Although anthropogenic dust has received more attention from the climate
research community, its dominant role in the production process is still not
identified. In this study, we analysed the relationship between anthropogenic
dust and population density/change over global semi-arid regions and found
that semi-arid regions are major source regions in producing anthropogenic dust.
The results showed that the relationship between anthropogenic dust and
population is more obvious in cropland than in other land cover types (crop
mosaics, grassland, and urbanized regions) and that the production of
anthropogenic dust increases as the population density grows to more
than 90 persons km
It is well known that anthropogenic activities play an important role in drylands' climate change. Salinization, desertification, loss of vegetative cover, loss of biodiversity, and other forms of environmental deterioration are partly caused by anthropogenic activities (Huang et al., 2016a, b). With rapid economic development, more fossil fuels have been consumed, which produced a great deal of greenhouse gases (GHGs) as well as energy (Barnett and O'Neill, 2010). The released GHGs and heat have induced a strong influence on temperature spatial distribution in recent years (Li and Zhao, 2012), especially in developing countries, where the economic policy relies on extensive growth, which favours results, despite lower resource efficiency and energy waste.
Jiang and Hardee (2011) noted that main factors influencing anthropogenic effects on aerosol emission are economic growth, technological change, and population growth, which cannot be easily simulated using numerical models (Zhou et al., 2010). Recently, better understanding about the effects of human activities on dryland expansion in various scenarios has been achieved (Huang et al., 2016b). It appears that higher densities of younger workers are strongly correlated with increased energy use (Liddle, 2004), carbon dioxide emission (Liddle and Lung, 2010; H. Huang et al., 2014), and energy consumption, and the combined production of heat has been released into the atmosphere along with GHGs. Although human activities play an important role in the process of regional climate change, our understanding of their relationship is extremely limited, especially in drylands (Jiang, 2010).
Huang et al. (2012) showed that drylands are most sensitive to global warming; this warming was induced by dynamical and radiative factors. Guan et al. (2015a) found that the enhanced warming in drylands was a result of radiative-forced temperature, which has a close relationship with aerosol column burden. The aerosol in drylands has an obvious warming effect (Huang et al., 2006a, 2008; Chen et al., 2010; Ye et al., 2012; Jin et al., 2015), and the aerosol has a wide distribution and tends to have a relatively large optical depth (H. Huang et al., 2010; Bi et al., 2011; Liu et al., 2011; Xu and Wang, 2015; Xu et al., 2015), leading to a significant radiative effect in the drylands. According to the results of Tegen and Fung (1995), the existing atmospheric dust load is hard to explain by natural sources alone. The atmospheric dust load that originates from soil and is disturbed by human activities, such as various land-use practices, can increase the overall dust load and in turn affect radiative forcing. Efforts to quantify the relative importance of different types of dust sources and the factors that affect dust emissions are critical for understanding the global dust cycle, as well as historical and possible future changes in dust emission (Okin et al., 2011; Huang et al., 2015). Therefore, studies on different types of aerosols are necessary in the study of radiative effect (J. Huang et al., 2009, 2014; Wang et al., 2010; Yi et al., 2014).
Generally, the aerosols in drylands are divided into two categories: natural and anthropogenic dusts. Anthropogenic dust originates predominantly from agricultural practices (e.g. harvesting, ploughing, and overgrazing) and changes in surface water (e.g. shrinking of the Caspian Sea, the Aral Sea, and Owens Lake), as well as urban (e.g. construction) and industrial practices (e.g. cement production and transport) (Prospero et al., 2002). Over the past few decades, a combination of higher frequency of warmer and dryer winters/springs in semi-arid and semi-wet regions and changes in vegetated land cover due to human activities have likely increased anthropogenic dust emission over different regions (Mahowald and Luo, 2003). Mulitza et al. (2010) studied the development of agriculture in the Sahel, which was associated with a large increase in dust emission and deposition in the region, and they found that dust deposition is related to precipitation in tropical West Africa on the century scale. Due to the importance of anthropogenic dust in climate study, Huang et al. (2015) developed a detection method of anthropogenic dust emission and presented a global distribution of anthropogenic dust aerosol. The current consensus is that up to half of the modern atmospheric dust load originated from anthropogenically disturbed soils (Tegen et al., 2004). Such a great proportion of anthropogenic dust will greatly influence local radiative forcing. Therefore, influence of human activities on production of anthropogenic dust is critical for predicting and estimating the radiative effect of aerosol in regional climate change.
Most of previous results focused on the emission of natural dust aerosol (Z. Huang et al., 2010; Li et al., 2011; Yi et al., 2011, 2012); the study on anthropogenic dust is relatively limited. In this study, the anthropogenic dust over semi-arid regions is identified by CALIPSO data, and its relationship with human activities is investigated. The method used to distinguish anthropogenic dust from the total dust aerosols is based on that of Huang et al. (2015). This paper is organized as follows. Section 2 introduces the data sets used in this study. Section 3 presents the method used to identify the anthropogenic dust aerosols in the semi-arid regions. Section 4 discusses anthropogenic dust emission over global semi-arid regions and its relationship to human activities, including a comparison among four different semi-arid regions. Our major findings, followed by a discussion of the radiative effect of anthropogenic dust on regional climate change in semi-arid regions, are given in Sect. 5.
In this study, we use the AI to classify different types of
regions. The AI is defined as the ratio of annual precipitation to annual
potential evapotranspiration, representing the degree of climatic dryness.
The AI data set used in this study (Feng and Fu, 2013; Huang et al., 2016b)
is
based on the Climate Prediction Center (CPC) data sets. Drylands are
identified as regions with AI values less than 0.65 and are further
classified into hyper-arid (AI
The population data are from the Gridded Population of the World data set,
version 3 (GPWv3,
The instrument used to detect anthropogenic dust is the CALIPSO Cloud-Aerosol
Lidar with Orthogonal Polarization (CALIOP). CALIOP acquires vertical
profiles of elastic backscatter at two wavelengths (532 and 1064 nm) and
linear depolarization at 532 nm from a near-nadir viewing geometry for both
day and night (Hu et al., 2007a, b, 2009; Liu et al., 2008). The data sets
detail the information of Level-1 backscatter, depolarization ratio, and
colour ratio profiles along with the Level-2 Vertical Feature Mask (VFM)
product and the 5 km aerosol profile product. The CALIPSO algorithm uses
volume depolarization ratio (
The Collection 5.1 MODIS global land cover type product (MCD12C1) in 2011 is
used to identify types of anthropogenic dust sources. It includes
17 different surface vegetation types and was developed based on the data
from the International Geosphere–Biosphere Programme (IGBP) (Friedl et al.,
2010), with a spatial resolution of 0.05
Recently, Huang et al. (2015) developed a new method of separating natural dust and anthropogenic dust at the global scale using CALIPSO measurements. They defined a schematic framework of dust sources and used vertical and horizontal transport processes as the foundation for their approach to discriminate anthropogenic dust from natural dust in CALIPSO data, which proceeds in a sequence of four steps. The first step is to detect the total dust load (both natural and anthropogenic). The second step is to determine the source region from which the dust originates. The third step is to determine the height of a planetary boundary layer (PBL), and the final step is to determine the proportion of dust, i.e. that subset of the total dust within the PBL.
After the anthropogenic dust was identified by the detection method described
above, the anthropogenic dust column burden was calculated as follows. First,
we determined dust extinction coefficient from the “atmospheric volume
description”, which is used to discriminate between aerosols and clouds in
the CALIPSO Level-2 aerosol extinction profile products. Then the dust
extinction coefficients with the highest confidence levels (
Figure 1 shows the global distribution of semi-arid regions along with the mean anthropogenic dust column burden from 2007 through 2010, demonstrating the wide spread of anthropogenic dust. Most of the areas with high anthropogenic dust loading are located in the middle to high latitudes of the Northern Hemisphere, such as North China, Mongolia, northern India, central western North America, and Sahel. The highest values are generally distributed throughout East China and India. Note that the Northern Hemisphere has much more anthropogenic dust than the Southern Hemisphere. Therefore, we select four geographical regions that encompass semi-arid regions and are influenced by anthropogenic dust in order to quantify the recent changes. These regions marked in Fig. 1 include East China, India, North America, and North Africa. From a visual inspection of the overlap between the anthropogenic dust distribution and the semi-arid regions, it can be seen that most semi-arid regions coincide with regions of high anthropogenic dust. However, the anthropogenic dust column burdens are different over the selected semi-arid regions: East China and India appear to have greater amounts of anthropogenic dust than North America and North Africa.
Global distribution of mean anthropogenic dust column burden
(g m
Figure 2 displays the total global anthropogenic dust column burden as a
function of climatological annual AI during the period of 1948–2004. The
mean AI varies from 0.0 to a maximum of 2.0. Note that the intervals in this
figure are non-uniform because they are from the classification standard for
different types of regions based on the AI, as defined in Sect. 2. Semi-arid
region is the transition zone between arid and semi-wet regions; it is
defined as the area where precipitation is less than potential evaporation
and is characterized by high temperatures (30–45
Total global anthropogenic dust (AD) column burden (Tg) as a function of the climatological mean aridity index (AI).
Figure 3 compares the anthropogenic dust column burdens in summer (blue), spring (green), autumn (red), and winter (black) as a function of the climatological mean AI. The curves are similar in all four seasons, and the anthropogenic dust column burden exhibits a dominant peak in semi-arid regions in all four seasons, with values much larger than those in the other regions. For the semi-arid regions, the total anthropogenic dust column burden is the greatest in summer, followed by spring, autumn, and winter, which may relate with the different frequency of human activities (Huang et al., 2015), such as the construction activity is likely to be greater in summer.
Comparison of the global anthropogenic dust (AD) column burden (Tg) in spring (green), summer (blue), autumn (red), and winter (black) as a function of the climatological mean aridity index (AI).
In order to illustrate the key role of anthropogenic dust in generating dust aerosols in the semi-arid regions, we compared the dust column burdens corresponding to natural with mixed dust (natural and anthropogenic dusts) in the semi-arid regions of the globe, North America, East China, North Africa, and India, in Fig. 4. It is evident that mixed dust aerosol column burden is greater than the pure natural dust of the globe. Both mixed and pure natural dust column burdens are the greatest in India, followed by North Africa and East China. The mixed dust burden of North America is a little less than that of the natural dust. Among these regions where the mixed dust is greater than natural dust, the difference between mixed dust and natural dust is the largest in North Africa, followed by India and East China. For the mixed dust aerosol, the dust column burdens of natural and anthropogenic dusts are presented separately in Fig. 5. It shows that the anthropogenic dust column burden is greater than that of natural dust. Additionally, the highest value of anthropogenic dust column burden is in India, followed by North Africa, East China, and North America; among these regions, the natural dust burden is the highest in North Africa, followed by India, North America, and East China.
Mean dust column burdens (g m
Table 1 reports the detailed values of the annual mean anthropogenic and
natural dust column burden from mixed dust areas over the semi-arid regions
of East China, India, North America, and North Africa. In the semi-arid
regions of India, the mean anthropogenic dust column burden is
0.38 g m
Mean anthropogenic (red) and natural (blue) dust column burdens
(g m
Mean dust column burdens (g m
Figure 6 is the distribution of mean population density. The population density in semi-arid regions exhibits dramatic regional variability. For the four selected semi-arid regions, both India and East China have higher population densities, most semi-arid regions of North Africa have relatively lower population density, and the population density in the semi-arid region of North America is the lowest. The regional difference of population indicates that influences of human activities are not uniformly distributed in the semi-arid areas. Figure 7 illustrates the global distribution of population change between 1990 and 2010. India exhibits the most obvious population change, followed by North Africa and East Asia. North America exhibits an obvious difference between east and west areas, a similar spatial pattern of population change occurred in China. The difference between these respective western and eastern areas may be related to their economic status. The eastern areas of both North America and China are more industrialized than their western counterparts. In a comparison of Figs. 6 and 7, the inconsistent distribution between population density and population change reveals that the regions with the higher population densities do not always have the more obvious population change. Population density and change are related to various factors, such as population policies, economic development status, and political divisions.
Global distribution of mean population density (persons km
Global distribution of mean population change (persons km
Figure 8 compares the mean population density and change in the four selected
regions; it is apparent that India has the highest population density, which
reaches almost 290 persons km
Mean population density (persons km
Mean population density/change (persons km
Figure 9 is the mean anthropogenic dust column burden as a function of
population density. The population varies from 0 to 400 persons km
Mean anthropogenic dust (AD) column burden (g m
Mean anthropogenic dust (AD) column burden (g m
Different land cover areas (km
In the semi-arid regions, four typical land covers in semi-arid regions are
urban, grassland, cropland, and croplands mosaics. Figure 11 shows the global
mean anthropogenic dust column burden in semi-arid region as a function of
population density over cropland (blue line), cropland mosaics (which are
lands with a mosaic of croplands less than 60 % of the landscape
according to Friedl et al., 2002; green line), urban (red line), and
grassland (orange line). For population density less than 90 persons km
Global mean anthropogenic dust (AD) column burden (g m
Different land cover area percentage (%).
The percentage of different type of land cover in the semi-arid regions
of East China, India, North America, and North Africa is illustrated in
Fig. 12a–d; they show that the components of cropland, grassland, urban, and
cropland mosaics are quite different. In the four selected regions, the
Indian semi-arid region is dominated by croplands, which has an area of
5.92
Percentage of different types of land cover in semi-arid regions of
East China
Figure 13a–d illustrate the anthropogenic dust probability distributions are
quite different in East China, India, North America, and North Africa with
intervals of population and dust column burden are 20 persons km
Anthropogenic dust (AD) probability distribution in different population
density (persons km
The comparison in Fig. 13 highlights the representative relationship between
anthropogenic dust and population in India, and Fig. 14 shows that quantified
influences of population on anthropogenic dust probability in typical
croplands of Indian semi-arid regions with intervals of population
density/change are 20 persons km
Anthropogenic dust (AD) probability as a function of population
density (persons km
In this paper, we focused on the relationship between anthropogenic dust and population. It was found that the total anthropogenic dust column of globe exhibited an obvious peak in the semi-arid regions, which were much higher than it in the other regions. Four geographical semi-arid regions of East China, India, North America, and North Africa were chosen as our study areas according to their anthropogenic dust levels and population. Both population density and population change were correlated with anthropogenic dust, indicating that these population features had effects on the production of anthropogenic dust column burden in these semi-arid regions. In particular, typical croplands in the Indian semi-arid region showed a normal relationship between anthropogenic dust with population density/change; the relationship indicated the influence of human activities on environment can be quantified in the process of climate change. Additionally, it also proposed a typical influence of human activities on anthropogenic dust in cropland.
Dust aerosols exert a key impact on regional radiative forcing over semi-arid regions (Huang et al., 2006b) and are closely related to local climate change (Guan et al., 2015b). Historical statistics revealed that population change occurs in parallel with economic growth and with increases in energy consumption, GHG emission, and anthropogenic dust. Further studies are needed to gain a better understanding of the influence of anthropogenic dust aerosols on climate change in semi-arid regions. Under the current dynamic economic conditions throughout the world, there are still many developing countries in semi-arid regions that are undergoing extensive economic development or are in the process of transforming from an extensive economic mode to an intensive economic model. Developing countries exhibit high rates of population growth, which must be considered when forming economic development strategies. In the developed countries, population change may also result in increased consumption, higher energy demands, and enhanced GHG production. Therefore, further investigations into the influence of human activities on anthropogenic dust aerosol production and the consequent impacts on regional climate change in semi-arid regions are needed, with an emphasis on understanding the feedback between regional climate change and societal development with the intent to apply more reasonable policies in the process of economic development.
This work was jointly supported by the National Basic Research Program of China (2012CB955301), the National Science Foundation of China (41305009, 41575006, 41521004, 41175084), the China 111 project (no. B 13045), and the Fundamental Research Funds for the Central Universities (lzujbky-2015-2, lzujbky-2015-ct03). Edited by: D. Covert