2.1 Study area and in-boundary FFE
The 14 reported global major cities according to the research of Chen et al. [24] are selected as the study area and include Bangkok, Beijing, Shanghai, Delhi, Cape Town, Sao Paulo, Tokyo, Greater Paris, Greater London, Los Angeles, Manhattan, New York City, Washington D.C., and Greater Toronto (see Table 4 and Additional file 1: Figure S1). The definitions of the 14 cities ranges from ‘district’ to ‘metropolitan’ (see also Table 4) [25]. The CO2 from human and livestock respiration is directly emitted within global city boundaries, which is equivalent to scope 1 [26]. To compare the CO2 emissions from human and livestock respiration within these global cities, we retrieved the in-boundary anthropogenic FFE from Chen et al. [24], who estimated the total FFE directly within the city boundaries of these 14 cities and metropolitan areas around the world.
Our study also separated city areas into two subcategories, urban and suburban. The urban extent of each city is based on the 1:10 m urban areas shapefile from Nature Earth (https://www.naturalearthdata.com), which is derived from 2002-2003 MODIS satellite data at 1 km resolution [27]. Urban areas are defined in this study as urban and built-up areas with high population densities, high radiance levels in commercial/industrial areas and high-density residential land cover, instead of being based on impervious surfaces (Schneider et al., 2009).
2.2 Estimate methods for HLR
The CO2 release from respiration of per person (HR) or per head of livestock (LR) is obtained according to the basal metabolic rate (BMR). The BMR refers to the minimum level of energy required to sustain vital functions of organs at complete rest in a neutrally temperate environment and in a fasting state. It is measured by heat production or oxygen consumption and can be expressed as Cal m-2 h-1, Cal kg-1 h-1 or O2 g-1 h-1 for individuals [29,30]. For various mammals, the oxygen consumption rate per body mass consistently decreases with increasing body size, while the rate of oxygen consumption for individuals against body mass tends to decrease along regression lines in logarithmic coordinates (birds have a similar equation to mammals) [31]. Additionally, oxygen is combined with carbon according to the respiration reaction. Therefore, based on the BMR of each species, we can estimate the CO2 produced by respiration according to the oxygen consumption.
The BMR of humans was given as 6279 kJ day-1 per person by Johnstone et al. [32] (minimum=4301 kJ day-1, maximum=10455 kJ day-1), which was the mean of seven experiments including 155 adults between the ages of 21 and 64. We also estimated the human BMR as 5698 kJ day-1 per person (minimum=4778 kJ day-1, maximum=6612 kJ day-1), which was the weighted average for different age groups and for both sexes from the daily BMR of 92 individuals predicted by the FAO (Additional file 1: Table S1) [33]. The fractions for different ages and both sexes for the global total population come from the World Bank (The World Bank: Featured indictors of Health, 2019). We take the average of the two sources as the globally averaged BMR and convert the heat production (kJ day-1) into oxygen consumption (L O2 day-1) by introducing the thermal equivalent of oxygen (20.2 kJ L-1). Finally, the HR is approximately 57.97 kg C yr-1.
The BMR (ml O2 g-1 h-1) of mammalian livestock and chickens are measured values from previous experimental results that controlled the environmental temperature, nutrition, age and activity level [35–37]. The LR is estimated from the following equation:
where BMRl is the BMR, with units of ml O2 g-1 h-1; Body_weightl is the average of different breeding ages and genders; is the molecular mass of O2 in g mol-1; and Vm is the molar volume of gas in 22.4 L mol-1. The ratio of carbon (C) and O2 is set to 12/32 according to the processes of respiration, which can be expressed by the following chemical equation [38]:
where CH20 represents the composition of biological material.
The amount of LR represents the total carbon released during the days the animals are alive (see Table 2). Therefore, we assumed that the life span of poultry is 42 days, that the life span of pigs is half a year [39,40], and that all species except poultry and pigs live for more than one year.
Table 2. The parameters of eight types of livestock.
Livestock
|
BMR
(kg O2 yr-1)
|
LR 1
(kg C yr-1 per head)
|
Global total production (million head in 2010)2
|
References
|
horse
|
813.43
|
305.04
|
59.66
|
M. A. Elgar and P. H. Harvey (1987)
|
pig
|
103.24
|
19.36
|
974.41
|
cattle and buffalo
|
578.66
|
217.00
|
1603.86
|
goats
|
85.60
|
32.10
|
910.83
|
sheep
|
127.65
|
47.87
|
1076.36
|
chicken and duck
|
26.95
|
1.25
|
22311.21
|
B. M. Freeman (1963)
|
- LR is the CO2 release from respiration of per head of livestock.
- Data coms from FAOSTAT, http://www.fao.org/faostat/en/#home.
What’s more, concerning the metabolic enhancement caused by exercise metabolism and other factors, the physical activity level (PAL) was defined in terms of three levels of physical activity [41]. For simplicity, we assume that the WHO recommended PAL=1.55 could be used as an uniform parameter for global countries and for different gender and age groups for both human and livestock [42]. Finally, the HR is 89.90 kg C yr-1.
The total HLR is estimated by multiplying the CO2 emission of each individual by the total population/livestock production within city boundaries. Since the population and livestock production are reported as high-resolution datasets (see sector 2.3), the HLR in grid i is estimated by HR and LR, as well as the population/production in each grid based on the following equations:
where h is for humans; l is for the species of livestock with a total of eight in this study; and Livestock(i) is the total production of each species of livestock in grid i.
2.3 Datasets of humans and livestock
The HLR in each city are extracted from high-resolution vector datasets (see Table 3). The Gridded Livestock of the World (GLW) datasets include global distributions of eight major livestock species (also see Additional file 1: Table S2). It should be noted that the total cattle and poultry production in Beijing from the high-resolution datasets is 17 times higher than the statistical data from the National Bureau of Statistics of China (NBS, http://data.stats.gov.cn/english/), while cattle production is consistent with census statistics for Shanghai, Delhi and Sao Paulo (Additional file 1: Table S3-S4). As the detailed cattle census statistics of Beijing for GLW were mined from statistical yearbooks of China, we consider the values from official source of China are more reliable. Thus, we first corrected the livestock production in each grid in Beijing according to the spatial distribution from GLW and the total livestock production from NBS.
The city boundaries we used in this study come from the database of Global Administrative Areas (GADM) version 2.0 (http://gadm.org/). The shapefile with polygon features of 14 cities was first converted to a high-resolution vector-form dataset at a resolution of 30 arc-second and can be used as a region mask to extract values for population and livestock production within different cities.
To match the years with FFE in 14 cities from Chen et al [24], we used linear interpolation to obtain the annual human respiration after extracting the total CO2 release from each city every 5 years. For livestock, we assume that the trend of livestock production in each city is close to that of the country (for the data, please see Table 3). We obtained the livestock respiration in each country through the method described in section 2.1. Finally, the total annual livestock respiration within cities is scaled from that of each country.
Table 3. Sources of data on humans and livestock
Data
|
Data source
|
Resolution
|
Time range
|
Gridded Population of the World, Version 3 (GPWv3)2
|
Socioeconomic Data and Applications Center (SEDAC)
|
2.5 arc-minute regridded to 30 arc-second
|
1990, 1995
|
Gridded Population of the World, Version 4 (GPWv4)2
|
30 arc-second
|
2000, 2005, 2010, 2015
|
Gridded Livestock of the World (GLW)1
|
FAO
|
5 arc-minutes regridded to 30 arc-second
|
2010
|
Livestock production3
|
FAOSTAT
|
national total
|
1960-2014
|
- Gilbert et al., 2018b, 2018c, 2018a, 2018d, 2018e, 2018f, 2018g, 2018h
- Center for International Earth Science Information Network - CIESIN - Columbia University, 2018; Center for International Earth Science Information Network - CIESIN - Columbia University and Centro International de Agricultural Tropical - CIAT, 2005
- FAO. FAOSTAT, 2017