Why Trees Are Important: Heat Islands
SCRAMBLED OR SUNNY SIDE UP: LIVING WITHIN THE HEAT ISLAND
Heat Island research SCRAMBLED OR SUNNY SIDE UP: LIVING WITHIN THE HEAT ISLAND http://geography.hunter.cuny.edu/~nycesrt/heat_island.html Joy Romanski Anyone who has spent a summer in New York City has undoubtedly heard the phrase, "It's so hot you could fry an egg on the sidewalk!" This is not mere hyperbole -- with sidewalk temperatures sometimes surpassing 120°F on a hot summer day (Landsberg, 1981), it really is possible to cook the proverbial egg. Alternative methods of food preparation aside, urban environments create significant perturbations in the local climatic regime, altering temperature, wind and precipitation patterns in their immediate area and downwind. The primary urban meteorological influence is the urban heat island (e.g., Leahey and Friend, 1971; Landsberg, 1973; Landsberg, 1981; Hildebrand and Ackerman, 1984; Policy Statement of the American Meteorological Society, 1992). A difference in surface-level air temperatures between a city and its rural surrounds was first noted in 1820 by Luke Howard, an English chemist and amateur meteorologist (Landsberg, 1981). He performed measurements of urban vs. rural temperatures in and around London over a ten year period from 1807 to 1816, finding a pronounced nocturnal urban temperature elevation of 3.7°F. Since that time, numerous other studies have demonstrated urban-induced temperature anomalies (Leahey and Friend, 1971, citing Renou, 1862; Hammon and Duenchal, 1902; Schmidt, 1929; Middleton and Millar, 1936; Sundborg, 1950; Chandler, 1967). The study of urban heat islands now encompasses more than just a qualitative description of urban-rural temperature differences for various cities. Urban climatologists probe the underlying causes of thermal anomalies, for instance, how roofing and street materials influence the urban energy budget (Rosenfeld, et al., 1997) and how radiation exchange occurs within urban canyons (Johnson and Watson, 1984). Others investigate how heat islands affect and are affected by mesoscale weather conditions (e.g., Grillo and Spar, 1971; Landsberg, 1981, citing Angell, et al., 1966, Colacino and Dell'Osso, 1978, Bornstein and Fontana, 1979, Hildebrand and Ackerman, 1984). Studies of several cities have demonstrated an average ground-level temperature increase of 2-3.5°C above nearby rural areas with urban readings sometimes elevated as much as 17°C in the case of New York City (Bornstein, 1968; Landsberg, 1973; Landsberg, 1981 Vukovich, 1983; Brazel, et al., 1993; Changnon, 1992). Several characteristics of the urban environment contribute to the formation of heat islands; these include the low-albedo, heat-retentive and high heat conductivity properties of building materials such as concrete and asphalt, the rapid runoff of water caused by paved surfaces, which leaves little available for evaporative cooling, the lack of trees to cool the air through evapotranspiration, heat released from anthropogenic sources such as building furnaces and vehicle exhausts, reduced wind speeds within the urban canopy (the subrooftop level) and high concentrations of pollutants such as SOx, NOx and O3, which trap longwave radiation. The intensity of a city's heat island is directly related to its size (Landsberg, 1981; Changnon, 1992). Regressions determined by Oke (Oke, 1976) and Landsberg (Landsberg, 1975) show a linear relationship between population and heat island intensity for numerous North American and European cities. Studies conducted by Landsberg of the planned community of Columbia, Maryland demonstrate the evolution of an urban heat island as a function of population. When the community was founded in 1968 it had 100 residents and a maximum heat island of 1°C. In 1974, with a population of 20,000, the maximum heat island had grown to 7°C (Landsberg, 1975). This value is similar to that observed by Hutcheson, et al. in their study of Corvallis, Oregon (Landsberg, 1981, citing Hutcheson, et al., 1967), a town of 21,000 with a maximum heat island of 6.7°C. In terms of urban heat islands, not all air masses are equal. Large scale stagnation is most favorable for the development of urban heat islands. The greatest differences between rural and urban temperatures are associated with synoptic scale high pressure systems, characterized by clear skies, low wind speeds and stable, subsiding air. (Landsberg, 1981; Hildebrand and Ackerman, 1984). In the absence of high winds or large scale air movement, microscale climatic effects such as urban heat islands reach their greatest intensity. Conditions are especially favorable for the development of heat islands on calm, clear nights. Under these conditions, rural areas cool rapidly by radiating heat to the atmosphere while the city is kept warm by the heat absorbed during the day by its streets and buildings. The taller the buildings and the narrower the streets, the less longwave radiation escapes (Johnson and Watson, 1983). The resulting diurnal variation in intensity of the urban heat island has been documented in a number of studies (e.g., Bornstein, 1968; Landsberg, 1981; Leahey and Friend, 1971 citing Renou, 1862; Hammon and Duenchal, 1902; Schmidt, 1929; Middleton and Millar, 1936; Sundborg, 1950; Chandler, 1967). Description of the horizontal aspect of the urban heat island is only part of the picture. The three-dimensional character of the urban heat island cannot be inferred from surface-level data -- the vertical temperature structure must be determined as well. This was first undertaken in 1947 by Balchin and Pye (Leahey and Friend, 1971) in Bath, England. Bath is located in a valley, permitting researchers to take measurements from the surrounding hills. Technological advances later enabled studies of vertical temperature profiles to be performed using remote sensing techniques such as data taken from instrumented television towers (De Marrais, 1961), tetroons (tetrahedral volumetric balloons which float at a constant density level) (Angell, et al., 1968), helicopters (Bornstein, 1968) and aircraft (Hildebrand and Ackerman, 1984). Thermal infrared data from radiometers aboard various satellites, including the Improved TIROS Operational Satellite, the NOAA-3 and NOAA-5 Satellites and the Heat Capacity Mapping Mission (HCMM) Satellite, has been used to detect and characterize heat islands in various regions of the United States (Vukovich, 1983). In a study published in 1968, Bornstein described the three-dimensional structure of New York City's heat island, using data obtained over 42 predetermined test mornings from July 1964 through December 1966 (Bornstein, 1968). He found a low incidence of surface temperature inversions compared to surrounding rural areas. Characteristically, temperature decreases with increased elevation at a rate referred to as the lapse rate, which varies depending on the particular air mass. Under certain conditions, however, temperatures increase with increased elevation - this is called a temperature inversion and can occur either at the surface or at various levels within the troposphere. The point at which temperatures begin to rise with increasing altitude is considered to be the bottom of the inversion layer and the point at which temperatures begin to decrease with increasing altitude marks the top of the inversion layer. The air within an inversion layer is stable and thus inversions form a barrier to vertical motion in the atmosphere. The type of clear, windless conditions favorable for the development of heat islands also favor the development of stable surface inversions. Generally, surface inversions form at night when the heat from the ground is lost and are destroyed after sunrise by solar heating, which warms the atmosphere from the ground up. In urban areas, however, the nocturnal surface inversion never develops because streets and buildings radiate their stored heat to the surrounding atmosphere, creating an urban zone of discontinuity within the rural inversion. Instead, urban areas are more often characterized by weak elevated nocturnal inversions, thought to result from the absorption and reradiation of heat within elevated haze layers comprised of smoke, water vapor, CO2 and SO2 (Bornstein, 1968). Shortly after sunrise, convection associated with insolation destroys the haze layer. The layer below the inversion is characterized by instability and turbulent mixing and is consequently referred to as the mixing layer. The vertical extent of the heat island is considered to be the area below the elevated inversion layer. The urban influence extends downwind of the city, with warm, unstable urban air overriding the cooler rural air (the "urban plume") (Leahey and Friend, 1971; Oke, 1978). Bornstein found an average depth at the center of the heat island of 300 m, with the temperature elevations greatest at the surface and decreasing to zero at the base of the inversion layer (Bornstein, 1968). In order to understand the urban heat island effect and its impact on local and regional atmospheric conditions, it is necessary to consider not only the spatial and temporal three-dimensional distribution of temperature fields, but also the variation of parameters such as heat flux, moisture, and wind (Landsberg, 1981; Hildebrand and Ackerman, 1984). Direct measurements are taken of these atmospheric conditions (e.g., Hildebrand and Ackerman, 1984) which are used in conjunction with mathematical models (Leahey and Friend, 1971) to predict the evolution over time of various aspects of the urban heat island. Heat flux has been shown to be greater in urban areas than in rural areas (Landsberg, 1981; Hildebrand and Ackerman, 1984). The reason for this is that urban building materials such as concrete, brick and asphalt have a lower albedo and a higher heat capacity than the vegetation found in rural regions. They absorb and store almost three times as much solar radiation (4.53 Wm-2 in a city versus 1.67 Wm-2 in the country) than does vegetation and radiate it back into the city at night (Landsberg, 1981). Investigation into the urban versus rural humidity field reveals a 4% drop in relative humidity over urban areas (Landsberg, 1981, citing Landsberg and Maisel, 1972). Landsberg and Maisel attribute half of the reduction to higher temperatures within the heat island and half to reduced evapotranspiration from impermeable urban surfaces. Measurements of moisture flux by Hildebrand and Ackerman show a marked alteration in urban areas. While in rural regions the moisture flux profile is characterized by a decrease with height, in urban areas the reverse is true. This variation in moisture flux reflects the greater convection and entrainment present in the urban atmosphere. (Entrainment is a process by which heated air within an air column rises and cools, causing its load of water vapor to condense. The condensed vapor then raises the humidity of the surrounding air. Since there is an updraft, more air flows upward, repeating the process. The result is an increase in humidity at higher levels of the atmosphere. This process is largely responsible for the formation of convective cloud types, such as cumulus.) This has the net effect of transporting large quantities of energy, in the form of latent heat, into the air above cities. The elevated urban temperatures also affect wind circulation patterns, inducing the development of a torroidal or roll vortex circulation structure. In such a structure, warm air converges at the surface and rises in the center of the city. The air column then diverges aloft, flowing outward over the city, to eventually descend and flow in toward the surface convergence at the urban center. This circulation scheme is markedly different from that found in rural areas (Hildebrand and Ackerman, 1984; Oke, 1978) and has a pronounced impact on local precipitation patterns and air pollutant dispersion. Synoptic conditions favoring the formation of heat islands are those associated with the most severe air pollution episodes -- a stable, anticyclonic air mass characterized by subsiding air, clear skies and weak winds. Within the heat island, conditions are even worse. The elevated inversion layer typifying urban heat islands traps and concentrates pollutants within the mixing layer. When pollutants are emitted in an area where a surface inversion prevents them from reaching the ground the polluted plume disperses horizontally but not vertically (a condition known in dispersion meteorology as fanning, so called because the plume appears fan-shaped when viewed from above). When such a fanning plume travels downwind and enters the unstable urban sub-inversion mixing layer, it joins with urban pollutants to create an extremely hazardous situation in which pollutants are carried to the surface by turbulence within the mixing layer, a condition known as fumigation. The alarming implications for New York City's air quality become clear upon considering the location of many chemical and petroleum plants immediately upwind in New Jersey. The large scale torroidal urban circulation typically accompanying heat islands then recirculates the polluted air, resulting in ever greater pollutant concentrations (Oke, 1978). As well as trapping pollutants in the city, urban heat islands actually contribute to the formation of additional pollutants. Higher temperatures increase the rate of reaction of the chemical processes that result in the formation of NOx, SOx and O3 (Mestel, 1995; Rosenfeld, et al., 1997). For example, according to Rosenfeld, et al., the Los Angeles heat island raises tropospheric O3 levels by 10 to 15 percent. Higher pollutant concentrations then serve to intensify the heat island by absorbing and scattering more longwave radiation, which catalyzes the formation of even more pollutants, with the only limit being the amount of pollutant precursors available. This type of nonlinear feedback effect can rapidly concentrate urban pollutants. Under such adverse conditions, health risks to susceptible individuals and to the general population increase dramatically. Sometimes the adverse health impacts of urban heat islands are more direct. In July, 1966 a heat wave encompassed the central and eastern United States, lasting for several weeks. During this time, New York and St. Louis recorded 50 percent more deaths than would normally be expected for that time period. Analysis shows that many of the fatalities were due to strokes, with persons between 45 and 64 years of age and persons over 80 years of age being most at risk. During this period, New York City recorded its highest temperature ever, 107°F, at La Guardia Airport. One can only image how much hotter it was in the densely built inner city area, where the most deaths occurred (Landsberg, 1981). A linear relationship has been demonstrated between heat stroke probabilities and daily maximum temperatures in New Orleans. Nearly 80% of heat strokes occur when the maximum temperature exceeds 86°F, a fairly low maximum for this humid subtropical climate, where temperatures above 90°F are common (Landsberg, 1981). A 1995 Midwestern heat wave brought record daily maximum temperatures to numerous locations across the Great Plains, but particularly in Chicago, where daily maxima often exceeded 100°F, with maximum apparent temperatures (an index based on actual air temperature and relative humidity) exceeding 118°F. This record heat was accompanied by a high number of fatalities; in fact, 65% of all deaths in the nation during this period occurred in Chicago (Kunkel, et al., 1996). These examples show that during periods of hot weather, urban heat islands are capable of raising temperatures well above the danger point for many residents. The influence of human activities on urban precipitation has been clearly demonstrated in a number of studies, including those of Ashworth, who showed that the industrial town of Rochdale experienced 13% greater rainfall during the week (Landsberg, 1981, citing Ashworth, 1929) and Dettweiler, whose 8 year study of Paris demonstrated a 24% increase in weekday precipitation as compared to Saturdays and Sundays (Landsberg, 1981, citing Dettweiler, 1970). This clear delineation between weekdays and weekends strongly indicates an anthropogenic influence. Human activities within cities influence cloud processes in two ways; by increasing convection and by loading the atmosphere with aerosols. Increased convection caused by heat islands leads to a higher incidence of cloudiness and precipitation in urban areas and downwind (Landsberg, 1981; Hildebrand and Ackerman, 1984; Changnon, 1992; Policy of the American Meteorological Society, 1992). Heating warms and destabilizes air in the city, causing it to become less dense than the surrounding air. The air will then rise until it attains neutral buoyancy with respect to the surrounding atmosphere. As it rises, the weight of the air column above decreases, lowering the pressure on the parcel of rising air. The air parcel then expands and cools at a constant rate (the dry adiabatic lapse rate, dry meaning that the air parcel is not saturated and adiabatic meaning that there is no net gain or loss of enthalpy). As it cools, the capacity of air to hold water vapor decreases and thus, since the air parcel contains the same amount of water vapor, the relative humidity of the air parcel increases. Eventually, if it rises and cools enough, the air parcel will reach saturation (which is a function of temperature), at which point condensation is possible. Condensation requires not only saturation of the air parcel but also the presence of hygroscopic particles called cloud condensation nuclei. Without these, an air parcel can even be supersaturated and yet condensation will not occur. Since urban aerosols such as H2SO4 and HNO3 have excellent hygroscopic properties there is usually no shortage of condensation nuclei in an urban area. The combination of these pollutants is particularly effective in enhancing rain, as the sulfates tend to form small droplets (<0.1µm) while the nitrates form larger droplets (>1 µm). Smaller droplets stabilize clouds while larger droplets incorporate smaller ones and eventually grow large enough to fall as rain, a process known as condensation and coalescence which is thought to be the primary mechanism by which warm-cloud precipitation occurs (Landsberg, 1981). The urban environment thus provides optimum conditions for cloud formation and precipitation, providing both convection and condensation nuclei. Urban influences most strongly affect warm-season clouds, causing a 10-20% increase in precipitation within the urban area and adjacent downwind regions (Landsberg, 1981; Policy Statement of the American Meteorological Society, 1992). Urban enhancement of convective precipitation is most pronounced when the atmosphere is highly unstable, producing heavy rainfall events (Changnon, 1992). Short-term, intense convective showers of this type are the ones which most endanger heavily paved urban areas, overwhelming storm sewers and catchment basins, resulting in more frequent urban flooding (Landsberg, 1981). Increased temperatures place greater demands on energy supply. People use air conditioners more frequently and keep them on for longer periods of time. Brazel, et al. plotted the spatial variation of cooling degree hours (a measure of energy demand based on cumulative degrees of temperatures above 75°F) over the Phoenix metropolitan area and found a strong correlation between increased cooling degree hours and urban land-use patterns (Brazel, et al., 1993). Research conducted at Lawrence Berkeley Laboratories shows that one-sixth of the electricity used in the United States goes toward cooling buildings, for a cost of billion per year. For every additional degree of heat within a heat island, Americans use $1 million more of electricity per hour to cool off (Rosenfeld, et al., 1997). In addition to costing money, higher energy demands increase the consumption of fossil fuels and the concomitant deleterious effects on the environment, such as the release of greenhouse gases. Urban heat islands do bring a few benefits to city residents by lowering winter heating costs and by reducing the amount of snow that falls in the city as compared to adjacent rural areas (Grillo and Spar, 1971). By and large, however, they make city life unpleasant and even dangerous. Computer simulations generated by Lawrence Berkeley Laboratories and UCLA have demonstrated that a few simple measures, if adopted on a widespread basis in urban areas, could significantly reduce heat islands. For example, simulations of the effect on the Los Angeles heat island indicate a possible reduction of 5°F (Rosenfeld, et al., 1997). This simulated reduction was achieved easily and inexpensively by planting trees and using lighter colors for roofs and pavement. Trees reduce urban heat loads in several ways; they shade streets and buildings, they filter out heat-trapping pollutants such as SOX and NOX, they remove CO2 from the air through photosynthesis and they cool the air through evapotranspiration, which transfers sensible heat from the atmosphere to latent heat in water vapor (Landsberg, 1981; Moll, 1989; Mestel, 1995; Rosenfeld, et al., 1997). Not all trees are suitable for use in urban areas, however, as research conducted by Arthur Winer of UCLA shows (Mestel, 1995). Winer found that certain tree species actually emit large quantities of volatile hydrocarbons, which then react with NOx to form smog. Another factor to consider in selecting tree species suitable for mass urban plantings is tolerance to urban conditions. University of Washington researchers Clark and Kjelgern have classified urban arboreal environments into three types: park, canyon and plaza(Moll, 1989). Park environments most closely approximate natural settings, with a mixture of various plant species, full access to sunlight and no restrictions on root growth. Trees within this optimal environment are the healthiest and have the highest rate of growth. Canyon environments are those where soil access is limited and the area is partially shaded by buildings. Streetside plantings are usually of the canyon type. Trees within this habitat have actually adapted to the reduced sunlight available to them; their leaves are thinner and broader and they have more branches than their counterparts of the same species in plaza and park environments. The well-adapted canyon trees have a rate of growth comparable to that of park trees (Moll, 1989). Plaza environments are found in open areas with restricted soil availability, such as parking lots. Within a plaza environment, pavement reduces available water while additional sunlight is reflected from buildings and streets. These factors create stress on plaza trees and have resulted in reduced growth, higher incidence of disease and even tree death. Trees selected for the highly stressful plaza habitat must be hardy, drought-resistant species. Urban heat island intensity can also be reduced through the use of high-albedo building materials. For instance, roads can be made lighter in color by using a lighter asphalt aggregate or by using cement as a binder instead of asphalt. Although initially costlier, cement is cheaper in the long-term because it is stronger and more durable than asphalt and so would require less frequent repair and replacement (Rosenfeld, et al, 1997). Using roofing materials designed to reflect radiation would also cool urban areas. Lighter colored shingles aren't necessarily better, in fact some white shingles actually absorb more radiation in the near infrared spectrum than do red terra cotta shingles. The use of shingles made with titanium dioxide, which reflects in both the visible and the infrared spectra, can raise the albedo of a city over 35%. The Department of Energy and the EPA are working with Lawrence Berkeley Laboratories to produce an index of reflectivity for labeling roofing materials, raising consumer awareness of thermal pollution and enabling them to make an informed choice (Rosenfeld, et al., 1997). "Cool communities" techniques don't have to be adopted all at once; as existing roofing and paving materials wear out, they can be replaced with new, cooler surfaces. Computer simulations indicate that it would take about 15 years to obtain a reduction of 5°F in the Los Angeles heat island. This assumes an increase in city albedo of 7.5% and a tree cover of 5%, both modest assumptions (Rosenfeld, et al., 1997). Reducing vehicle use would also cool cities, both by eliminating heat directly generated by cars, trucks and buses, and by diminishing the amount of heat-trapping pollutants in the urban atmosphere. Furthermore, space presently devoted to parking could be transformed into cooling green spaces. Nowhere have humans altered the natural environment more than in cities. The changes we have made have had unforeseen consequences, one of which is climate modification. Analysis of urban-induced climate change is instructive in its own right and gives insight into the mechanisms of climate modification on a global scale (e.g., anthropogenic alterations in energy balance and cloud processes) (Changnon, 1992). The understanding gained from such investigation can be used to improve life for city residents and to serve as a springboard to address the larger issue of global climate change which affects all of us. References Angell, J.K., Pack, D.H., Hass, W.A. and Hoecker, W.H. (1968) "Tetroon Flights Over New York City," Weather: vol. 23; pp. 184-191. Bornstein, R.D. (1968) "Observations of the Urban Heat Island Effect in New York City," Journal of Applied Meteorology: vol. 7; pp. 575-582. Brazel, A.J., Verville, H.J. and Lougeay, R. (1993) "Spatial-Temporal Controls on Cooling Degree Hours: An Energy Demand Parameter," Theoretical and Applied Climatology: vol. 156, pp. 1-12. Changnon, S.A. (1992) "Inadvertent Weather Modification in Urban Areas: Lessons for Global Climate Change," Bulletin of the American Meteorological Society: vol. 73; pp. 619-627. Grillo, J.N, and Spar, J. (1971) "Rain-Snow Mesoclimatology of the New York Metropolitan Area," Journal of Applied Meteorology: vol. 10, pp. 56-61. Hildebrand, P.H. and Ackerman, B. (1984) "Urban Effects on the Convective Boundary Layer," Journal of the Atmospheric Sciences: vol. 41, pp. 76-91. Johnson, G.T. and Watson, I.D. (1984) "The Determination of View-Factors in Urban Canyons," Journal of Climate and Applied Meteorology: vol. 23, pp. 329-335. Kunkel, K.E., Changnon, S.A., Reinke, B.C. and Arritt, R.W. (1996) "The July 1995 Heat Wave in the Midwest: A Climatic Perspective and Critical Weather Factors," Bulletin of the American Meteorological Society: vol. 77, pp. 1507-1518. Landsberg, H.E. (1973) "The Meteorologically Utopian City," Bulletin of the American Meteorological Society: vol. 54, pp. 86-89. Landsberg, H.E. (1981) The Urban Climate. New York: Academic Press. Leahey, D.M. and Friend, J.P. (1971) "A Model for Predicting the Depth of the Mixing Layer Over An Urban Heat Island with Applications to New York City," Journal of Applied Meteorology: vol. 10, pp. 1162-1173. Mestel, R. (1995) "White Paint On A Hot Tin Roof," New Scientist: vol. 25, pp. 34-37. Moll, G. (1989) "Cooling Our Baking Cities," American Forests: vol. 95, pp. 21-23, 76-77. Oke, T. (1978) Boundary Layer Climates. London: Methuen; New York: Wiley. Policy Statement of the American Meteorological Society (1992) "Planned and Inadvertent Weather Modification," vol 73, pp. 331-337. Rosenfeld, A.H., Romm, J.J., Akbari, H. and Lloyd, A.C. (1997) "Painting the Town White - And Green," Technology Review: Feb/March, pp. 52-60. Vukovich, F.M. (1983) "An Analysis of the Ground Temperatures and Reflectivity Pattern about St. Louis, Missour, Using HCMM Satellite Data," Journal of Climate and Applied Meteorology: vol. 22, pp. 560-571.