Comprehensive Assessment of the Effect of Urban Built-Up Land Expansion and Climate Change on Net Primary Productivity

Key Laboratory of Geospatial Technology for Middle and Lower Yellow River Region, Research Center of Regional Development and Planning, Institute of Agriculture and Rural Sustainable Development, Henan Overseas Expertise Introduction Center for Discipline Innovation (Ecological Protection and Rural Revitalization Along the Yellow River), Henan University, Kaifeng 475004, China Collaborative Innovation Center on Yellow River Civilization of Henan Province, Henan University, Kaifeng 475001, China Key Laboratory of Guang dong for Utilization of Remote Sensing and Geographical Information System, Guang dong Open Laboratory of Geospatial Information Technology and Application, Guangzhou Institute of Geography, Guangzhou, China


Introduction
As an important part of the global carbon cycle, the terrestrial biosphere is affected by urban expansion and climate change [1][2][3]. e trend of global terrestrial net primary productivity (NPP) is still uncertain [4]. In the context of climate change, the threat of rapid global urbanization to terrestrial ecosystem productivity, environment, livelihoods, and food security has gradually become one of the most critical issues in the world [5,6]. e net primary productivity refers to the amount of organic matter accumulated by green plants in unit time per unit area [7,8]. It is an important indicator in the determination of the carbon source, carbon sink health, and sustainable development of ecosystems and is the main factor regulating ecological processes. It also helps in assessing the carrying capacity of ecosystems [9][10][11]. Terrestrial vegetation provides a great deal of food, fuel, and building materials for human beings; therefore, in the context of global change, more and more researchers are beginning to pay attention to the trend of NPP in terrestrial ecosystems [12].
With the rapid growth in the world economy, urbanization is increasing. Urbanization is a complex process involving population transfer, land use change, urban function change, and urban form [13,14]. Urbanizationinduced changes may have a significant impact on the ground and thus affect the structure and function of ecosystems, as well as regional climates [15][16][17]. erefore, studying the response of NPP to urban expansion and climate change can provide a better understanding of the function of ecosystems, which is important for balancing the relationship between development and environment and for the rational use of natural resources [18][19][20]. e NPP is an important ecological indicator for judging sustainable development and can help assess the carbon budget of terrestrial ecosystems [9,21,22]. e NPP has been widely used to monitor the state of carbon cycles in regions of different sizes [23][24][25]. Changes in NPP over a specific period can help quantify vegetation growth, which is related to the amount of vegetation and the environment in which it grows. Different models have been used to enrich the research results provided by NPP trends, primarily at a global level [26,27], at a national level [28], and in ecosensitive areas [22,29]. By comparing annual and seasonal NPP estimates from 15 global models in latitude zones and biomes, Cramer et al. [26] found that NPP estimates vary over time and space. Most previous research studies have been conducted at the global, national, or other macrolevels; studies at the city level are limited. e impact of urbanization on terrestrial ecosystems has been assessed based on the NPP indicator [30][31][32]. In addition, model estimation is a convenient method for determining NPP, as field measurements require significant human and material resources and data on urban areas, which are difficult to obtain. With the development of remote sensing technology, the surface information of any region can be comprehensively and continuously obtained [26,33]. To analyze the impact of urban expansion on NPP in the past few decades, a longterm NPP time series with a high time resolution is required. e spatial distribution of NPP in urban areas can be determined using data from the moderate resolution imaging spectroradiometer (MODIS), with proven accuracy. e spatial changes in NPP in urban areas can thus be better reflected.
e response of NPP to urban expansion and the factors influencing NPP have been widely studied [34,35]. Most studies have shown that urban landscape and land use changes lead to carbon loss [36] and that land use changes have a negative impact on urban NPP [37]. However, these studies employed two time nodes when quantifying the impact of land use change on NPP and only few studied the spatialization of land expansion by urbanization and the impact of land expansion on NPP from a time and space perspective. Several scholars have relied on DMSP/OLS night light data to carry out a wide range of spatial research, such as those on population migration [38], anthropogenic carbon emissions [39,40], and night lighting data particularly for determining built-up land and other aspects for a wide range of applications [41,42], with good results. However, these studies were limited to the time period from 1992 to 2013; only few scholars have connected MODIS and NPP-VIIRS (Visible Infrared Imaging Radiometer Suite (VIIRS) onboard the Suomi National Polar-Orbiting Partnership (Suomi NPP) spacecraft) to simulate urban expansion [43]. Appropriate time series updates are thus required. e factors influencing NPP include land use change, vegetation, climate, and topography. Although the factors influencing NPP have been widely studied, particularly the correlation between NPP and climate factors, the studies have primarily focused on a single dynamic change perspective. In other words, the correlation between climate change and NPP change was analyzed by establishing a relationship between them, with few comprehensive studies on the impact of land expansion combined with climate factors on NPP. erefore, quantifying the impact of human activities and climate change on NPP is an important step in formulating sustainable development of urban ecosystems in the context of climate change and human activities.
Henan province is located in central China, has a temperate continental climate, and is suitable for multicrop growth. is province is known for its agriculture, grain, and population. Zhengzhou is the capital of Henan province and is China's national center. In recent years, with the change in national policies, urbanization and industrialization have been taking place at a rapid rate, leading to an increase in the number of urban areas and population explosion. e urban landscape in Zhengzhou has undergone rapid changes, which has affected the local ecological environment. In this study, we used relevant data to analyze the response of NPP to urban expansion and climate change in Zhengzhou during the period of 2000-2015. We provide a reference for urban land use and ecological environment development in the regional centers of developing and agricultural countries.
is study can fill the gaps discussed above, including broadening the time series of night lighting data, innovatively combining two kinds of night lighting data to simulate the change of urban construction land, and comprehensive studies on the impact of land expansion combined with climate factors on NPP. Strong data and new method will enhance the accuracy compare with previous studies. Detailed contents of this research include the following: (1) analysis of land use type changes and NPP changes; (2) analysis of the relationships between NPP and climate factors; (3) analysis of changes in land use intensity and urban built-up area; (4) exploration of the impact of urban built-up land on NPP change.

Study Area.
Zhengzhou is the capital of Henan province, bordering the Yellow River in the north (Figure 1). It is located at 112°42′-114°14′E and 34°16′-34°58′N. e region has a warm temperate continental climate, with an annual precipitation of approximately 639.2 mm and an average 2 Complexity annual temperature of 14.2°C. e dominant striped vegetation is the temperate deciduous evergreen mixed broadleaved forest belt, and the distribution of the flora is in the middle north temperate zone and east Asia. By 2018, Zhengzhou had a total area of 7446 km 2 , a built-up land area of 830.97 km 2 , a total population of 101.36 million, and a total GDP of 101.433 billion yuan. In recent years, due to policy guidance, urbanization in Zhengzhou has been accelerating, and the urbanization rate in 2018 reached the top of the national rate of growth, at 1.59%. is rapid urbanization process is accompanied by several urban land development projects; land use change is thus inevitable.

Data Sources.
e datasets used in this study and data preprocessing conducted are as follows: (  Figure 1: e location of study area. Complexity persistent light sources, with the background noise eliminated. e NPP-VIIRS NSL data for the period of 2014-2015 were obtained from the National Environmental Information Center website (https:// www.ngdc.noaa.gov). Prior to data processing, the monthly average data of 2014 and 2015 from January to December were combined into annual data through ENVI 5.1. NPP-VIIRS NSL data processing included noise cancellation and continuity correction of OLS night light data using DMSP-OLS. First, the DMSP-OLS night light data of 2013 were extracted as a dark background mask, and the mask was then used to remove unexpected noise from the NPP-VIIRS night lighting data for 2014 and 2015.
Second, according to Li et al. [45], the average light value (DN) of the NPP-VIIRS night light data is exponentially correlated with the DN value of DMSP-OLS night light data. Accordingly, we can obtain the corrected NPP-VIIRS night light data. e formula is as follows: After further processing, equation (1) can be converted to the following: Here, Y represents the DN value of DMSP-OLS night light data, X represents the DN value of NPP-VIIRS night light data, and a and b are coefficients. (4) e temperature and precipitation data in the 2000-2015 time series were derived from the Resource Environmental Science Data Center of the Chinese Academy of Sciences (http://www.resdc.cn/doi/doi.aspx?doiid�32).

NPP Trend Analysis.
e trend of change in NPP at the cell level is analyzed and predicted using a one-way linear regression analysis, and the formula is as follows: Here, n represents the number of years (n � 16), NPP i is the NPP for year i (i � 1, 2, 3, . . ., 16), and slope is the slope for the NPP at the individual cell level of the slope. If the slope > 0, an increasing trend is indicated; the greater the value, the more evident the increasing trend. If the slope < 0, a decreasing trend is indicated; the lower the value, the more evident the decreasing trend.

Correlation Analysis.
e trend of change in NPP and the temperature and precipitation correlation coefficient on the space-time scale can be calculated using the Pearson correlation coefficient method. e correlation coefficient (R xy ) is calculated as follows: Here, n is the year serial number, x ij is the value of the NPP in the first year of the j cell, and x j is the average of the NPP j cell over 16 years, i.e., from 2000 to 2015. Similarly, y ij is the value of the first j cell of temperature or precipitation, and y j is the average of the first j cell of temperature or precipitation over the 16 years of 2000-2015. To check the validity of the model, a p-test was used, and the tendencies were classified into 3 categories: highly significant, significant, and no significant change.

Urban Built-Up Land Expansion Simulation.
It is of great significance to understand the spatial expansion trend of urban built-up land for guiding the rational expansion of urban land [46][47][48][49][50][51]. Night light data is a common kind of remote sensing data analyzing the city scale [52][53][54]. e average night light data constitute a cloudless composite map generated by the DN value for each time period, with a spatial resolution of 1 km. e night lighting data from DMSP-OLS provides a tool for monitoring urban sprawl from time and space perspectives [55,56]. e city area is determined by the threshold of the DN. If the DN of the region is greater than the threshold, the region is defined as a city. Different scenarios for the total urban area are obtained by changing the threshold of DN. When the simulated total urban areas are close to each other, the threshold of DN is noted in the urban land use census data and applied to the process of measuring urban expansion. e recorded nighttime light image data were used as an indicator of urban expansion.

Land Use Intensity.
In recent years, Zhengzhou has experienced rapid urbanization, with a significant growth in economic development, leading to an increase in population and resource consumption. e land use intensity has changed dramatically. In this study, seven indicators were selected to measure the change in land use intensity: urban population, GDP, industrial output, agricultural output, fixed asset input, quantity of shipments, and electricity consumption.

Changes in Land Use Types.
To determine the land use change in Zhengzhou during the period of 2000-2015, the land use transfer matrix for the period was obtained (Table 1). Table 1 shows that the transfer of cropland to built-up land is the main form of land use transfer; the area is approximately 367.51 km 2 , accounting for approximately 51.08% of the total land use transfer area. Water area is the second largest type of land occupied by cropland in addition  Figure 2(a) shows that cropland was the main land use type in Zhengzhou and was spread throughout the region. e builtup land was concentrated in the central and northern regions in the form of small patches scattered in some parts. Ecological land was mainly distributed in the western part of the region in the form of large patches, and the water area was mainly distributed at the borders of the city in the northern area in the form of a ribbon. According to the land transfer situation (Figure 2(b)), eight land use types were selected in the order from highest to lowest use to show the spatial distribution of land use transfer; the total area of the eight land transfer types accounted for 92.07% of the total area of the land transfer. Among them, the transfer from cropland to built-up land was mainly distributed in the central urban area, as a result of urban land expansion. e transfer from cropland to water area was mainly distributed around the land along the waters. Figure 3(a) shows that the NPP in general was distributed in areas other than the builtup land and water area, showing a generally increasing trend. Among them, the low value of NPP was mainly distributed in the eastern, southwestern, and southern border areas of Zhengzhou City in the form of small patches, and the areas with a high NPP value were mainly distributed throughout the city in the form of large patches. As shown in Figure 3  e negative values, which represent a negative correlation between precipitation and NPP values, were primarily observed throughout the region in the form of small patches.

Change in Land Use Intensity and Urban Built-Up Area.
Land use intensity can reflect the rates of energy emissions and socioeconomic development to a certain extent, and it has been documented that socioeconomic development indicators can be a good indicator of land use intensity [57,58]. We selected seven main indicators to reflect the land use intensity of Zhengzhou City and calculated the ratio between typical years to analyze the changing trends (Table 2). According to Table 2, the total population increased steadily from 2000 to 2015, but the growth amount was greater, and the growth rate showed a steady downward trend. e increase in GDP was most pronounced, increasing from 73

Discussion
e rapid expansion of urban areas has significantly affected regional ecosystems, making it extremely important to quantify the impact of urban expansion and climate change on NPP. Using land use data, night light data, NPP data, climate data, and a series of socioeconomic data, we explored the expansion of urban built-up land in Zhengzhou City during the period of 2000-2015 and the response of NPP to urban expansion and climate change. e study provides a reference for land managers to formulate land policies towards low carbon and sustainable development.
Built-up land expansion is the main form of land use change in Zhengzhou and is quite common in China, as the country is undergoing rapid urbanization. However, the land use changes in China and those in Europe [59,60], the United States [61], Australia [62], and other developed countries that have completed the process of urbanization are different. As Zhengzhou is located in the region with a general population and economic level, cropland and ecological land are the main land use types, and land use transformation is more characteristic, compared with the northwest, inland areas, and other underdeveloped areas of China. Cropland being occupied by built-up land is the main change in land use types that occurs in Zhengzhou, which is mainly due to the social and economic development. is result is consistent with previous studies on land use change [63,64]. From 2000 to 2015, the population increased to 9.569 million, and the urbanization rate in Zhengzhou increased from 55.1% to 69.7%. To accommodate more urban Complexity 7 residents and under the influence of the growing real estate market [65], urban areas expanded rapidly during the study period. Although the population in rural areas has been significantly reduced, idle rural settlements are widely distributed, and land consolidation can take a long time. As cropland accounts for more than 80% of the area in Zhengzhou and is distributed in various places, the expansion of urban land requires the occupation of a large area of cropland. e spatial distribution of land transformation, shown in Figure 2, proves our conclusion that socioeconomic development is the main driving force of change in land use type. e economic level of Zhengzhou city shows a characteristic growth from center to periphery, and the land use change shows the same trend. In other words, the land circulation in the developed areas of the urban economy is more intensive. Zhengzhou has experienced rapid urbanization in recent years, not only bringing about a growth in the economy and mass population, capital, and technology but also leading to a rapid increase in land use intensity. In this study, to represent land use intensity, we selected seven indicators, which showed a strong correlation between land use intensity and economic development level. e land use intensity gradually enhanced with the growing economy, which can be attributed to the regional differences in economic and natural resources between different periods. e differences in the economy and resources are consistent with the study conducted by Yang et al. [66], which showed that urban land use intensity increases with the increase in the level of urbanization. According to our research, the change in land use intensity has a significant effect on NPP. e areas with high land use intensity usually exhibit high levels of economic development and lower NPP values because such areas always have a large proportion of artificial vegetation and land, which can also explain the lower NPP values. In addition, areas with more natural and seminatural land use tend to have lower land use intensity [67], where cropland, forests, and grasslands account for a large proportion. With ecological land being increasingly occupied, policies and measures to promote intensive land use should be implemented to adjust land use intensity and to match the socioeconomic situation with the local condition [68]. Besides socioeconomic conditions, climate change [69], ecological conditions, and crop structures affect land use intensity to some extent.
Second, to correct and integrate the two types of night light data from different sources, we performed an exponential regression between MODIS-OLS night light data for 2013 and NPP-VIIRS night light data for 2014 and 2015 after noise processing and then obtained the revised NPP-VIIRS night light data. e reliability of this method has been verified [39,70]. In spatial simulations of energy-related carbon emissions, the night light data obtained from DMSP-OLS and NPP-VIIRS are suitable for simulating urban land  Compared with previous studies, the NPP simulation performed in this study has several advantages. e NPP estimates from MODIS have been validated to be consistent with the NPP values observed in the field [44,71]. Exist studies show that, for urban system, the MODIS NPP still has a good application [72,73]. As a result, the NPP simulations in previous studies have been validated, although more field observations are required for further research to improve the NPP simulation models for China (particularly in the west, where the density of field observations is low). e effects of global warming were widely demonstrated in China during the period of 2000-2015, with regions experiencing a rise in temperature accounting for 52.99% of that for the whole country. Although an increase in temperature can promote NPP to a higher extent than an increase in soil respiration, higher temperatures can also contribute to steaming and drought, leading to low vegetation productivity [74], particularly in an environment with insufficient water supply. In addition, a continuous increase in temperature will lead to increased soil respiration [75,76] and decrease the NPP. Continued global warming will ultimately damage carbon sequestration in terrestrial ecosystems. erefore, in China, reducing carbon emissions is urgently required for green, low-carbon development. Zhengzhou, which is an emerging and fast-growing city, should respond positively to the country's call for low-carbon green development. Moderate precipitation conditions are essential for vegetation growth; if the precipitation is too high or too low, vegetation growth will be affected, thus reducing NPP. For example, rainfall can increase cloud cover and thus reduce solar radiation, which is critical for vegetation growth [77]. Further water supply may create an aerobic environment in the root area, reduce soil nutrients [78], and inhibit vegetation growth. With the increase in temperature and a decrease in precipitation, ecological pressure will increase under primitive fragile environmental conditions [79,80]. However, rising temperatures may also cause other ecological problems, such as melting glaciers, extreme climate, and disease, which require enhanced ecological protection. Owing to extreme weather conditions, the precipitation in Zhengzhou city has been insufficient in recent years, and water resources are relatively scarce. Precipitation is a key factor for the growth of vegetation. e government should strengthen ecological protection.
For the effect of urban built-up land expansion on NPP, the impact of urban built-up land expansion on NPP may be positive or negative, depending on socioeconomic and biophysical factors [81,82]. e results in this study show there is negative relationship between urban built-up land and NPP at edge of the urban expansion areas, which means that the urban built-up expansion can damage the vegetation productivity in some extent. is is because with the continuous expansion of built-up land, the composition and quantity of vegetation decreased, then the net primary productivity of vegetation will decrease, which is consistent with the existing study [83]. While on the other areas, the relationships between urban built-up land and NPP are positive on the whole. e negative impact of urban expansion on NPP will disturb carbon balance to some extent. In addition, in Zhengzhou, human activities, such as desertification, loss of agricultural use, and deforestation, hamper green development; most of these activities may also disrupt the carbon balance. erefore, government departments should take effective environmental protection measures to strictly prohibit such activities.

Conclusions
Unlike previous studies, this study first simulated urban expansion from night light data by integrating the night light data obtained from DMSP-OLS and NPP-VIIRS and explored the response of NPP to urban expansion and climate change. is study can serve as a reference for urban green development.
We found that cropland is the main land use type in Zhengzhou City, and from 2000 to 2015, the land transfer of cropland to built-up land was the main pattern of land use change, with a total area of 367.51 km 2 being converted; this was common in most cities of China. Areas other than the built-up land and water area exhibited a generally increasing NPP trend; these areas were mainly distributed in the southern and northwest regions of Zhengzhou. Areas around the development of built-up land exhibited a downward trending NPP. Both precipitation and temperature had obvious effects on NPP. e average correlation coefficients between temperature and NPP and precipitation and NPP were 0.267 and 0.020, respectively, indicating that an increase in temperature and precipitation can promote NPP, despite significant spatial differences. Land use intensity gradually increased with economic growth. In terms of urban expansion, Zhengzhou expanded into the central city, and urban built-up expansion was mainly distributed in the central and northern parts of the city. From 2000 to 2015, most expansion areas exhibited an increasing NPP trend, indicating that the influence of urban expansion on NPP is mainly characterized by the evident influence of the expansion area.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request. Disclosure e funding sources had no role in the study design, data collection, analysis or interpretation, or the writing of this manuscript.

Conflicts of Interest
e authors declare that they have no conflicts of interest.

Authors' Contributions
Pengyan Zhang and Wenlong Jing designed and carried out the study. Yanyan Li, Dan Yang, and Yu Zhang participated Complexity in the analysis and presentation of analytic results. Ying Liu and Wenliang Geng collected and analyzed data. Tianqi Rong, Jiaxin Yang, and Jingwen Shao contributed the data used, and Mingzhou Qin revised the paper.