SEARCH

SEARCH BY CITATION

Keywords:

  • anthropogenic heat;
  • energy use;
  • waste heat

Abstract

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Sources of anthropogenic heat and moisture
  5. 3. Methods for estimating anthropogenic heat and moisture emissions
  6. 4. Inter-comparison of methods
  7. 5. Conclusions
  8. Acknowledgements
  9. References

Energy consumption in the urban environment impacts the urban surface energy budget and leads to the emission of anthropogenic sensible heat and moisture into the atmosphere. Anthropogenic heat and moisture emissions vary significantly both in time and space, and are not readily measured. As a result, detailed models of these emissions are not commonly available for most cities. Furthermore, most attempts to quantify anthropogenic emissions have focused on the sensible heat component, largely ignoring moisture emissions and invoking assumptions—such as the equivalence of energy consumption and anthropogenic sensible heating—which limit the accuracy of the resulting anthropogenic heating estimates. This paper provides a historical perspective of the development of models of energy consumption in the urban environment and the associated anthropogenic impacts on the urban energy balance. It highlights some fundamental limitations of past approaches and suggests a roadmap forward for including anthropogenic heat and moisture in modelling of the urban environment. Copyright © 2010 Royal Meteorological Society


1. Introduction

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Sources of anthropogenic heat and moisture
  5. 3. Methods for estimating anthropogenic heat and moisture emissions
  6. 4. Inter-comparison of methods
  7. 5. Conclusions
  8. Acknowledgements
  9. References

In the early decades of computer-based regional-scale atmospheric modelling, urban surfaces were poorly represented in the models. As the resolved scale of these models improved and interest in urban applications of the models grew, researchers sought better representations of the urban surface. Initially these representations took the form of modifications in surface characteristics such as roughness length, albedo, thermal conductivity, and thermal diffusivity. In some cases, they involved attempts to represent the complex structure of buildings in the urban environment as a heat storage layer (Tso et al., 1990). Many studies have been conducted to try to understand the urban thermal climate or the potential for heat island mitigation using this sort of simplified modelling framework (e.g. Myrup, 1969; Richiardone and Brusasca, 1989; Anthes, 1990; Sailor, 1995; Taha, 1996). These approaches typically represent the urban area as a flat surface (soil) that simply has thermal properties that correspond to a weighted average of the construction materials in the city. In more recent efforts, researchers have incorporated more sophisticated parameterisation schemes that seek to represent the complicated impacts of urban morphology on energy and momentum exchange (e.g. Masson, 2000; Lemonsu and Masson, 2002; Martilli et al., 2002; Otte et al., 2004; Dupont and Mestayer, 2006).

One aspect of the urban environment that has been largely ignored or overly simplified in many studies of the urban climate is heat and moisture emissions associated with energy consumption in cities. These emissions result from sources including human metabolism, vehicles, commercial and residential buildings, industry, and power plants. While prior studies have used the term anthropogenic heat (Qf) synonymously with energy consumption (e.g.Oke, 1988), the present paper differentiates energy consumption from the associated emission of both sensible heat and moisture. Sensible anthropogenic heat emission into the atmosphere can be directly through tailpipes, chimneys, and air-conditioning or heating equipment, or indirectly by conduction through a building envelope and then convection and radiation into the urban environment. The sensible heat emissions can lag the timing of the energy consumption and can differ substantially in magnitude. Anthropogenic moisture emissions also take two forms. Heat removed from buildings can be exhausted through evaporative cooling equipment. Such heat removal is often referred to as anthropogenic latent heating, with the net effect on the urban atmosphere being a source of moisture. The second mechanism of anthropogenic moisture emissions is the chemical reaction that occurs in the combustion of hydrocarbon fuels either in vehicle engines or combustion furnaces. This process creates anthropogenic water vapour as a result of a chemical reaction rather than phase change and is thus distinct from latent heat emissions. In this paper, the term anthropogenic heat is used to refer to sensible emission of heat associated with energy consumption. The latent and chemical generation of anthropogenic moisture resulting from energy consumption in the urban environment is referred to collectively as anthropogenic moisture emissions.

The focus of this paper is on methods for estimating anthropogenic sensible heat and moisture emissions for use in models of the urban atmosphere. It starts by describing the characteristics of the various energy consumption sources that lead to these emissions. It then describes the three distinct categories of approaches that have been used to develop estimates of the magnitude, spatial distribution, and diurnal profiles of anthropogenic emissions. Examples are given where researchers have developed sophisticated estimation techniques and applied them to specific cities of interest. The paper concludes with a discussion of next steps needed to enable widespread and accurate representation of anthropogenic heat and moisture emissions in cities around the world.

2. Sources of anthropogenic heat and moisture

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Sources of anthropogenic heat and moisture
  5. 3. Methods for estimating anthropogenic heat and moisture emissions
  6. 4. Inter-comparison of methods
  7. 5. Conclusions
  8. Acknowledgements
  9. References

As anthropogenic heat is closely tied to (but not identical to) energy consumption, this inventory of sources focuses primarily on the three classically defined sectors of the economy that consume energy—transportation, buildings, and industry. A fourth, less important, source that is also discussed is anthropogenic heat release directly from human metabolism. It is important to note that total anthropogenic heat release results from the sum of these sources and is accompanied by an emission of moisture and also an impact on the urban long-wave radiation budget. Anthropogenic emissions models that treat only one sector, or ignore moisture emissions, may significantly misrepresent both the timing and magnitude of anthropogenic emissions.

2.1. Metabolism

To estimate heat release from human metabolism, we first note that metabolic rates vary over the course of a day. Specifically, referring to data from Fanger (1972) and Guyton (1986), the sleeping metabolic rate for a typical 70-kg man is about 75 W. During the daytime, this metabolic rate increases depending upon the activity. A typical person has a metabolic rate that may range from about 100 W at rest to 200 W while walking and more than 300 W during strenuous activity. The magnitude of anthropogenic heating from human metabolism in a city clearly depends upon population density. Daytime population densities in US cities are typically on the order of 5000 persons per square kilometre and the corresponding metabolic heat flux is on the order of only about 1 W/m2. While many cities around the world have higher population densities in their urban cores, at any instant in time much of the urban population is located within buildings. Thus, while representing total anthropogenic heating in an urban area, it is important to recognise that the building sector anthropogenic heating model (discussed later) likely accounts for occupancy loads, and human metabolism should not be treated separately as it will lead to a double counting of that source of heating. In fact, most studies either integrate metabolism as a component of their building energy model (e.g. Hsieh et al., 2007; Heiple and Sailor, 2008) or simply ignore it (Grimmond, 1992; Sailor and Lu, 2004). On the basis of the reasonable assumption that more than half of metabolic heat occurs in buildings, it can be shown that outdoor human metabolism is generally less than 1% of the total anthropogenic heating for a city, and can thus be ignored.

2.2. Industry

It is well known that energy use in the industrial sector is relatively insensitive to variations in weather and has a much more uniform diurnal and seasonal distribution than in other sectors. In fact, once the spatial distribution and intensity of industry within a city is determined, it is fairly common to simply assume that the energy consumption is uniformly distributed among the 8760 h of the year (e.g. Sailor and Fan, 2004; Sailor and Lu, 2004; Torrance and Shum, 1975). Depending upon the region or city under study, the monthly or annual energy consumption within the industrial sector can be obtained from power utility and/or governmental energy agency data bases (e.g. EIA, 2003a, 2003b, 2004). Land use data can then be used to apportion the industrial sector energy consumption into the corresponding regions within the city where industry is prevalent. While much of the energy consumption in the industrial sector is converted directly into sensible heat, there are instances where heat is removed using evaporative cooling towers or by exchanging heat with a large body of water such as a river. As a result, it is difficult to estimate how much of the energy consumption in the industrial sector results in sensible anthropogenic heat and the magnitude of resulting moisture emissions. Large industrial consumers are generally reticent when it comes to sharing details of their energy consumption patterns. Therefore, reasonable estimates must be made on the basis of publicly available data regarding equipment characteristics and usage within the predominant industries within the city under study.

2.3. Buildings

Energy use in the building sector can be thought of as falling into three primary categories: lighting loads; plug & appliance loads; and energy use for heating, ventilation and air conditioning (HVAC) equipment. Energy use for lighting is fairly substantial, often comprising 20 to 30% of the total building electricity use (EIA, 1998, 1999). Energy use for plug loads and lighting are both directly tied to building occupancy schedule and hence depend upon the day type (work day vs non-work day) and usage category of the building. Residential buildings typically have peak energy consumption early in the morning hours and again in the mid-evening hours. Depending upon the climate, the afternoon peak may be dominant in the summer and the morning peak may be dominant in the winter. In the commercial sector, loads begin to ramp up on workday mornings levelling out during midday and ramping down again in the late afternoon. Weekend load patterns are distinctly different, particularly in the commercial sector where building loads are relatively minimal during non-workdays. Energy use in buildings for heating, ventilation, and air conditioning is a complicated function of occupancy, internal loads, and environmental loads (Figure 1). During occupied hours, the building's HVAC controls manage internal thermal comfort conditions through use of heating and cooling systems. As heat is transmitted into the building from the exterior environment (E) and internal heat is generated by lighting (L), plug loads (P), and human metabolism (M), the cooling system consumes additional energy (AC) to reject this heat to the outdoor environment. This heat rejection, R, can thus be represented as follows:

  • equation image(1)

Depending upon the type of mechanical equipment used to condition the building, the rejected heat may comprise both a sensible and latent component. Specifically, cooling towers may reject up to 80% of their cooling load through evaporation as moisture.

thumbnail image

Figure 1. Waste heat rejected from a building (a) is the sum of the environmental loads (b), the internal building loads (c), and the building air-conditioning energy consumption (d), less any heat that is stored or exhausted through other means such as in waste water (e)

Download figure to PowerPoint

Environmental loads are loads associated with environmental heating or cooling of the building interior. For example, in winter, internal heat is lost to the outside through conduction and both intentional air exchange and unintentional leakage. The role of such air exchanges is particularly important. For example, residential buildings typically have 0.3 to 0.5 air exchanges per hour. That is, within 1 h, up to half of the volume of air is replaced by outside air. In summer, the important environmental loads include conduction, infiltration, and solar radiation penetrating through windows. While the solar loading is also present in winter, it serves to reduce the need for heating demand during cold months. In summer, however, the shortwave transmissivity of windows (typically 60 to 80%) admits a significant environmental load into buildings that must be removed by the air-conditioning system. As a result, the heat removed from a building in summer by the air-conditioning system is significantly larger than the thermal load associated with energy consumption within the building. For example, whole building energy simulations for office buildings in Houston, Texas reveal that the total heat rejected from the buildings in summer can be 40 to 70% greater than the actual energy consumption of the buildings (Sailor and Brooks, 2009). Another complicating factor is that some air-conditioning systems use evaporative cooling to exchange heat with the outside environment. In such systems, the majority of the heat may be removed as evaporated water.

As a result of these complexities, it is particularly problematic to represent anthropogenic heat emission from the building sector with an energy use inventory approach. Such an approach that simply equates energy consumption with sensible heat emissions can either over- or under-estimate anthropogenic heat depending upon the season and types of buildings under consideration. Furthermore, inventory approaches generally neglect the energy-related emissions of moisture, which can be quite large in cities where a significant fraction of the air-conditioning load is met using evaporative systems such as residential ‘swamp coolers’ or cooling towers, which are common in large commercial buildings.

Another challenging aspect of assessing anthropogenic heat and moisture emissions from the building sector is the estimation of the actual vertical location of the emissions. Some of the anthropogenic emissions occur as conduction through the building envelope. A larger fraction occurs as a result of air exchanges through the façade and through natural operation of windows and doors. The largest fraction of anthropogenic emissions from buildings comes in the form of heat and moisture rejected from mechanical heating, cooling, and ventilation systems. It is thus important to know where these systems are located. In buildings that are fewer than 10 stories tall, the HVAC equipment may be located adjacent to the building on the ground level, on a mezzanine-level roof, or on the rooftop of the top floor. In mid- and high-rise commercial buildings, however, the emissions may be more uniformly distributed vertically on utility floors (perhaps every 10 or 12 floors) or consolidated at the rooftop level. As a result, it is difficult to specify the height of anthropogenic emissions from buildings unless specific building characteristics are known. As a general rule, this author suggests that it is reasonable to simply assume that such emissions are uniformly distributed over the height of the building.

2.4. Vehicles

As vehicles travel along roadways, they release heat and moisture associated with the combustion of gasoline or diesel fuel. A typical heating value for vehicle fuels is 45 MJ/kg (Annamalai and Ishwar, 2006). Also, as fuel is burned, the chemical reaction leads to generation of water vapour. In fact, for each litre of gasoline or diesel fuel burned 0.9 to 1.0 kg of water vapour is generated. This corresponds to roughly 100 g of water vapour for every kilometre driven. As with the building sector, this emission of water vapour, in itself, may be a significant aspect of the urban climate system. The challenging aspect of determining anthropogenic emissions from the vehicle sector involves estimating the spatial and temporal distribution of vehicles on major and minor roadways within the city. Further, the actual fuel economy of vehicles varies among vehicle types, and the distribution of vehicle types on roads changes over the course of the day. Fleet fuel economy depends very much on the city and country of analysis, but typically is in the range of 8 to 16 km/l (20 to 40 miles per gallon). Many regional transportation authorities track vehicle use in one form or another. Generally one can find statistics on average vehicle distance travelled per day by urban residents in major cities. In the United States, these data are compiled by the US Department of Transportation (USDoT, 2003) and range from about 30 to 60 km/person/day. In other countries throughout the world, these numbers are often substantially lower. As detailed hourly traffic count data are not available for all road types in most cities, some level of disaggregation and assignment of diurnal profiles is needed. As an example, Hallenbeck et al. (1997) created useful diurnal profile data illustrating the typical morning and evening peaks in rush hour traffic in US cities. They found that about 16% of all daily traffic occurs between 1600 and 1800 local time, and another 13% occurs between 0700 and 0900 in the morning rush hours. The analysis of Sailor and Lu (2004) suggests that in US cities, such as Houston TX, the total anthropogenic heating from the vehicle sector is as high as 300 W/m2 during afternoon rush hour averaged over 500-m square grids centred over major freeways.

2.5. Relative importance of each sector

As a simplification, we use energy consumption as the basis for estimating the relative importance of each sector of the economy in producing anthropogenic heat. It is crucially important to recognise that there is large variation from country to country, and even from city to city within a country. Nevertheless, for illustrative purposes, consider the annual energy consumption statistics for the United States. According to the Energy Information Administration (e.g. EIA, 2005), about 40% of all end-use energy consumption in the United States occurs in the building sector, about 30% is in transportation, and about 30% is in industry/manufacturing. Data from the International Energy Association (www.iea.org) suggests that all three sectors are generally important in most OECD (Organisation for Economic Co-operation and Development) countries. In contrast to the United States, however, in most European countries the building sector accounts for a larger fraction of the country's total energy consumption, and the transportation sector is relatively smaller. In some countries, such as Turkey and the Czech Republic, the transportation sector accounts for as little as 15 to 20% of the total end-use energy consumption. France and Denmark are examples of countries where the building sector is clearly dominant, representing nearly 50% of total end-use energy consumption. Therefore, as a starting point, it is clear that any accounting of anthropogenic heat must include each sector, but that the relative importance of each sector will vary.

As noted above, each sector is not uniformly distributed among all cities within any country. For example, Houston, Texas, is a city with a large petrochemical and manufacturing industry, located within the city. In fact, a number of these large facilities have on-site electric generation power plants. In contrast, many other cities are dominated by the other sectors with relatively small industry and manufacturing components. Therefore, making any blanket statements regarding the relative size of the various components of anthropogenic heating is dangerous. Cities with a large industrial base such as Houston may have on the order of half their energy consumption in the industrial sector with the remainder somewhat evenly split between buildings and vehicles. In cities with a relatively small industrial base, the building and vehicle sectors may comprise 80% or more of the total energy consumption.

As a general rule, the diurnal profile of total anthropogenic heating has local peaks in the morning and mid-afternoon, corresponding to peaks in transportation and building energy use. As illustrated in Figure 2, the diurnal profile also depends upon the day of the week, with total emissions on weekends and holidays being significantly lower than emissions on workdays. The relative magnitude of the morning and evening peaks depends upon the underlying climate and the season. In summer months, the afternoon peak becomes more pronounced as a result of air-conditioning loads that tend to peak between 1500 and 1700 local time. In the winter, the morning peak may become more pronounced as a result of heating demand in the building sector.

thumbnail image

Figure 2. Typical shapes of diurnal profiles of anthropogenic heating (Qf) for (a) work days, and (b) non-work days, illustrating local peaks in the morning and early evening hours

Download figure to PowerPoint

3. Methods for estimating anthropogenic heat and moisture emissions

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Sources of anthropogenic heat and moisture
  5. 3. Methods for estimating anthropogenic heat and moisture emissions
  6. 4. Inter-comparison of methods
  7. 5. Conclusions
  8. Acknowledgements
  9. References

As suggested in the previous sections, there are different approaches that have been used to estimate anthropogenic emissions in the urban environment. These include inventory approaches, micrometeorologically based energy budget closure methods, and building energy modelling approaches. As discussed below, each approach has its advantages and limitations.

3.1. Inventory approaches

An inventory approach is one in which energy consumption data are gathered, typically at coarse resolution in time and space. These approaches may rely on utility scale energy consumption data, or data resulting from energy consumption surveys. The vast majority of inventory applications make the assumption that consumption is equal to anthropogenic sensible heat emissions, with no time lag, no accounting for environmental loads, and no partitioning into sensible and latent components. Further, they generally must use some mechanism to map annual or monthly consumption data onto diurnal profiles. They must also map city-wide or regional emissions estimates onto a finer scale grid to represent spatial variability within the city. Taha (1997) gathered data from a number of early inventory-based approaches and found that the annual city-wide estimates of anthropogenic heating typically ranged from 15 to150 W/m2. One of the earliest of these inventory studies was that of Torrance and Shum (1975) who related annual energy consumption data to population density to arrive at annual estimates of 83.7 W/m2 for a densely populated city. In this work, Torrance and Shum represented space heating through a simplified empirical model that scaled space heating energy use with heating degree days. Their industrial component of anthropogenic heat was simply assumed to be constant in time. They represented the transportation sector by using hourly traffic count data averaged over the city. Their model represented the spatial variation of anthropogenic heating only in terms of differentiating ‘urban’ from ‘rural’ components. Despite its limitations, this study was perhaps the first to estimate anthropogenic heat emissions (up to 135 W/m2 for a summer day in an unspecified dense city) and the resulting impact on urban air temperatures. Specifically, through use of a transient one-dimensional atmospheric model, Torrance and Shum estimated that anthropogenic heating could result in up to a 4 °C warming in summer. Subsequent studies improved on both the spatial and temporal disaggregation of inventoried energy consumption data. Kimura and Takahashi (1991), for example, used detailed land use data in combination with energy consumption data to estimate anthropogenic heating and its impact on urban air temperatures in Tokyo. While they used a fairly simple approach to estimate the diurnal profile (early morning minimum equal to half the midday value), they were able to take advantage of detailed (2 km by 2 km) gridded land use data to create a somewhat realistic spatial map of anthropogenic heat emissions. Peak anthropogenic heating from their analysis was on the order of 100 W/m2. By feeding these sensible heat emissions into a hydrostatic three-dimensional numerical model, they then estimated that most of the nocturnal summertime heat island in Tokyo (2–3 °C) is due to anthropogenic heating. This was a particularly significant finding, as it clearly illustrated the important role that anthropogenic heating plays in the formation of the urban heat island.

Similar inventory approaches have been applied in a range of urban regions around the world (e.g. in Korea by Lee et al. (2009) and in London by Harrison et al. (1984)). One such example is the work of Klysik, (1996), who used an inventory approach to estimate anthropogenic heat emissions in Lodz, Poland. In his work, the annual heating energy consumption data were mapped to monthly profiles using variability in monthly air temperatures. Population density was used as a metric for discerning spatial variation in electricity use. Anthropogenic heat emissions from the vehicle sector were estimated on the basis of annual fuel sales. While limited to coarse resolution in space and time, Klysik's analysis suggests monthly averaged anthropogenic heat flux in Lodz ranging from 12 W/m2 in summer to 54 W/m2 in winter.

Ichinose et al. (1999) took a major step forward in spatially resolving anthropogenic heating. They linked energy statistics for various building types and end uses in Tokyo [obtained from Ichinose et al. (1994)] with detailed digital geographic land use data (12 land use categories) to estimate hourly anthropogenic heating at a 25 m by 25 m grid resolution. As expected, the high spatial resolution led to some rather large estimates of heating associated with grids that were centred on large high-rise office buildings. Specifically, they found that hourly energy consumption at this scale (assumed to be equal to anthropogenic sensible heating) could be as large as 1590 W/m2 in the winter and was more than 400 W/m2 during summer daytime hours. Their analysis also integrated these estimates of anthropogenic heat into a numerical simulation (using the Colorado State University Mesoscale Model). They found that anthropogenic heating resulted in summer air temperature elevations of less than 1 °C during the day, but as much as 1.5 °C at night. In winter, when the magnitude of anthropogenic heating was much higher (and mixing heights presumably lower), they found that the nocturnal warming reached 2 to 3 °C.

Fan and Sailor (2005) implemented the inventory-based approach of Sailor and Lu (2004) in a study of anthropogenic heating in Philadelphia, Pennsylvania. That study found peak anthropogenic heating at the city scale to be 60 W/m2 in summer and approximately 90 W/m2 in winter. The impacts of this anthropogenic heating on urban air temperatures ranged from 1 °C in summer to 3 °C in winter. One of the unique characteristics of the method presented in Sailor and Lu (2004) is that it emphasised use of data that are widely available for most large US cities. The modelling approach was inventory based and diurnally distributed based on widely applicable profile functions. As with most inventory-based approaches, the work of Sailor and Lu implemented the assumptions that energy consumption equals anthropogenic heat and that all of this heating is sensible. Another limitation of this work was that it provided no mechanism for mapping city-wide anthropogenic heating data onto finer spatial scales. Nevertheless, its wide applicability led to its application in Sailor and Hart (2006) to 50 US cities.

In a study of London, Hamilton et al. (2009) used a similar approach to estimate inventories of building energy consumption across the city, and then mapped this consumption onto 1-km grids based on detailed land use and building characteristic data. They found spatially and temporally averaged anthropogenic heating in London to be 9 W/m2, but noted that the spatial and diurnal variations were significant.

Recently, there has been an increased interest in implementing anthropogenic heating in large-scale general circulation models. Thus far, these efforts have also used an inventory-based approach with necessarily crude representations of spatial and temporal variability. While observing anthropogenic heating impacts on regional climates in central Europe, for example, Block et al. (2004) used regional estimates of energy consumption and population density as a proxy for anthropogenic heat emissions. Their work simply assumed this heating to be constant in time, and used such a large estimation area that the magnitude of anthropogenic heating was 14.5 W/m2 or less throughout their domain. Flanner (2009) also used an inventory approach to estimate anthropogenic heating for use in global climate models. His work mapped country-wide energy consumption data into 2.5 by 2.5 min latitude/longitude grids using population density. On the basis of the work of Sailor and Lu (2004), Flanner used weighting functions to create seasonal and diurnal profiles. Owing to the relatively coarse scale of the analysis, the resulting magnitude of anthropogenic heat was relatively low.

These inventory approaches all suffer from two main limitations: first, they can only resolve energy consumption at relatively coarse spatial and temporal scales and must use some technique to create estimates at finer scales and, second, they make the assumption that energy consumption is equivalent to anthropogenic sensible heat emissions.

3.2. Estimates based on energy budget closure

In contrast to inventory approaches that are generally based on relatively coarse data, direct measurement of the urban energy budget enables estimation of anthropogenic heating over targeted neighbourhoods within a city. In theory, if one draws a control volume around a portion of a city as in Figure 3, it is possible to track all energy movement into and out of this control volume. In this figure, Q* is the net radiation, QE and QH are latent and sensible heat transported across the top of the control volume, ΔQA is the total heat advected out of the sides of the control volume, ΔQs is the storage of thermal energy within the control volume, and QF is the total source of anthropogenic heat within the control volume. From the micrometeorological standpoint, one can use net radiometers to track net radiation (Q*), and eddy covariance techniques to estimate sensible and latent heat fluxes (QH and QE). If the net advection out of the control volume is assumed to be negligible and the total energy stored in the control volume is either estimated through surface measurements or assumed to be stationary from one day to the next, the energy balance can be simplified and solved for QF:

  • equation image(2)

This equation is applicable for small time steps (say 1 h), or if one integrates the energy balance over the course of a 24-h period, the net storage term may be assumed small, and ignored.

thumbnail image

Figure 3. A control volume analysis of the energy budget of a portion of a city in close proximity to a micrometeorological tower (a). As illustrated, the nominal control volume (b), radiation measurement footprint (c), and eddy flux footprint (d) may not coincide

Download figure to PowerPoint

One particularly attractive yet limiting aspect of this method is that it inherently includes anthropogenic heating from all sources. This sort of approach has been applied by a relatively small number of investigators with differing levels of success. Offerle et al. (2005) applied this concept to data gathered in a downtown region of Lodz, Poland. These data included surface temperature and heat flux measurements, measurements of net radiation, and an eddy covariance approach to estimate sensible and latent heat flux. The challenge with using an energy balance residual approach is that any errors within the measurement system are simply incorporated into the estimate of the residual. Specifically, due to significant challenges in using eddy flux methods in a non-homogeneous urban system, the magnitude of the errors in estimating sensible and latent fluxes may be large. Also, as illustrated in Figure 3, the source area for radiation measurements can be thought of as being a circular area with the centre at the instrument tower. The flux measurements, however, essentially represent a source footprint area dictated by upwind flow and dispersion characteristics. As a result, the nominal source area for the radiation measurements may not coincide with or be similar to the source area for the eddy flux measurements. The assumptions regarding thermal storage and advection add uncertainty to estimates of anthropogenic heat flux. Therefore, it is not surprising that the results of Offerle et al. include unrealistically negative numbers for anthropogenic heating in summer months. Their estimates for wintertime anthropogenic heating, however, were more consistent with what is expected from inventory methods—magnitudes around 50 W/m2.

In a contrasting approach, Kato and Yamaguchi (2005) employed an energy budget residual approach in which remote-sensed satellite data were used in conjunction with surface meteorological data to estimate anthropogenic heating in Nagoya Japan. Owing to the nature of their measurements (surface radiation), Kato and Yamaguchi's results do not allow estimation of anthropogenic heat directly, but rather, the increase in surface sensible heat flux resulting from anthropogenic heating and the associated warming of the near-surface air.

Pigeon et al. (2007) present one of the most recent and comprehensive energy-budget-based estimates of anthropogenic heat release. Their research focused on Toulouse, France as part of the CAPITOUL experiment. This study included both an inventory-based approach with some similarities to Sailor and Lu (2004) as well as an energy budget residual approach similar to Offerle et al. (2005). The particularly useful aspect of their study was that it enabled direct comparison of the two approaches. They found the two approaches to be in general agreement with wintertime anthropogenic heating as high as 70 W/m2 and summertime values closer to 15 W/m2.

Overall, the energy budget micrometeorological approach to estimating sensible anthropogenic heating has value as a local validation for other estimation approaches. The key limitations of this approach include a high level of uncertainty associated with some of the assumptions required to deduce anthropogenic heating as a residual in an energy budget. Further, these approaches cannot differentiate among the different source terms and often cannot reliably produce diurnal profiles of heat release.

3.3. Estimates based on building energy models

While the inventory and energy-balance-based approaches are readily capable of estimating anthropogenic heating from all sources in the urban environment, anthropogenic heating from the building sector is particularly complicated and has been the subject of a number of focused studies. These studies generally involve explicit modelling of energy consumption within buildings and careful evaluation of heat rejection. The town energy budget (TEB) model of Masson (2000) was an early attempt to model the anthropogenic heat from buildings into the urban environment. While TEB represented a significant advancement with respect to modelling turbulent processes and energy budgets in urban canyons, its representation of buildings was rather simplified. It treated buildings as envelope structures with conduction through the envelope dictating the need for heating or air conditioning, which is then emitted into the environment. This formulation is insufficient as it ignores internal loads in buildings, which can lead to increased cooling demand and reduced heating demand. It also ignores shortwave heating of internal spaces through glazing (windows), which typically accounts for 20 to 40% of the wall surface area of buildings and air exchange between the building and the outdoor environment (infiltration and ventilation). Nevertheless, the TEB model has undergone additional development and improvement over the years as well as validation efforts that verify its ability to accurately represent the urban surface energy balance (e.g. Pigeon et al., 2008).

Kikegawa et al. (2003) presented one of the first studies that integrated building energy simulation results with an urban canopy meteorological model. Their building energy submodel explicitly accounted for building occupants, radiative transfer through windows, type of air-conditioning heat exchanger (air-cooled vs evaporatively cooled), and performance of air-conditioning systems. As a result of using such a detailed building energy submodel, they were able to partition HVAC heat rejection between latent and sensible components and provide a more realistic estimate of anthropogenic emissions. In a follow-on study, Kondo and Kikegawa (2003) used this modelling approach to explore the relationship between anthropogenic energy use and air temperature within the urban canopy. In this study, they simply assumed a one-to-one ratio between latent and sensible emissions from air-conditioning equipment. In 2006, the same research team published a study using this modelling approach to explore impacts of urban heat island countermeasures on building sector air-conditioning energy use (Kikegawa et al., 2006). This work was followed by an application where they applied their building energy modelling approach to explore the impact of anthropogenic heat on summertime urban air temperatures in Tokyo (Ohashi et al., 2007). In that study, they found that air-conditioning heat rejection resulted in summertime air temperature elevation of 1 to 2 °C in Tokyo. It should be emphasised that, as with most building energy focused modelling approaches, this work ignored anthropogenic heat from other sources (vehicles and industry). In a similar set of studies, Dhakal and colleagues (Dhakal et al., 2003, 2004) used building energy simulation to estimate the magnitude and timing of anthropogenic heat emissions from residential and commercial buildings in Tokyo. They were among the first to document the time lag and magnitude difference between energy consumption and heat emissions.

In another category of modelling studies, researchers have integrated detailed building energy simulations for prototypical buildings with geographical information system data bases containing characteristics of building types and sizes. Such studies allow researchers to model a representative building stock and then for each land use parcel (or taxlot) within a domain associate a floor-space scaled anthropogenic heating and moisture emissions from the corresponding building prototype. As illustrated in Figure 4, the anthropogenic emissions data can then be aggregated from the taxlot scale to the scale of the atmospheric model grid cell. Hsieh et al. (2007) conducted such a study for Taipei, Taiwan using the EnergyPlus building energy simulation tool and a total of 10 building prototypes (4 residential and 6 commercial). Heiple and Sailor (2008) conducted a very similar study in which they used a total of 30 building prototypes (8 residential and 22 commercial) to represent the building stock in Houston, Texas. Initially they simply applied this approach to estimate hourly energy consumption at grid resolutions of 100 m. This approach was later implemented in a study that extracted building sector anthropogenic sensible heat and moisture and linked these data to an inventory-based transportation sector model (Sailor et al., 2007). This study estimated that, due to significant environmental loads, heat rejection from buildings can be 50 to 100% greater than the energy consumption of the building. It also found that 50 to 80% of the heat rejected from buildings in the commercial core of a city such as Houston (with many large buildings with evaporative cooling towers) is rejected as latent heat. While the generation of building prototypes may appear to be a significant obstacle to implementing this method, it should be noted that there are many efforts in the building energy research community that have addressed this need, beginning with the work of Huang et al. (1991) and culminating with the recent release of the US Department of Energy's commercial building benchmarks—a compilation of prototypes for 16 building types across 16 climate zones (Torcellini et al., 2008). Such prototype models are readily linked at the taxlot level to GIS data bases containing building size and categorisation data.

thumbnail image

Figure 4. A series of prototypical buildings such as (a) are modelled and then mapped to individual tax lots wihin a GIS (b), and then aggregated up to an atmospheric model grid cell (c)

Download figure to PowerPoint

In these sorts of building energy modelling approaches, it is important to note that the building energy simulation tool (e.g. EnergyPlus) requires input meteorological data. The simulations are usually driven by typical meteorological year (TMY) data files that are a climatological representation of typical weather at the nearest airport weather station. Such representations have the obvious drawbacks of possible differences between the urban location under study and the airport site that was used to generate the TMY file. Of course, it is also possible to use actual measurements from a local weather station. For the most part, however, these building sector prototype approaches have focused on generating representative profiles of anthropogenic emissions of heat and moisture rather than estimates for any specific episode. The latter would require a dynamic linking of an urban atmospheric model and an anthropogenic heating model.

4. Inter-comparison of methods

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Sources of anthropogenic heat and moisture
  5. 3. Methods for estimating anthropogenic heat and moisture emissions
  6. 4. Inter-comparison of methods
  7. 5. Conclusions
  8. Acknowledgements
  9. References

The three approaches to estimating anthropogenic heating in urban environments each have their limitations and represent different pieces of the anthropogenic emissions puzzle. The inventory approach is problematic due to the available resolution of data in both space and time. While electricity and traffic data may be available hourly, the electricity data are typically at the scale of the city or the utility district and the traffic data typically are limited to counters on major freeways. Furthermore, it is often difficult to estimate the distribution of vehicles in the on-road fleet at any hour of the day, leading to significant uncertainty in fleet fuel economy. Energy consumed in buildings for space heating is typically more difficult to resolve temporally than electricity used in that natural gas consumption data are not usually available at the hourly or even daily time scales. As a result, the inventory approach implements a variety of techniques to map coarse resolution data onto finer scales. It is also problematic in that its representation of the building sector assumes that the energy consumption is equal to anthropogenic sensible heat—an assumption that is clearly inaccurate and not easily corrected. On the positive side, inventories lend themselves to widespread use. The data resource needed to estimate energy use in one city often will have similar data available for many cities.

The energy budget residual approach suffers from uncertainties associated with the accuracy with individual measurement of heat flux terms. It also makes the assumption that the micrometeorological flux measurements all have comparable flux footprints. Perhaps the most significant drawback, however, is that this approach requires a dedicated micrometeorological flux tower for each location where anthropogenic heat data are desired. Such a tower is expensive, and must be in place for a full year or more to gather useful data. Placement of towers in non-homogenous settings (most urban locations) invalidates the assumptions necessary for eddy flux measurements, and such measurements cannot readily be made in downtown settings, particularly with tall buildings. Thus, the most useful aspect of the energy budget residual approach is as a tool for local validation of other estimation approaches.

The building energy modelling approach corrects many of the problems associated with the inventory approach. It enables inclusion of time-dependant occupancy, energy use, environmental loads, and HVAC schedules in buildings. It also allows representation of both anthropogenic sensible heat and moisture emissions. When implemented in a prototypical building analysis framework and linked with a GIS database of building data, it is both data intensive and powerful. Of course, to be useful, it must be linked with an inventory-based estimate of emissions from the transportation and industry sectors. Unfortunately, while various forms of traffic models are relatively common in the air quality and transportation communities for estimation of mobile source pollutant emissions and congestion (see the discussions of Kinnee et al., 2004; Smit et al., 2008), such rigorous modelling efforts have not yet been applied to the problem of estimating anthropogenic heat and moisture emissions from the transportation sector.

5. Conclusions

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Sources of anthropogenic heat and moisture
  5. 3. Methods for estimating anthropogenic heat and moisture emissions
  6. 4. Inter-comparison of methods
  7. 5. Conclusions
  8. Acknowledgements
  9. References

The ultimate goal of estimating anthropogenic heat and moisture emissions is to develop information that can readily be incorporated into an urban atmospheric model. As a result, there is a need for a balance between accuracy and simplicity. It is particularly important that any modelling approach adopted be readily integrated into an atmospheric modelling code and that the necessary data be obtained with a modest amount of effort for a large number of cities. This author suggests then that the necessary characteristics of any anthropogenic heat and moisture emissions modelling approach should be as follows: (1) it should accurately account for all major sources of energy use including buildings, industry, and vehicles; (2) it should account for both moisture and sensible heat emissions; and (3) it should be based on easily constructed and accessed data bases of urban characteristics (e.g. GIS land use data resources, transportation, and building energy statistics). On the basis of the survey of the anthropogenic heating literature presented here, it is this author's opinion that the ideal approach to creating anthropogenic heat and moisture emissions inputs for atmospheric models is a combination of an inventory approach for industrial and transportation sector emissions, and a simplified building energy modelling approach for the building sector. This modelling framework is summarised conceptually in Figure 5. The transportation sector model should include a GIS-based road link model that enables differential assignment to major and minor roadways and integration of mobile source emissions modelling techniques to develop accurate temporal profiles of spatially explicit vehicle emissions. The building model should involve prototypical building simulations that represent actual building characteristics (internal loads, environmental loads, HVAC schedules, and equipment) linked to a GIS resource that enables easily scaled assignment of heat and moisture emissions to rasterised land use parcels. Ultimately, all of the modelling techniques, simulation tools, and data resources necessary to create a comprehensive anthropogenic emissions data resource are already available within many developed countries.

thumbnail image

Figure 5. A proposed framework for using inventory approaches in conjunction with building energy modelling approaches to arrive at comprehensive estimates of sensible and latent heat emission, which can then be validated using local micrometeorological measurement

Download figure to PowerPoint

There is clearly a need in the atmospheric modelling community for anthropogenic heating data. At the present time, however, the development of anthropogenic emissions estimates for urban modelling are generally done on an application-specific basis with no standardisation. Just as global data sets for land use and land cover have been developed for atmospheric modelling applications, it is now time for the atmospheric modelling community to work towards the development of a similar anthropogenic heat and moisture emissions data product.

Acknowledgements

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Sources of anthropogenic heat and moisture
  5. 3. Methods for estimating anthropogenic heat and moisture emissions
  6. 4. Inter-comparison of methods
  7. 5. Conclusions
  8. Acknowledgements
  9. References

This material is based upon work supported by the National Science Foundation under Grant No. 0410103. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.

References

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Sources of anthropogenic heat and moisture
  5. 3. Methods for estimating anthropogenic heat and moisture emissions
  6. 4. Inter-comparison of methods
  7. 5. Conclusions
  8. Acknowledgements
  9. References
  • Annamalai K, Ishwar KP. 2006. Combustion Science and Engineering. Computational Mechanics and Applied Analysis, CRC Press: Boca Raton, FL, USA; 1184 pp.
  • Anthes RA. 1990. Recent applications of the Penn State/NCAR mesoscale model to synoptic, mesoscale, and climate studies. Bulletin of the American Meteorological Society 71: 16101629.
  • Block A, Keuler K, Schaller E. 2004. Impacts of anthropogenic heat on regional climate patterns. Geophysical Research Letters 31:L12211, 14.
  • Dhakal S, Hanaki K, Hiramatsu A. 2003. Estimation of heat discharges by residential buildings in Tokyo. Energy Conversion and Management 44: 14871499.
  • Dhakal S, Hanaki K, Hiramatsu A. 2004. Heat discharges from an office building in Tokyo using DOE-2. Energy Conversion and Management 45: 11071118.
  • Dupont S, Mestayer PG. 2006. Parameterization of the urban energy budget with the submesoscale soil model. Journal of Applied Meteorology and Climatology 45: 17441765.
  • EIA. A Look at Commercial Buildings in 1995: Characteristics, Energy Consumption, and Energy Expenditures. DOE/EIA-0625 (95), 1998. 402 pp.
  • EIA. A Look at Residential Energy Consumption in 1997. DOE/EIA-0632 (97), 1999. 310 pp.
  • EIA. Natural Gas Monthly. 2003a. Available online from www.eia.doe.gov [Access 2009].
  • EIA. Electric Power Monthly. 2003b. Available online from www.eia.doe.gov [Access 2009].
  • EIA. Form EIA-826 Database: Monthly Electric Utility Sales and Revenue Data. 2004. Available online from http://www.eia.doe.gov/cneaf/electricity/page/eia826.html [Access 2009].
  • EIA. Annual Energy Review 2005DOE/EIA-0384(2005). 2005.
  • Fan H Sailor DJ. 2005. Modeling the impacts of anthropogenic heating on the urban climate of Philadelphia: a comparison of implementations in two PBL schemes. Atmospheric Environment 39: 7384.
  • Fanger PO. 1972. Thermal Comfort: Analysis and Applications in Environmental Engineering. McGraw-Hill: 244 pp.
  • Flanner MG. 2009. Integrating anthropogenic heat flux with global climate models. Geophysical Research Letters 36: L02801, DOI: 10.1029/2008GLO36465.
  • Grimmond CSB. 1992. The suburban energy balance: methodological considerations and results for a mid-latitude west coast city under winter and spring conditions. International Journal of Climatology 12: 481497.
  • Guyton AC. 1986. Textbook of Medical Physiology. WB Saunders: 1057 pp.
  • Hallenbeck M, Rice M, Smith B, Cornell-Martinez C, Wilkinson J. 1997. Vehicle Volume Distribution by Classification, 54 pp.
  • Hamilton IG, Davies M, Steadman P, Stone A, Ridley I, Evans S. 2009. The significance of the anthropogenic heat emissions of London's buildings: A comparison against captured shortwave solar radiation. Building and Environment 44: 807817.
  • Harrison R, McGoldrick B, Williams CGB. 1984. Artificial heat release from Greater London, 1971-1976. Atmospheric Environment 18: 22912304.
  • Heiple SC, Sailor DJ. 2008. Using building energy simulation and geospatial modeling techniques to determine high resolution building sector energy consumption profiles. Energy and Buildings 40: 14261436.
  • Hsieh C-M, Aramaki T, Hanaki K. 2007. Estimation of heat rejection based on the air conditioner use time and its mitigation from buildings in Taipei City. Building and Environment 42: 31253137.
  • Huang YJ, Akbari H, Rainer L, Ritschard R. 1991. 481 Prototypical Commercial Buildings for 20 Urban Market Areas. Technical Documentation LBL-29798, 208 pp.
  • Ichinose T, Hanaki K, Matsuo T. 1994. Analyses on geographical distribution of urban anthropogenic heat based on very precise geographical information. Environmental Engineering Research 31: 263273 (in Japanese with English Abstract).
  • Ichinose T, Shimodozono K, Hanaki K. 1999. Impact of anthropogenic heat on urban climate in Tokyo. Atmospheric Environment 33: 38973909.
  • Kato S, Yamaguchi Y. 2005. Analysis of urban heat-island effect using ASTER and ETM+ Data: Separation of anthropogenic heat discharge and natural heat radiation from sensible heat flux. Remote Sensing of Environment 99: 4454.
  • Kikegawa Y, Genchi Y, Yoshikado H, Kondo H. 2003. Development of a numerical simulation system toward comprehensive assessments of urban warming countermeasures including their impacts upon the urban buildings' energy-demands. Applied Energy 76: 449466.
  • Kikegawa Y, Genchi Y, Kondo H, Hanaki K. 2006. Impacts of city-block-scale countermeasures against urban heat-island phenomena upon a building's energy-consumption for air-conditioning. Applied Energy 83: 649668.
  • Kimura F, Takahashi S. 1991. The effects of land-use and anthropogenic heating on the surface temperature in the Tokyo Metropolitan area: a numerical experiment. Atmospheric Environment, Part B 25 B: 155164.
  • Kinnee EJ, Touma JS, Mason R, Thurman J, Beidler A, Bailey C, Cook R. 2004. Allocation of onroad mobile emissions to road segments for air toxics modeling in an urban area. Transportation Research Part D: Transport and Environment 9: 139150.
  • Klysik K. 1996. Spatial and seasonal distribution of anthropogenic heat emissions in Lodz, Poland. Atmospheric Environment 30: 33973404.
  • Kondo H, Kikegawa Y. 2003. Temperature variation in the urban canopy with anthropogenic energy use. Pure and Applied Geophysics 160: 317324.
  • Lee SH, Song CK, Baik JJ, Park SU. 2009. Estimation of anthropogenic heat emission in the Gyeong-In region of Korea. Theoretical and Applied Climatology 96: 291303.
  • Lemonsu A, Masson V. 2002. Simulation of a summer urban breeze over Paris. Boundary-Layer Meteorology 104: 463490.
  • Martilli A, Clappier A, Rotach MW. 2002. An urban surface exchange parameterisation for messocale models. Boundary Layer Meteorology 104: 261304.
  • Masson V. 2000. A physically-based scheme for the urban energy budget in atmospheric models. Boundary-Layer Meteorology 94: 357397.
  • Myrup LO. 1969. A numerical Model of the Urban Heat Island. Journal of Applied Meteorology 8: 908918.
  • Offerle B, Grimmond CSB, Fortuniak K. 2005. Heat storage and anthropogenic heat flux in relation to the energy balance of a central European city centre. International Journal of Climatology 25: 14051419.
  • Ohashi Y, Genchi Y, Kondo H, Kikegawa Y, Yoshikado H, Hirano Y. 2007. Influence of air-conditioning waste heat on air temperature in Tokyo during summer: numerical experiments using an urban canopy model coupled with a building energy model. Journal of Applied Meteorology and Climatology 46: 6681.
  • Oke TR. 1988. The urban energy balance. Progress in Physical Geography. 12: 471508.
  • Otte TL, Lacser A, Dupont S, Ching JKS. 2004. Implementation of an urban canopy parameterization in a mesoscale meteorological model. Journal of Applied Meteorology 43: 16481665.
  • Pigeon G, Legain D, Durand P, Masson V. 2007. Anthropogenic heat release in an old European agglomeration (Toulouse, France). International Journal of Climatology 27: 19691981.
  • Pigeon G, Moscicki MA, Voogt JA, Masson V. 2008. Simulation of fall and winter surface energy balance over a dense urban area using the TEB scheme. Meteorology and Atmospheric Physics 102: 159171.
  • Richiardone R, Brusasca G. 1989. Numerical experiments on urban heat island intensity. Quarterly Journal of Royal Meteorological Society 115: 983995.
  • Sailor DJ. 1995. Simulated urban climate response to modifications in surface albedo and vegetative cover. Journal of Applied Meteorology 34: 16941704.
  • Sailor DJ,Brooks A. American Meteorological Society. 2009. Quantifying anthropogenic moisture emissions and their potential impact on the urban climate. 8th Symposium on the Urban Environment: Phoenix, AZ, USA.
  • Sailor DJ, Brooks A, Hart M, Heiple S. American Meteorological Society. 2007. A bottom-up approach for estimating latent and sensible heat emissions from anthropogenic sources. 7th Symposium on the Urban Environment, San Diego.
  • Sailor DJ, Fan H. 2004. The importance of including anthropogenic heating in mesoscale modeling of the urban heat island. 84th annual meeting of the AMS, Symposium on Planning, Nowcasting, and Forecasting in the Urban Zone, Seattle.
  • Sailor DJ, Hart M. American Meteorological Society. 2006. An anthropogenic heating database for major U.S. cities. 6th Symposium on the Urban Environment, Atlanta.
  • Sailor DJ, Lu L. 2004. A top-down methodology for developing diurnal and seasonal anthropogenic heating profiles for urban areas. Atmospheric Environment 38: 27372748.
  • Smit R, Brown AL, Chan YC. 2008. Do air pollution emissions and fuel consumption models for roadways include the effects of congestion in the roadway traffic flow? Environmental Modelling and Software 23: 12621270.
  • Taha H. 1996. Modeling impacts of increased urban vegetation on ozone air quality in the south coast air basin. Atmospheric Environment 30: 34233430.
  • Taha H. 1997. Urban climates and heat islands: albedo, evapotranspiration, and anthropogenic heat. Energy and Buildings 25: 99103.
  • Torcellini P, Deru M, Griffith B, Benne K, Halverson M, Winiarski D, Crawley DB.2008. DOE Commercial Building Benchmark Models, 15 pp.
  • Torrance KE, Shum JSW. 1975. Time-varying energy consumption as a factor in urban climate. Atmospheric Environment 10: 329337.
  • Tso CP, Chan BK, Hashim MA. 1990. An improvement to the basic energy balance model for urban thermal environmental analysis. Energy and Buildings 14: 143152.
  • USDoT. 2003. Urbanized Areas—2000 Selected Characteristics. www.fhwa.dot.gov [Access year 2009].