. However, the lack of published literature on integrating lidar-derived parameters is a strong driver of this work. It is hypothesized that the inclusion of UCPs, even degraded to 1 km, in coupled WRF–Noah simulations will have noticeable effects on the representation of urban–coastal-induced circulations, convergence, and vertical motion.
As stated previously, in this study we simulate an actual sea-breeze case day but acknowledge that aspects of our simulation may not produce the exact timing and magnitude of the event in question. However, we note that our primary interests are related to theoretical considerations of dynamic–thermodynamic responses in the model realization. Although, we find that the model results exhibit a uniform cool bias relative to observed temperatures, the sea breeze and urban heat island evolution is physically realistic and useful for this theoretical comparative analysis.
4. Methods
a. Study area and case period
The study area is a region approximately 175 km × 175 km centered over the city of Houston, Texas, which also includes Galveston Bay and much of the nearby coastline (see Fig. 1). This comprises the fourth domain of the various WRF simulations used herein (Fig. 3). This domain is represented within the model as a 160 × 160 point grid with a 1.1-km spatial resolution and 27 vertical levels.
The four WRF model domains for the simulations used herein. The primary study area is the fourth domain.
Citation: Journal of Atmospheric and Oceanic Technology 29, 3; 10.1175/2011JTECHA1524.1
The study employs a theoretical case study approach. Case studies are commonly used to investigate physical mechanisms within a convective–mesoscale framework. While results must be carefully interpreted and not generalized, the literature is overwhelmingly clear concerning the contributions of properly designed case studies using models. This study is initialized using atmospheric conditions that were present on 17–18 August 2006, as reproduced in the North American Regional Reanalysis (NARR). This case was chosen based on two criteria, the first of which is the suitability of the synoptic and mesoscale weather conditions. Analysis of satellite data, surface weather observations, Doppler radar, and operational forecast model products shows that for these case days, there existed a well-developed sea-breeze circulation, as well as an absence of large-scale synoptic forcing mechanisms. Thus, any circulations that we are able to resolve should be due primarily to the mesoscale environment. The second criterion was that case study days fall within the study period of the Second Texas Air Quality Survey (TexAQS II) campaign. TexAQS II (Parrish et al. 2009) was an intensive effort to measure air quality and pollutant transport in southeast Texas using a dense network of surface and upper air data collection systems. While the primary focus of the TexAQS II campaign was air quality, and thus a large portion of the data collected was related to atmospheric pollutants such as ozone, all of the surface monitoring stations also collected basic meteorological data including temperature, dewpoint temperature, wind speed, and wind direction, which are used for model verification in this study. These data have been made available to us in digital format and can be accessed online at http://atmo.tamu.edu/texaqs2/ (J. Nielsen-Gammon 2009, personal communication).
b. Data
1) WRF–Noah
This study utilizes a series of coupled WRF–Noah simulations of the atmosphere. The WRF model was developed as a collaborative effort by the National Center for Atmospheric Research (NCAR), the National Centers for Environmental Prediction (NCEP), the Forecast Systems Laboratory (FSL), the Air Force Weather Agency (AFWA), Oklahoma University (OU), and the university community. The version of the model used is the ARW-WRF version 2.2. The ARW dynamic core is designed to integrate the compressible, nonhydrostatic Euler equations using a terrain-following mass vertical coordinate (Skamarock et al. 2005) and is structured to perform case studies based on real input data. The input data used in this case study are from the NARR dataset maintained by NCEP. For each case, the model is run for the duration of the case study period, totaling 36 h (0000 UTC on day 1 to 1200 UTC on day 2). The NARR input data have a spatial resolution of 32 km, which is then enhanced by using a series of nested domains within the model. Each domain increases the resolution by a factor of 3, such that the fourth domain, centered over Houston, has a resolution of 1.1 km.
To more accurately represent the effects of the urban environment and land surface interactions within the urban environment, the Noah LSM and WRF UCM have been coupled to the ARW-WRF for these simulations. The Noah LSM was developed by NCEP, Oregon State University, AFWA, and the National Weather Service Hydrologic Research Laboratory and is used to provide latent heat fluxes and surface skin temperature (Kusaka and Kimura 2004). The WRF UCM is used to represent the physical processes involved in the exchange of heat, momentum, and water vapor in urban environment by providing a representation within the model of shadowing from buildings, reflection of shortwave and longwave radiation, wind profiles in the canopy layer, and multilayer heat transfer equations for roof, wall, and road surfaces (Tewari et al. 2007).
Table 1 shows the physics and dynamics options used for the primary study area (fourth domain). These values are held constant throughout all the simulations used in this study, so that any differences between simulations should be due only to changes in the representation of urban land cover.
Physics and dynamics options used for the simulations described herein.
2) Urban canopy parameters
The UCPs used in this study originate from a unique dataset provided by coauthors Burian and Jeyachandran. This dataset is derived using airborne light detection and ranging (lidar) measurements of the urban environment and describes the height and density characteristics of the urban landscape in great detail and at high resolution (Fig. 4). Burian and Ching (2009) described the lidar data acquisition (from TerraPoint LLC), the processing of the UCPs, and the general trends of the UCPs in the Houston metropolitan area. The Houston UCP dataset used in this study was derived from a full-feature lidar digital elevation model with 1-m horizontal spatial resolution covering Harris County, Texas, an area of approximately 5800 km2, encompassing the Houston metropolitan area. The raw lidar data was subjected to a number of quality checks and processing steps using the ArcGIS geographic information system (GIS) software package as described in Burian and Ching (2009). The preprocessed lidar data were then analyzed to determine the three-dimensional characteristics of the buildings, trees, and other roughness elements, which are then used to compute the UCPs at 1-km spatial resolution fitting the modeling domain used in this study. For a complete description of the database, the interested reader should refer to Burian and Ching (2009).
(top) Representation of mean building height from the Burian dataset aggregated to model grid resolution. (bottom) Representation of mean building height of the three aggregated enhanced urban land use categories.
Citation: Journal of Atmospheric and Oceanic Technology 29, 3; 10.1175/2011JTECHA1524.1
The UCPs used for this study are as follows [following recommendations of Holt and Pullen (2007)]: building height (m), momentum roughness length above canyon (m), and sky view factor (implicit). A grid representing the model simulations’ finest resolution was overlaid over the dataset and a mean value for each UCP was calculated for each cell. Then, a “natural breaks” classification was used to separate these cells into three new land classes, analogous to “low-intensity residential,” “medium-intensity residential,” and “industrial or commercial” in the standard USGS classification scheme, which are the land use classes expected by the urban canopy model. The land classification was based on building height although all UCPs are included in the urban canopy model during the simulations. A mean value of each of the UCPs was then calculated for each of these new UCP-based land classes, and that value was used to replace the standard value within the urban canopy model’s parameter table. For those UCPs, which are not included in the dataset, the value from the analogous USGS land class was used, following Tewari et al. (2007). This new UCP-based urban land cover data was overlaid on the standard USGS 30-arc-second land cover to generate a new hybrid land cover dataset, which was used in the enhanced UCP simulations.
3) Verification data
To validate the model simulations, meteorological data were used that were collected as part of the TexAQS campaign, as well as surface data from the Automatic Surface Observation System (ASOS) network, the Automated Weather Observation System (AWOS), and the Texas Mesonet. Additional surface and satellite-based observations were used to identify a suitable case day. Figure 5 reveals that the model does capture the prevailing low-level east to northeast wind flow as indicated in the gridded observations at 2100 UTC. Generally, the simulations are realistically reproducing the diurnal pattern and spatial distribution of temperatures on the selected day, although the model results are typically 3°–5°C cooler than observed temperatures. Holt and Pullen (2007) also reported a cool bias in their simulations. Because this cool bias is consistent and uniform throughout the simulations, it does not appear to disrupt the representation of the diurnal cycle and allows for meaningful consideration of the theoretical impact of the degraded UCPs. It is also possible that the cool bias could also affect the mixing heights and thereby the depth of the sea breeze and urban circulations within the planetary boundary layer. Future work will need to consider a more careful validation as applications and forecasting implications are considered.
(top) Surface temperature (°C) and wind barbs from 2100 UTC 17 Aug 2006 illustrating conditions during the study period (data source is WxScope-gridded AWOS and ASOS). (bottom) Model output 2-m temperature (°C) and wind barbs from 2100 UTC 17 Aug 2006.
Citation: Journal of Atmospheric and Oceanic Technology 29, 3; 10.1175/2011JTECHA1524.1
c. Research methodology
1) Sea breeze–urban interactions and a novel application of bulk Richardson shear
2) Sensitivity to UCPs
To characterize changes in the sea breeze and sea-breeze front due to interaction with the urban area, the same SimpleUrban and NoUrban simulations are used. The sea-breeze front is identified based on the vertical velocity anomaly that occurs at this mesoscale boundary interface. Differences are identified between the SimpleUrban and NoUrban scenarios to determine the relative roles of urbanization and coastal morphology in sea-breeze evolution.
To characterize the sensitivity of these simulations to enhanced representation of the urban canopy, as well as identify possible implications regarding urban convective forcing and pollutant and contaminant transport, a third simulation was conducted, hereafter referred to as UCPUrban. This simulation is similar to the SimpleUrban scenario; however, the enhanced land classes and associated UCP values derived from the lidar dataset were incorporated into the land use dataset.
5. Results
a. Evolution of the urban–sea-breeze circulation
The results of the SimpleUrban simulation clearly show the initiation and evolution of the urban circulation over the city of Houston during the study period. Figure 6 shows that the SimpleUrban simulation was able to reproduce this initial forcing mechanism, with a persistent skin temperature anomaly (i.e., an urban heat island) over the urban area during the late afternoon and early evening hours. The exact magnitude of the anomaly varies from location to location and from hour to hour, but it approaches 10°C in some areas.
SimpleUrban skin temperature (K) from 2100 to 0200 UTC (1600–2100 CDT) showing the persistent skin temperature anomaly over the urban area.
Citation: Journal of Atmospheric and Oceanic Technology 29, 3; 10.1175/2011JTECHA1524.1
Figure 7 clearly indicates that the BRN shear is an indicator of the sea-breeze circulation. Stable, nonconvective, low-shear air masses exhibit higher values while environments with greater low-level shear are found along and rearward of the leading edge of the sea-breeze circulation. The lighter blue areas indicating low BRN shear values inland are likely caused by convective updrafts from daytime heating, such as often results in a field of cumulus humilis clouds. The green, yellow, and red areas denoting higher BRN shear values are those where the stable marine air mass has begun to intrude.
BRN shear field (m−2 s−2) at 2100–0200 UTC (1600–2100 CDT) showing the time evolution (2100–0200 UTC) of the low-level mesoscale circulations in the coastal–urban environment.
Citation: Journal of Atmospheric and Oceanic Technology 29, 3; 10.1175/2011JTECHA1524.1
Figure 8 shows the difference field of BRN shear when subtracting SimpleUrban from NoUrban simulation. An anomaly of lower BRN shear values can be seen both over and to the north of the city, confirming that the forcing may be urban induced. The positive BRN anomaly to the west of the city is due to the difference in sea-breeze front morphology caused by interaction with the urban area. We argue that small BRN shear indicates vertical mixing. Forthcoming discussions in section 5b (see also Fig. 10) further support this notion.
Difference field resulting from the subtraction of the NoUrban BRN shear from the SimpleUrban BRN shear (m−2 s−2) at 0000 UTC (1800 CDT). The negative BRN shear anomaly is caused primarily by the presence of the urban area in these simulations.
Citation: Journal of Atmospheric and Oceanic Technology 29, 3; 10.1175/2011JTECHA1524.1
b. Sea-breeze morphology
The SimpleUrban simulation was able to reproduce a realistic sea-breeze event, allowing for the characterization of the interactions between the sea breeze and urban environment. Figure 9 shows the evolution of the mesoscale boundaries, which in some cases may not develop until midafternoon. At 1900 UTC [1400 central daylight time (CDT)] the 2-m air temperature over the ocean is cooler than over land, on the order of 4°–5°C. By 2100 UTC (1600 CDT), the sea breeze has begun to set up, and the marine air mass begins to intrude inland, creating a sharp mesoscale temperature gradient. At 2300 UTC (1800 CDT) the sea breeze has begun to interact with the southern extent of the urban area, and by 0100 UTC (2000 CDT), the marine air mass has begun to bifurcate around the urban heat island. By 0300 UTC (2200 CDT) the sea breeze has completely split around the urban heat island, which has itself begun to advect northward. At 0500 UTC (0000 CDT), the sea breeze has completely overwhelmed the city, although an urban heat island signature remains.
The 2-m air temperature (°C) and wind direction from 1900 to 0500 UTC (1400–0000 CDT) showing the development of the sea breeze and intrusion of the marine air mass.
Citation: Journal of Atmospheric and Oceanic Technology 29, 3; 10.1175/2011JTECHA1524.1
Figure 10 is a differencing analysis in which the NoUrban near surface vertical velocity field is subtracted from the SimpleUrban vertical velocity field. The position of the sea-breeze front in the SimpleUrban scenario is indicated by darker colors in the vertical velocity field. The position of the boundary in the NoUrban scenario is indicated by lighter colors. One interesting finding is that the leading edge of positive vertical velocities in SimpleUrban are consistent with the leading edge of lower BRN shear (cf. the 0100 UTC panels in Figs. 7 and 10, respectively). This supports the notion that smaller BRN shear may be a proxy for vertical mixing.
Difference field resulting from the subtraction of the NoUrban near surface vertical velocity from the SimpleUrban vertical velocity (cm s−1) from 0100 to 0330 UTC (2000–2230 CDT) illustrating the changes in sea-breeze front morphology caused by the presence of the urban area in these simulations. The light colored boundary indicates the position of the NoUrban SBF while the dark colored boundary indicates the position of the SimpleUrban SBF.
Citation: Journal of Atmospheric and Oceanic Technology 29, 3; 10.1175/2011JTECHA1524.1
It is clear that, through its interaction with the urban-induced circulation, the sea-breeze boundary is both strengthened and accelerated in its inland progress on the western flank of the city, similar to the findings of Ohashi and Kida (2002) while at the same time being slowed somewhat over the urban core, which is consistent with Yoshikado (1992, 1994). Dandou et al. (2009) attributed this slowing to frictional retardation of the sea breeze front by the urban environment. This has the effect of reducing the dwell time of the sea-breeze front over the outlying urban areas, while simultaneously increasing the strength of the vertical velocity anomaly.
At 0100 UTC (2000 CDT), the NoUrban boundary lags behind the SimpleUrban boundary on the western edge of the city by as much as 10–15 km. There is some evidence of horizontal convective rolls in the figure (Fovell and Dailey 2001), but an analysis of their evolution is beyond our scope, although they are also known to interact with urban–sea-breeze circulations (Miao et al. 2009).
It is also possible that the more rapid propagation on the western flank is due to downstream advection of urban heat. The 2-m air temperature plots (Fig. 9) at 2100 and 2300 UTC, respectively, suggest that some westward advection of heat is occurring. This increased enhancement of the sea-breeze propagation on the western side of the city as opposed to the eastern side is an interesting result and warrants future study to determine whether this is caused by some mechanism common to all coastal urban areas or by some unique characteristics of this particular case such as the synoptic conditions of the day, the topography, or the particular characteristics of the Houston urban area. The SimpleUrban boundary continues to surge ahead of the NoUrban boundary until 0300 UTC (2200 CDT) when the sea-breeze front has begun to pass the urban area, and the acceleration of the boundary due to urban influences has slowed. Also at this time, a combined urban–coastal vertical velocity anomaly has developed on the northern edge of the city. By 0330 UTC (2230 CDT) the SimpleUrban and NoUrban boundaries have synchronized. Some of our results are similar to findings of Lo et al. (2007) in the Pearl River Delta of China but the western edge acceleration is not well understood. In the case of Houston the city does not cause the sea-breeze front to penetrate farther inland than it would if it were not present.
c. Sensitivity to urban canopy parameters
Another important goal of this study was to evaluate the sensitivity of ARW-WRF simulations to the inclusion of reduced resolution urban canopy parameters (see Fig. 11). Figure 12 shows the results of including this enhanced land cover on skin temperature. While early in the afternoon the differences are mainly confined to areas where the enhanced land cover is significantly different from the SimpleUrban land cover, by 2100 UTC (1600 CDT) the coarsened UCPs have led to a more complex surface thermal structure.
Land cover data used to initialize WRF in the UCPUrban simulation. Urban areas for which enhanced UCPs were available are shown in shades of red; all other urban areas are shown in gray. Other land classes are as in Fig. 2.
Citation: Journal of Atmospheric and Oceanic Technology 29, 3; 10.1175/2011JTECHA1524.1
Difference field resulting from the subtraction of the SimpleUrban skin temperature from the UCPUrban skin temperature (K) showing the complex differences which arise in skin temperature from the inclusion of enhanced urban canopy parameters: (top) 1900 and (bottom) 2100 UTC.
Citation: Journal of Atmospheric and Oceanic Technology 29, 3; 10.1175/2011JTECHA1524.1
Analysis of the BRN shear difference between the UCPUrban and SimpleUrban simulations shows that the primary impact of coarsened UCPs on the evolution of convergence occurs early in its development, around 2100 UTC (1600 CDT) (Fig. 13). In general, this fits well with our understanding as the surface is where the UCPs would have the greatest impact. Finally, to address the issue of transport, a series of parcel trajectories were plotted (Fig. 14). These trajectories simulate the path of a series of parcels released along a southwest to northeast transect through the urban core and surrounding areas at 2100 UTC (1600 CDT) and allowed to advect for 3 h. The “chimney effect” of enhanced urban forcing is clearly visible, as parcels released from within the urban boundary reach a higher altitude (larger arrow head) than those released outside the city. The effect of the urban area on the atmosphere to the west of the city is evident as well, as those parcels follow a similar trajectory as those released within the city but do not reach as great an altitude because of the lack of urban forcing. Conversely, parcels released to the northeast of the city where the urban impact is minimized remain primarily near the surface and move with the environmental wind.
Difference field resulting from the subtraction of the SimpleUrban BRN shear field from the UCPUrban BRN shear field (m−2 s−2) at 2100 UTC (1600 CDT), illustrating the complex differences in the convective field caused by the inclusion of enhanced urban canopy parameters.
Citation: Journal of Atmospheric and Oceanic Technology 29, 3; 10.1175/2011JTECHA1524.1
Plot showing the trajectories of parcels released along a southwest to northeast transect through the urban area at 2100 UTC (1600 CDT). The endpoints of each trajectory show the position of the parcel 0000 UTC (1900 CDT). Arrowhead size indicates altitude. Base map is 2-m air temperature (°C) and wind vectors at 0000 UTC.
Citation: Journal of Atmospheric and Oceanic Technology 29, 3; 10.1175/2011JTECHA1524.1
6. Conclusions
a. Urban convergent forcing
Through a series of model simulations, this study has been able to characterize the structure and evolution of the sea breeze–urban circulation near Houston, Texas. This study also introduced a novel use of the BRN shear term, normally used to diagnose thunderstorm environments, to identify shallow mesocirculations and possible areas of vertical mixing and to chart the evolution of the urban convergent forcing. Areas of enhanced convergence were associated with the leading edge of the sea breeze as well the urban environment.
b. Urban–sea breeze interactions
The sea breeze is a long studied aspect of the coastal environment, but its interactions with urban centers are recently receiving more attention. The implications of the sea breeze as a transport mechanism for pollutants are clear. The results of these simulations show that while coastal morphology can itself lead to complex sea-breeze front structures, including preferred areas of vertical motion, the urban environment also has a large impact on the evolution of the sea-breeze mesoscale boundary. The sea-breeze front begins to strengthen and accelerate, particularly on the western edge of Houston, as it comes into contact with the urban area. This acceleration decreases the dwell time of the sea-breeze front over the urban area, but also increases the strength of the vertical velocity anomaly.
c. Sensitivity to urban canopy parameters
This study shows that even reduced resolution urban canopy parameters can alter the simulation of complex circulations in the region. Because of a pronounced cool bias in the simulations and the lack of a robust verification analysis, we cannot conclusively state the significance of the changes. It appears, at least qualitatively, that even degraded to the model resolution (1 km) of our experiments, processes are simulated that would appear to be more realistic than a typical two-dimensional urban slab. The inclusion of building heights and associated parameters into the model’s land surface representation led to differences in patterns of skin surface temperature, which has implications for all aspects of urban weather. Further, this change in skin temperature patterns leads to a higher degree of complexity in the evolution of the urban convective forcing, especially during its incipient phase, which is the most important stage of its development with regard to pollutant transport. The effects of urban canopy parameters are also not limited to the city itself. This analysis shows that perturbations generated by the enhanced urban environment begin to radiate outward from the city, preferentially to the west. This is particularly important for the evolution sea-breeze front, which is most strongly affected by urban forcing on the western flank of the city.
d. Future directions
This analysis established some degree of motivation for exploring ways to integrate three-dimensional urban canopy parameters into coupled atmosphere–land surface models. Future work should seek to find an optimal convergence point between model resolution and the resolution of the lidar-derived parameters.
Additional work in developing these datasets for use with other urban areas, as well as expanding the number of urban canopy parameters available within the datasets, will be necessary to explore this avenue of research to its full potential. Further field campaigns, such as TexAQS, will be an excellent resource for collecting model verification data with the goal of improving model performance.
Future work investigating pollutant transport under different flow regimes could also make an important contribution for societal impacts and urban planning. Another possible use of this basic framework is the investigation of urban-induced precipitation and lightning anomalies (Shepherd 2005; Shepherd et al. 2010). Previous work has shown that urban areas are capable of affecting precipitation patterns over a large area, and better understanding of the urban-induced convergence and vertical velocity structures resolved by this study could lead to greater understanding of the forcing mechanisms that lead to these anomalies.
Coastal urban environments generate complex atmospheric forcing mechanisms that alter dispersion and transport patterns of chemical, biological, and radiological agents and are of interest not only because of their implications for air quality and for accurate weather forecasting but also because of potential homeland and national security concerns. It is hoped that this study will stand as a useful contribution toward greater understanding of the atmospheric circulations which result from urban–coastal interactions, and that this framework will lead to further improvements in our understanding of urban weather.
Acknowledgments
The authors would like to acknowledge support from the Defense Threat Reduction Agency under Grant HDTRA1-07-C-0065.
REFERENCES
Banta, R. M., and Coauthors, 2005: A bad air day in Houston. Bull. Amer. Meteor. Soc., 86, 657–669.
Bao, J.-W., Michelson S. A. , McKeen S. A. , and Grell G. A. , 2005: Meteorological evaluation of a weather–chemistry forecasting model using observations from the TEXAS AQS 2000 field experiment. J. Geophys. Res., 110, D21105, doi:10.1029/2004JD005024.
Bornstein, R. D., 1987: Mean diurnal circulation and thermodynamic evolution of urban boundary layers. Modeling the Urban Boundary Layer, Amer. Meteor. Soc., 53–93.
Bornstein, R. D., and Oke T. , 1981: Influence of pollution on urban climatology. Adv. Environ. Sci. Eng., 3, 171–202.
Bornstein, R. D., and Lin Q. , 2000: Urban heat islands and summertime convective thunderstorms in Atlanta: Three case studies. Atmos. Environ., 34, 507–516.
Brown, M. J., 2000: Urban parameterizations for mesoscale meteorological models. Mesoscale Atmospheric Dispersion, Z. Boybeyi, Ed., WIT Press, 193–255.
Burian, S. J., and Ching J. , 2009: Development of gridded fields of urban canopy parameters for advanced urban meteorological and air quality models. EPA Rep. EPA/600/R-10/007, 73 pp. [Available online at http://ams.confex.com/ams/pdfpapers/79765.pdf.]
Cheng, F. Y., and Byun D. W. , 2008: Application of high resolution land use and land cover data for atmospheric modeling in the Houston–Galveston metropolitan area. Part I: Meteorological simulation results. Atmos. Environ., 42, 7795–7811.
Crossett, K. M., Culliton T. J. , Wiley P. C. , and Goodspeed T. R. , 2004: Population trends along the coastal United States: 1980–2008. U.S. Department of Commerce Coastal Trends Reports Series, 54 pp. [Available online at http://oceanservice.noaa.gov/programs/mb/pdfs/coastal_pop_trends_complete.pdf.]
Dandou, A., Tombrou M. , and Soulakellis N. , 2009: The influence of the city of Athens on the evolution of the sea-breeze front. Bound.-Layer Meteor., 131, 35–51.
Ek, M. B., Mitchell K. E. , Lin Y. , Rogers E. , Grummann P. , Koren V. , Gayno G. , and Tarpley J. D. , 2003: Implementation of Noah land surface model advances in the National Centers for Environmental Prediction operational mesoscale Eta model. J. Geophys. Res., 108, 8851, doi:10.1029/2002JD003296.
Fovell, R. G., and Dailey P. S. , 2001: Numerical simulation of the interaction between the sea-breeze front and horizontal convective rolls. Part II: Alongshore ambient flow. Mon. Wea. Rev., 129, 2057–2072.
Glickman, T. S., Ed., 2000: Glossary of Meteorology. 2nd ed. Amer. Meteor. Soc., 855 pp.
Haurwitz, B., 1947: Comments on the sea-breeze circulation. J. Meteor., 4, 1–8.
Hjelmfelt, M. R., 1982: Numerical simulation of the effects of St. Louis on mesoscale boundary-layer airflow and vertical motion: Simulations of urban vs non-urban effects. J. Appl. Meteor., 21, 1239–1257.
Holt, T., and Pullen J. , 2007: Urban canopy modeling of the New York City metropolitan area: A comparison and validation of single- and multilayer parameterizations. Mon. Wea. Rev., 135, 1906–1930.
Holton, J. R., 2004: An Introduction to Dynamic Meteorology. 4th ed. Academic Press, 511 pp.
Jeyachandran, I., Stetson S. , and Burian S. J. , 2010: Estimating urban canopy parameters using synthetic aperture radar data. J. Appl. Meteor. Climatol., 49, 732–747.
Kitada, T., Okamura K. , and Tanaka S. , 1998: Effects of topography and urbanization on local winds and thermal environment in the Nohbi Plain, coastal region of central Japan: A numerical analysis by mesoscale meteorological model with a k–ε turbulence model. J. Appl. Meteor., 37, 1026–1046.
Kusaka, H., and Kimura F. , 2004: Thermal effects of urban canyon structure on the nocturnal heat island: Numerical experiment using a mesoscale model coupled with an urban canopy model. J. Appl. Meteor., 43, 1899–1910.
Kusaka, H., Kimura F. , Hirakuchi H. , and Mizutori M. , 2000: The effects of land-use alteration on the sea-breeze and daytime heat island in the Tokyo metropolitan area. J. Meteor. Soc. Japan, 78, 405–420.
Lo, J. C. F., Lau A. K. H. , Chen F. , Fung J. C. H. , and Leung K. K. M. , 2007: Urban modification in a mesoscale model and the effects on the local circulation in the Pearl River Delta region. J. Appl. Meteor. Climatol., 46, 457–476.
McPherson, R. D., 1970: A numerical study of the effect of a coastal irregularity on the sea breeze. J. Appl. Meteor., 9, 767–777.
Miao, S., Chen F. , LeMone M. , Tewari M. , Li Q. , and Wang Y. , 2009: An observational and modeling study of characteristics of urban heat island and boundary layer structures in Beijing. J. Appl. Meteor. Climatol., 48, 484–501.
Nielsen-Gammon, J. W., 2000: The Houston heat pump: Modulation of a land–sea-breeze by an urban heat island. Proc. 2000 PSU/NCAR Mesoscale Modeling System Users’ Workshop, Boulder, CO, NCAR. [Available online at http://box.mmm.ucar.edu/mm5/workshop/ws00/N-G-heat.pdf.]
Nielsen-Gammon, J. W., and Coauthors, 2008: Multi-sensor estimation of mixing heights over a coastal city. J. Appl. Meteor. Climatol., 47, 27–43.
Ohashi, Y., and Kida H. , 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.
Oke, T. R., 1981: Canyon geometry and nocturnal urban heat island: Comparison of scale model and field observations. J. Climatol., 1, 237–254.
Orville, R., and Coauthors, 2001: Enhancement of cloud-to-ground lightning over Houston Texas. Geophys. Res. Lett., 28, 2597–2600.
Parrish, D. D., and Coauthors, 2009: Overview of the Second Texas Air Quality Study (TexAQS II) and the Gulf of Mexico Atmospheric Composition and Climate Study (GoMACCS). J. Geophys. Res., 114, D00F13, doi:10.1029/2009JD011842.
Rozoff, C. M., Cotton W. R. , and Adegoke J. O. , 2003: Simulation of St. Louis, Missouri, land use impacts on thunderstorms. J. Appl. Meteor., 42, 716–738.
Shem, W., and Shepherd J. M. , 2009: On the impact of urbanization on summertime thunderstorms in Atlanta: Two numerical model case studies. Atmos. Res., 92, 172–189.
Shepherd, J. M., 2005: A review of current investigations of urban-induced rainfall and recommendations for the future. Earth Interact.,9. [Available online at http://EarthInteractions.org.]
Shepherd, J. M., Carter W. M. , Manyin M. , Messen D. , and Burian S. , 2010: The impact of urbanization on current and future coastal convection: A case study for Houston. Environ. Plann., 37B, 284–304.
Simpson, J. E., 1994: Sea-Breeze and Local Wind. Cambridge University Press, 234 pp.
Skamarock, W. C., Klemp J. B. , Dudhia J. , Gill D. O. , Barker D. M. , Wang W. , and Powers J. G. , 2005: A description of the Advanced Research WRF version 2. NCAR Tech. Note NCAR/TN-468+STR, 88 pp.
Souch, C., and Grimmond S. , 2006: Applied climatology: Urban climate. Prog. Phys. Geogr., 30, 270–279.
Taha, H., 2008: Episodic performance and sensitivity of the urbanized MM5 (uMM5) to perturbations in surface properties in Houston Texas. Bound.-Layer Meteor., 127, 193–218, doi:10.1007/s10546-007-9258-6.
Tewari, M., Chen F. , Kusaka H. , and Mio S. , 2007: Coupled WRF/Unified Noah/Urban-Canopy Modeling System. NCAR Research Applications Laboratory, 22 pp. [Available online at http://www.rap.ucar.edu/research/land/technology/urban/WRF-LSM-Urban.pdf.]
United Nations Population Fund, 2007: 2007 annual report. UNFPA Information and External Relations Division, 30 pp. [Available online at http://www.unfpa.org/upload/lib_pub_file/777_filename_unfpa_ar_2007_web.pdf.]
Yoshikado, H., 1992: Numerical study of the daytime urban effect and its interaction with the sea-breeze. J. Appl. Meteor., 31, 1146–1164.
Yoshikado, H., 1994: Interaction of the sea breeze and urban heat islands of different sizes and locations. J. Meteor. Soc. Japan, 72, 139–142.
Zhong, S., In H., and Clements C. , 2007: Impact of turbulence, land surface, and radiation parameterizations on simulated boundary layer properties in a coastal environment, J. Geophys. Res., 112, D13110, doi:10.1029/2006JD008274.
