Research Articles

Carbon sink response of terrestrial vegetation ecosystems in the Yangtze River Delta and its driving mechanism

  • ZHAO Haixia , 1 ,
  • FAN Jinding 1, 2 ,
  • GU Binjie 1, 3 ,
  • CHEN Yijiang 4
  • 1. Nanjing Institute of Geography and Limnology, CAS, Nanjing 210008, China
  • 2. School of Environmental Science and Engineering, Suzhou University of Science and Technology, Suzhou 215009, Jiangsu, China
  • 3. Nanjing College, University of Chinese Academy of Sciences, Nanjing 211135, China
  • 4. School of Agricultural and Food Science, The University of Queensland, Queensland 4072, Australia
*Zhao Haixia (1976-), PhD and Associate Professor, specialized in environmental economics and ecological economics. E-mail:

Received date: 2023-03-13

  Accepted date: 2023-07-28

  Online published: 2024-01-08

Supported by

National Key R&D Program of China(2018YFD1100101)


The carbon cycle of terrestrial ecosystems is influenced by global climate change and human activities. Using remote sensing data and land cover products, the spatio-temporal variation characteristics and trends of NEP in the Yangtze River Delta from 2000 to 2020 were analyzed based on the soil respiration model. The driving influences of ecosystem structure evolution, temperature, rainfall, and human activities on NEP were studied. The results show that the NEP shows an overall distribution pattern of high in the southeast and low in the northwest. The area of carbon sinks is larger than that of the carbon sources. NEP spatial heterogeneity is significant. NEP change trend is basically unchanged or significantly better. The future change trend in most areas will be continuous decrease. Compared with temperature, NEP are more sensitive to precipitation. The positive influence of human activities on NEP is mainly observed in north-central Anhui and northern Jiangsu coastal areas, while the negative influence is mainly found in highly urbanized areas. In the process of ecosystem structure, the contribution of unchanged areas to NEP change is greater than that of changed areas.

Cite this article

ZHAO Haixia , FAN Jinding , GU Binjie , CHEN Yijiang . Carbon sink response of terrestrial vegetation ecosystems in the Yangtze River Delta and its driving mechanism[J]. Journal of Geographical Sciences, 2024 , 34(1) : 112 -130 . DOI: 10.1007/s11442-024-2197-z

1 Introduction

Climate change and its impacts have become the most serious environmental issue and one of the world’s most complex challenges today (Liu et al., 2019; Rogelj et al., 2019). According to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC), global surface temperatures increased by 1.09°C between 1850-1900 and 2011-2020. International consensus exists to address climate change and promote global carbon emissions reductions. Up to now, more than 130 countries and regions worldwide have proposed “carbon neutrality” goals or visions (Wei et al., 2022). China has also committed to carbon peaking by 2030 and carbon neutrality by 2060. Achieving net-zero carbon emissions and reaching the Paris Agreement control targets become the new focus of the global response to climate change. Net-zero carbon emissions refer to the balance between anthropogenic emissions of greenhouse gases into the atmosphere and anthropogenic removals of greenhouse gases from the atmosphere (Fankhauser et al., 2022). Carbon sinks can reduce carbon dioxide in the atmosphere and offset carbon emissions, achieving carbon neutrality goals (Duffy et al., 2021). Therefore, it is imperative to estimate terrestrial ecosystem carbon sink capacity.
Carbon sinks play a vital role in mitigating climate change (Piao et al., 2005). Scholars commonly use remote sensing methods to define indicators of vegetation carbon sink capacity (Cao et al., 2023). These indicators include gross primary productivity (GPP), net primary productivity (NPP), net ecosystem productivity (NEP), and net biome productivity (NBP) (Liu et al., 2021). NEP is a physical quantity that characterizes the net carbon exchange between terrestrial ecosystems and the atmosphere. Accurate NEP monitoring provides direct data support for carbon neutrality (Ye and Chuai, 2022). In recent years, carbon sink research has expanded from the initial study of carbon sinks in a single forest ecosystem to that of global terrestrial ecosystems (Evans et al., 2014; Ahlstrom et al., 2015; Cui et al., 2021), and some scholars have also focused on marine ecosystem carbon sink potential (Tokoro et al., 2014; Zhang et al., 2022). In addition, carbon sink studies in small regions have been expanded to large regions or even globally (Keenan et al., 2016). To sum up, carbon emissions tend to be higher in economically developed regions due to industrial concentration. Carbon sink research in developed regions should be given more attention. It is the basis for balancing regional development and ecological conservation.
Carbon sinks are affected by many factors. First, climate change affects carbon sinks. Climate change has become one of the most critical factors in altering the carbon cycle (Pan et al., 2020). As a result, warming will likely shorten carbon turnover in terrestrial ecosystems and increase uncertainty in carbon sequestration estimates (Carvalhais et al., 2014). In the tropics, rising temperatures and reduced precipitation lead to forest death and elevated carbon emissions (Mitchard, 2018). In the middle and high latitudes, higher temperatures can lengthen vegetation’s growing season and increase the carbon sink (Post et al., 2018). Second, land use change has been linked to ongoing land degradation due to long-term irrational land development, resulting in carbon loss and climate change (Li et al., 2021). For example, the change in natural factors will affect the carbon balance of cultivated land in different regions, while man-made management measures such as irrigation, fertilization and agricultural mechanization account for the main factors contributing to the change in the carbon cycle of cultivated land ecosystems (Liu et al., 2016; Liu et al., 2022). At the same time, the influence of construction land proportion on the regional carbon sink increased year by year (Xu et al., 2018). Finally, as human activities increasingly affect ecosystems, carbon sinks are increasingly closely related to human activities. Strengthening environmental regulation, promoting economic growth, increasing the share of tertiary industry, promoting technological progress, and strengthening the construction of ecological restoration projects are all likely to promote carbon sinks and sink efficiency (Huang et al., 2022; Zhang and Deng, 2022). Human activity is intense in economically developed areas. How temperature, precipitation and human activities and ecosystem structure affect carbon sink changes in these regions still needs further exploration.
In summary, this study characterized the carbon sink of terrestrial ecosystems by NEP. It also investigated the carbon sink pattern and its influencing factors in economically developed regions. The Yangtze River Delta region (YRD) is one of China’s most economically developed regions and one of China’s windows to the outside world. With less than 4% of its land area, it generates nearly a quarter of China’s total economic output and 1/3 of its total imports and exports. In 2019, carbon emissions accounted for 13.6% of China’s emissions, making it the primary source region for carbon emissions. The YRD bears the burden of balancing development and conservation. The study of carbon sinks in the YRD contributes to carbon neutrality in economically developed regions.

2 Methods and materials

2.1 Study area

The Yangtze River Delta (YRD) is one of the regions with the most dynamic economic development, the highest degree of openness, and the strongest innovation capacity in China. It covers 41 cities in Shanghai municipality, Jiangsu, Zhejiang and Anhui provinces, covering 358,000 km2 (Figure 1). The YRD accounts for about a quarter of China’s total economic output. Its total labor productivity ranks among the highest in China. Permanent residents urbanize at 60%. Due to the concentration of industry and population, carbon emissions are high, so the goal of carbon neutrality in the region is huge.
Figure 1 Location of the Yangtze River Delta region

2.2 Ecosystem carbon sinks measurement

Net ecosystem productivity (NEP) is usually defined as the difference between the net primary productivity (NPP) within an ecosystem and the carbon emissions from heterotrophic respiration (RH) (Tang et al., 2016). It represents the rate of carbon exchange between terrestrial and atmospheric ecosystems, which is an important indicator for regional carbon balance estimation and is often used as a measure of carbon sinks (Zhang et al., 2014; Yu et al., 2014; Dai et al., 2016; Pathak et al., 2018). Its formula is as follows:
N E P = N P P R H
R H = 0.22 × E x p 0.0913 T + L n 0.3145 R + 1 × 30 × 45.6 %
where NEP indicates net ecosystem productivity of vegetation (gC∙m-2∙a-1), NPP indicates net primary productivity of vegetation (gC∙m-2∙a-1), and RH indicates heterotrophic respiration (gC∙m-2∙a-1), calculated from the regression equation of temperature, precipitation and carbon emissions established by Pei et al. (2009). T is the monthly mean temperature (℃) and R is the monthly total precipitation (mm). If NEP > 0, it indicates that the carbon fixed by vegetation is higher than that emitted by soil and behaves as a carbon sink; conversely, if NEP < 0, it acts as a carbon source (Yu et al., 2014).

2.3 Trends in ecosystem carbon sinks

2.3.1 Theil-Sen Median trend analysis

The Theil-Sen Median trend analysis method is a robust trend calculation method for nonparametric statistics that can reduce or avoid the effects of missing data and anomalies on statistical results (Liu et al., 2010). This analysis is commonly used to study trends in long-time series data, which can scientifically and intuitively reflect the trends in time series data over time (Gao et al., 2021). Its formula is as follows:
β = M e d i a n x j x i j i , i < j
where β is the trend of NEP, i and j are time series, xi and xj represent the NEP values at time i and j, respectively. β>0 reflects an increasing trend of NEP and, conversely, a decreasing trend.

2.3.2 Mann-Kendall significance testing

The Mann-Kendall significance testing is a nonparametric statistical test that is simple and convenient to calculate and is often used to analyze and test the trend of a time series (Gao et al., 2021). Its formula is as follows:
Z = S 1 var S , S > 0 0 , S = 0 S 1 var S , S < 0
S = i = 1 n 1 j = j + 1 n sgn x j x i
sgn θ = 1 , θ > 0 0 , θ = 0 1 , θ < 0
var S = n n 1 2 n + 5 18
; n indicates the length of the time series, sgn is the sign of the function, and the value range of the statistic Z is (-∞,+∞). The original hypothesis is set to be that the series has no trend, and the bilateral trend test is used. At a given significance level α, when
Z > U 1 α / 2
, the original hypothesis is rejected, which means the series trend is significant, and the opposite implies that the trend is not significant.

2.3.3 Future trend analysis

Hurst exponent can be used to predict the future trend of long-term data relative to the past and is a standard method to describe the persistence of time series trends (Jiapaer et al., 2015). There are more methods to estimate the Hurst exponent, and in this paper, we use the commonly used R / S analysis method, which has more reliable estimation results (Jiapaer et al., 2015; Jiang et al., 2015). Its formula is as follows:
For the time series, NEPi, i = 1, 2, 3,..., n, define the time series:
(1) Define the sequence of the time series:
N E P ( τ ) ¯ = 1 τ 1 τ N E P ( τ ) τ = 1 , 2 , . . . , n
(2) Calculate the accumulated deviation:
X τ = t = 1 τ N E P τ N E P τ ¯ 1 t τ
(3) Create the range sequence:
R ( τ ) = max 1 t τ X ( τ ) min 1 t τ X τ τ = 1 , 2 , . . . , n
(4) Create the standard deviation sequence:
S ( τ ) = 1 τ t = 1 τ N E P t N E P ( τ ) 2 τ = 1 , 2 , . . . , n
For the ratio
R τ / S τ R / S
, if there exists the following relationship
R / S τ H
, the time series exists Hurst phenomenon. H is the Hurst exponent, which can be obtained by log(R / S)n = a + H × log(n) using the least squares method.

2.4 Influencing factors

2.4.1 Climate change

Partial correlation analysis was used to measure the impact of climate change on NEP in the YRD. Partial correlation analysis characterizes the correlation between two variables among multiple variables by controlling for other variables, and compared with the correlation coefficient, the partial correlation coefficient eliminates the influence of other variables, and the results are more reliable, which is widely used in many studies (A et al., 2016; Qin et al., 2021). Its formula is as follows:
R x , y , z = R x y R x z × R y z 1 R x z 2 × 1 R y z 2
where Rx,y,z is the partial correlation coefficients of the control variables x, y, and z, which are the correlation coefficients of variables x and y, variables x and z, and variables y and z, respectively, and the calculated results are subjected to the t-significance test.

2.4.2 Human activities

The residue analysis method is used to distinguish the effects of climate change and anthropogenic factors on ecosystem carbon sinks (Dong et al., 2020). In this paper, we establish the relationship between NEP and temperature and rainfall on a pixel-by-pixel basis based on multiple regression analysis and obtain the predicted NEP value (NEPpre) on each pixel-by-pixel, and the difference between NEPpre and the actual value of remote sensing observation is used to indicate the impact of human activities on NEP, and achieve the quantitative separation of climate change and the impact of human activities on NEP. Its formula is as follows:
N E P r e s = N E P r a w N E P p r e
N E P p r e = a × T + b × R + ε
where NEPpre and NEPraw are the predicted NEP values of the multiple regression model and the original NEP values estimated based on remote sensing images, respectively; a, b and ε are the parameters of the multiple regression model; T and R are the interpolated annual average temperature and rainfall, respectively; and NEPres is the residual.

2.4.3 Ecosystem structure

The ecosystem structure will have an impact on the regional ecosystem carbon sink during its evolution, as expressed by the fact that each changing ecosystem type will cause a decrease or increase in the carbon sink. Its formula is as follows:
E L , k = E L , t + 1 E L , t A k T A
where EL,k is the contribution of change in k ecosystem types; EL,t+1 and EL,t are the NEP corresponding to a particular ecosystem type at the beginning and end of change, respectively; Ak is the area of change in k ecosystem types; T is the study period; and A is the total area of the study area.

2.5 Data source and pre-processing

The data involved in the paper are mainly land use, NPP, temperature, and rainfall in the YRD. These data sources and preprocessing methods are shown in Table 1. Specifically, the land use data are obtained from the GlobleLand30 ( with a resolution of 30 m. The land use types are classified into eight categories, including cultivated land, forest, shrubland, grassland, wetland, water body, developed land, and bare land, according to the actual regional and research needs. The DEM data were obtained from the Shuttle Radar Topography Mission (SRTM) with a spatial resolution of 30 m. The net primary productivity (NPP) data for 2000-2020 were obtained from the MOD17A3HGF V6 product on the NASA website, with a spatial resolution of 500 m. Annual NPP data are pre-processed by the GEE (Google Earth Engine) cloud platform. 2000-2020 month- by-month rainfall data are obtained from the National Earth System Science Data Center (, global 0.5° climate data published by CRU, and global high- resolution climate data released by WorldClim are generated by the Delta spatial downscaling in China, and validated with data from 496 meteorological observation points, with credible validation results and a spatial resolution of 1 km. 2000-2020 month-by-month mean temperature data are obtained from GPRChinaTemp1km, generated by the Gaussian process regression based on meteorological station data with a spatial resolution of 1 km. The meteorological data were resampled to 500 m using the nearest neighbor interpolation method.
Table 1 Data source specific information and pre-processing
Data Sources Resolution Pre-processing
Land use GlobleLand30 ( 30 m Classified into eight categories, including cultivated land, forest, shrubland, grassland, wetland, water body, developed land, and bare land
DEM Shuttle Radar Topography Mission (SRTM) 30 m /
NPP MOD17A3HGF V6 product on the NASA website 500 m GEE platform pre-processing
gical data
Rainfall National Earth System Science Data Center ( 1 km Resampling to 500 m resolution
Temperature GPRChinaTemp1km 1 km

3 Results

3.1 Characteristics and patterns of carbon sinks pattern evolution

3.1.1 Overall characteristics

From 2000 to 2020, the overall NEP of the vegetation in the YRD showed a fluctuating upward trend (Figure 2), and the area of carbon sinks (NEP>0) accounted for a relatively large proportion. The average annual rate of change of the mean NEP was 3.89 gC∙m-2∙a-1, with the minimum and maximum annual NEP values occurring in 2000 and 2014, respectively. From 2000 to 2020, the average NEP value in the YRD ranged from 175.14 to 294.38 gC∙m-2∙a-1, with a multi-year mean value of 227.03 gC∙m-2∙a-1, which was at a high level. The overall carbon cycle of regional vegetation was dominated by carbon sinks, with an average annual increase of 71.34 Tg of total carbon sinks and a cumulative increase of 14.98×108 tC. The level of carbon sink in the past ten years is significantly higher than that of 10 years ago (1 TgC=1012 gC=106 tC).
Figure 2 Time series variation of annual mean NEP values in the Yangtze River Delta

3.1.2 Spatio-temporal distribution characteristics

The NEP of the YRD from 2000 to 2020 has apparent spatial heterogeneity (Figure 3a), and its change is generally decreasing from southeast to northwest, similar to the regional elevation change, with regional multi-year NEP mean values ranging from -278.92 to 1038.40 gC∙m-2∙a-1. The regions with significant NEP increase (NEP>120) from 2000 to 2010 are mainly observed in northern Jiangsu and Anhui. The decreasing regions are mainly in southern Jiangsu, southern Anhui, and most of Zhejiang and Shanghai. During this period, the rapid development of areas south of the Yangtze River, the expansion of urban construction land, and the large number of resources taken from natural ecosystems led to the reduction of regional vegetation cover and the weakening of ecosystem carbon sinks, while north of the Yangtze River, due to slower development and higher overall ecosystem integrity, the carbon sink capacity continued to increase. During 2010-2020, although the center of gravity of development is still south of the Yangtze River, the improvement of the economic level has prompted the strengthening of ecological civilization awareness and a significant increase in NEP in Zhejiang as well as the southern regions of Jiangsu and Anhui; while the carbon sink of ecosystems in the northern regions of the Yangtze River has declined (Figures 3b and 3c). However, in general, from 2000 to 2020, the NEP of the YRD increased significantly, and the ecosystem carbon sequestration capacity increased significantly, except for the NEP of the Shanghai urban area and Hangzhou Bay city cluster, which decreased significantly (Figure 3d).
Figure 3 Spatial distribution of NEP changes and multi-year mean values in the Yangtze River Delta from 2000 to 2020

3.1.3 Spatio-temporal trend analysis

Theil-Sen slope analysis shows that the vast majority of regions in the YRD show an increasing trend in NEP, and a limited number of regions are in a decreasing trend from 2000 to 2020. Further combined with the Mann-Kendall significance test (Figure 4), it is found that the area of the YRD with a significantly decrease and a moderately decrease in NEP accounts for a relatively small proportion, 2.85%, and 1.48%, respectively, and is mainly distributed along the Yangtze River and its estuary, around Jiangsu, Shanghai, and Hangzhou Bay. Only 0.84% of the area decreased, accounting for the smallest percentage. The area of regions that remained basically unchanged made up the largest proportion, 50.02%. This was widely distributed in Zhejiang, northwestern Anhui, and the coastal region of southeastern Jiangsu. The area of increased and moderately increased regions accounted for 8.49% and 15.47%, respectively, with a sporadic distribution. The area shares of significantly increased regions are higher, 20.85%, mainly in central Anhui, western Jiangsu, and northeastern coastal areas, and a small amount in central Zhejiang.
Figure 4 NEP trends in the Yangtze River Delta from 2000 to 2020
Further statistics on the trends of NEP changes in the provincial districts in the study area (Table 2) show that there are also significant differences in the trends of NEP changes within different regions of the YRD. Among them, Shanghai has a larger population and the most intense socio-economic activities, and the area of NEP decreased by 22.97%, which is significantly higher than that of other provinces. The area of NEP increased by 22.63%, which is significantly lower than that of other provinces. The development level of Jiangsu and Zhejiang is relatively similar, and the percentage of NEP decreasing areas is roughly the same, both around 7%. Zhejiang has better natural endowment and better protection of mountains and vegetation under the promotion of the concept of ecological civilization and the policy f closing mountains for forestry. The area of NEP basically unchanged has the highest percentage of 65.82%, while Jiangsu is mainly a plain with scattered and frequent human activities, and the area of NEP basically unchanged has the smallest percentage of 38.56%, but NEP with the construction of ecological civilization is increasing regional trend area accounted for a comparatively large 45.70%. Anhui has relatively weak socio-economic activities and a low impact of human activities in the YRD, with the smallest percentage of NEP reduction area of 1.84%.
Table 2 Area share of different NEP trends
Slope β Z NEP trends Proportion of change area (%)
Shanghai Jiangsu Zhejiang Auhui
β>0 2.58<Z Significant decrease 11.40 4.24 4.00 0.91
1.96<Z≤2.58 Moderate decrease 7.25 2.02 1.86 0.64
1.65<Z≤1.96 Decrease 4.32 1.10 1.20 0.29
β=0 Z≤1.65 Not significant 54.40 38.56 65.82 45.91
β<0 1.65<Z≤1.96 Increase 4.32 8.37 5.94 10.81
1.96<Z≤2.58 Moderate increase 7.43 18.63 9.28 18.47
2.58<Z Significant increase 10.88 27.07 11.89 22.96

3.2 Prediction of future trends in ecosystem carbon sinks

From the spatial distribution of the Hurst exponent, the Hurst exponent of the whole region ranged from 0.10 to 0.96, with only 26% of the area larger than 0.5 and 74% smaller than 0.5, indicating that the changes of NEP in the YRD have a strong anti-sustainability (Figure 5a).
Figure 5 Hurst Index and future trend changes of NEP in the Yangtze River Delta
To further explore the future trend of NEP change in the YRD, the Hurst exponent trend prediction method was combined with the Theil-Sen trend analysis method to classify the future trend of NEP change in the study area into four classes based on the H value and the slope of change β (Table 3). In the YRD, 69.69% of the regions showed a decreasing trend of future NEP, and 30.31% showed an increasing trend and the regional carbon sink level as a whole developed in a decreasing direction. However, there are differences in the future NEP trends in different regions, with 48.88% of the area in Shanghai showing an upward trend and 51.22% showing a downward trend, with little overall difference and a discrete distribution. The area of Jiangsu and Zhejiang with rising and falling NEP is relatively the same, both at 33% and 66%, with the falling area mainly concentrated in the northern part of the two provinces and the rising area more dispersed. The area share of NEP in Anhui was 25.46% and 75.54%, respectively, and the declining areas were mainly concentrated in the south and north. In contrast, the rising areas were concentrated in the central region and were more discrete (Figure 5b). Therefore, in the future, the regional development direction of Jiangsu, Zhejiang, and Anhui should be reasonably planned to improve the vegetation cover density in the declining areas and increase the regional carbon sink level.
Table 3 Area share of the future trend of NEP in the Yangtze River Delta
Slope β H index Future trends Proportion of change area (%)
Shanghai Jiangsu Zhejiang Auhui
β>0 H>0.5 Continuous increase 16.58 24.56 12.21 21.34
β<0 H>0.5 Increase-decrease 20.55 8.75 7.231 2.85
β>0 H<0.5 Continuous decrease 30.57 58.92 58.32 71.69
β<0 H<0.5 Decrease-increase 32.30 7.77 22.24 4.12

3.3 Driving mechanisms of the spatio-temporal patterns of carbon sinks changes

3.3.1 Impact of climatic conditions

Regional carbon sink level is comprehensively affected by various factors, and the hydrothermal condition is the key factor among them. The application of partial correlation analysis has found that the NEP and temperature partial correlation index of the YRD is between -0.83 and 0.84, and the rainfall partial correlation index is between -0.88 and 0.89. The spatial discrepancy is significant.
Some 45.64% of the regional area presents a positive correlation with the temperature, but only 1.12% show statistical significance (p<0.05). Most of them distribute in the west of Zhejiang and southern mountain regions. Results indicate that the rising temperature is conducive to the regional carbon sink enhancement (Figure 6a). 54.36% of the regional area presents a negative correlation with the temperature, which is mainly distributed in the middle plain area. 51.83% of the correlation data show statistical insignificance. The reasons are that rising temperature affects the plants’ enzyme activities during photosynthesis, transpiration, and nutrient transportation and thus inhibits growth in the plain area. This directly causes decreased fixed carbon content in plants, increased carbon emission of the soil respiration, and increased carbon source trending upward. 74.70% of the regional area displays a positive correlation between NEP and rainfall, and 10.93% shows statistical significance (p<0.05). Most of them are found in the Taihu lake basin, which indicates that rainfall is the critical factor of the regional NEP. The Taihu lake basin has a large water flow, a large amount of evapotranspiration, tense water circulation, and thus promoting increased rainfall, raising soil moisture content, enhancing plant growth, and strengthening plant productivity. Meanwhile, plant growth facilitates soil microbial activity, increasing soil carbon input and storage, which results in stronger carbon sinks. 25.30% of the regional area presents a negative correlation with rainfall, which is mainly distributed in the southern part of Zhejiang, southwestern forest areas of Anhui, etc. Some 23.76% of the correlation show statistical insignificance. In summary, rainfall significantly affects the YRD’s NEP variation, and its correlation is apparently stronger than temperature (Figure 6b).
Figure 6 NEP partial correlation with temperature and precipitation of the Yangtze River Delta

3.3.2 Impact of human activities

An ecosystem carbon sink is not only influenced by climate change but also disturbed by human activities as an important factor. Using the multiple regression residue analysis to quantify the effects of human activities on NEP variation in order to eliminate climate influence (Figure 7). Results show that the average residue from 2000 to 2020 is 0.53, and the average residue in the last three years has increased by 0.005, demonstrating that human activities mainly have positive effects on regional plants NEP. Positive impacts of human activities on ecosystem NEP have gradually increased due to the construction of the eco-civilization system and robust ecological protection and restoration force. From the perspective of spatial distribution, influential positive zones of human activities on NEP are mostly distributed in the middle of Anhui, the northern part, and the northern coastal areas of Jiangsu. Negative influential zones of human activities on NEP are mostly distributed in areas with intense human activities and rapid urbanization process along the Yangtze River, around the Shanghai Metropolitan Area, and Hangzhou Bay city cluster. Large areas of cultivated land, grassland, and forest ecosystem have been transformed into construction land, which causes the surface plants coverage rate to degrade and the ecosystem carbon sink function to weaken (Figure 8).
Figure 7 Effects of human activities on NEP in the Yangtze River Delta in 2000-2020
Figure 8 Ecosystem transformation matrix of the Yangtze River Delta in 2000-2020

3.3.3 Contribution of changes in ecosystem structure

First of all, using ecosystem type transformation and NEP variation data to conduct spatial overlay analysis. Then calculating different ecosystem structural changes’ contribution to NEP variations found the following results that ecosystem type and structural transformation affect regional NEP spatial pattern change through alteration of ecosystem integrity and health condition. In the recent 20 years, the contribution percentage of construction land transforming into a cultivated land ecosystem is 1.92%, and the contribution percentage of cultivated land ecosystem transforming into a forest ecosystem is 1.83% (Figure 8). Even though ecosystem type and structural transformation’s contribution level is low on NEP variation, this level still reflects the prompting role of increasing ecosystem carbon sink, which production and living land transform into ecological land. The stability of ecosystem structure plays a critical role in maintaining regional carbon sink levels. According to the results, cultivated land and forest ecosystem have the most significant contribution on NEP, with contribution percentages of 46.01% and 20.24%, respectively, which are 55.72% and 25.72% of the total NEP variations. These two ecosystems are also the major ecosystem types that are in charge of maintaining and increasing carbon sink levels in the Yangtze delta region (Table 4). Cultivated land and forest ecosystem are the predominant ecosystems in the Yangtze delta region. Cultivated land serves as product of the human activity; its land cover is the result of the human plantation. Its growing process is the process of ecosystem capturing and fixing carbon. Forests serve as the principal part of the natural carbon sink, and its photosynthesis, which growth and development require, is crucial for ecosystem carbon sink.
Table 4 Influential proportion of ecosystem-type transformation on NEP (%)
2000 2020
Cultivated land Forest Grassland Shrubland Wetland Water body Developed land Bareland
Cultivated land 55.72 1.83 0.43 0.01 0.04 0.40 1.12 0.00
Forest 1.51 25.72 0.71 0.04 0.01 0.14 0.04 0.00
Grassland 0.28 0.71 1.22 0.00 0.02 0.03 0.01 0.00
Shrubland 0.00 0.05 0.00 0.05 0.00 0.00 0.00 0.00
Wetland 0.06 0.00 0.01 0.00 0.03 0.02 0.00 0.00
Water body 0.84 0.05 0.03 0.00 0.08 1.10 0.05 0.00
Developed land 1.92 0.03 0.02 0.00 0.00 0.01 5.64 0.00
Bareland 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.01

4 Discussion

This paper uses remote sensing data from dense time series. The spatial and temporal patterns of NEP in the YRD and the effects of climate change, human activities and ecosystem structural changes on NEP were explored at the pixel scale. This study attempts to explain the carbon sink response of ecosystem structure. The research result is significant in terms of the YRD’s ecological protection and carbon neutrality goal.

4.1 Changes in carbon sink spatial and temporal patterns using NEP as an indicator

Carbon sinks are often characterized by NEP, which is the core data of this study, so it is imperative to evaluate its accuracy. There are currently two ways to evaluate precision. One way is field observation data-based precision evaluation. Another way is to compare previous simulation model results or study results approved by peer experts. The study area of this paper is large and rich in ecosystem types. Therefore, it is difficult to verify the actual measurement data, so the estimation results are compared with other research results. Zou et al. (2022) estimate the YRD’s NEP average to be 253.2 gC∙m-2∙a-1 between 2000 and 2019. This research estimates the multi-year mean of NEP to be 227.03 gC∙m-2∙a-1, which is considered a small disparity from the previous results. Guo and Fang (2021) estimate Jiangsu’s annual average NEP to be −373.37 to 1013.51 gC∙m-2∙a-1, which is close to this research. Therefore, this paper’s estimations of NEP numbers are reliable since there are only small gaps with other research findings.
Terrestrial ecosystems utilize photosynthesis to stabilize CO2. A large amount of CO2 is returned to the atmosphere through plants’ surface respiration, roots’ respiration, microbial heterotrophic respiration, etc. A small amount of CO2 is involved in the carbon cycle through burning, biological volatile organic compound emission, and soluble organic carbon dissolving into the river systems (Keller, 2019; Huang et al., 2023). Therefore, this research estimates terrestrial ecosystem productivity based on soil respiration and remote sensing data. However, the NPP estimation of the YRD’s terrestrial plants is based on the remote sensing method (top to bottom approach). Soil respiration effects are realized by using different sites’ soil respiration data to construct mathematical modelling and upscaling techniques (bottom to top approach), which have certain deviations (Jian et al., 2022). Furthermore, this research fails to estimate the carbon sink ability of two ecosystems: construction land and water bodies due to NPP data boundedness. Therefore, future research should focus on incorporating construction land and water bodies into regional-scale NEP estimation.

4.2 Factors influencing carbon sinks in terrestrial ecosystems

Changes in the temporal and spatial patterns of NEP in the YRD are the result of a combination of climate, human activities, and other factors. According to research findings, rainfall fluctuations are the main climate factor, consistent with other studies (Li et al., 2019; Zhang et al., 2019). Plants’ photosynthesis is not only affected by temperature and rainfall but also driven by sunlight, moisture, and other climate factors. There is a complex nonlinear relationship between NEP and weather factors. This research only considers two key climate factors, temperature and rainfall, and constructs a model to analyse terrestrial plant coverage variation. This research does not account for sunlight duration, soil moisture, and evaporation capacity, which could bias the results. Further research should consider multiple weather factors and provide a comprehensive evaluation. Secondly, the YRD has drastic differences in various regions’ terrain, landform, temperature, rainfall. This could cause spatial differences in various regions’ terrestrial plants’ growth status and response to climate factors. Furthermore, it can lead to different hysteretic characteristics of the relative data. In addition, the study found that NEP and terrain are relevant to space. Many existing studies pay more attention to temperature, rainfall, and human activities; hence, the following key points of research should revolve around terrain effects on NEP.
In general, human activity has a double-edged effect on NEP. On the one hand, human intensification may have an impact on ecosystem structure and cause a decline in carbon sink function. Ecological land such as forests may be converted into urban construction land, which is especially prevalent in economically developed regions. The leading urban agglomerations in the YRD all show NEP declines. On the other hand, human perceptions of the environment have improved. People are aware of the importance of the harmonious coexistence of man and nature, especially since the top-down construction of ecological civilization in China (Wang et al., 2020). Ecological protection and restoration projects have been carried out under policy guidance, ensuring the bottom line of ecological land and contributing to the restoration and enhancement of carbon sink functions. The areas within the YRD that prioritize ecological protection are often also areas with high NEP. This research utilizes residue analysis, which peels off the influence of climate on NEP, inferring the effects of human activity on NEP. However, this research does not quantify the double-edge effects of human activities, which should be a future research focus point.

4.3 Policy implications

Due to the large YRD area, there are obvious differences among various small regions. Therefore, different strategies for carbon sink enhancement should be proposed to carbon neutrality in response to the reality of different small regions. It is divided into urban construction zones and ecological protection zones. Urban construction zones are mainly located along the Yangtze River and coastal city clusters, supporting economic development. Ecological protection zones are areas with a good ecological foundation in the region, mainly in southern Zhejiang, southern Anhui and northern Jiangsu.
Urban construction areas are areas with high population concentration, industrial concentration and building density. These areas should aim to reduce carbon sources, but carbon sink capacity should not be neglected. On the one hand, these areas need to optimize their industrial structure and city layout, and improve technology to reduce carbon emissions. On the other hand, these regions can enhance their urban carbon sequestration capacity by building green areas such as urban green infrastructure (Zhao et al., 2023). In addition, the ecological assessment of construction projects needs to include a life cycle assessment for climate change adaptation (Salemdeeb et al., 2021).
Ecological protection zones are significant areas in the region that perform carbon sink functions. These areas need to optimize the layout of ecological construction and promote the optimization of the ecosystem’s internal structure. Ecological projects such as reforestation and afforestation have expanded the ecological land area in some small areas. In addition, they have enhanced the YRD’s carbon sink capacity. However, there is still a need to strengthen ecosystems’ ability of to respond to climate change and maintain ecosystem stability and integrity. Further, as forests mature, their photosynthetic carbon sequestration and respiratory carbon release are comparable, which results in a decrease in their carbon sequestration capacity (Liu et al., 2014; Cheng et al., 2023). Therefore, it is necessary to implement a set of periodic and reasonable measures for harvesting over-mature forests to ensure that the ecosystem’s carbon sequestration capacity is maximized in these regions.

5 Conclusions

(1) The YRD’s ecosystem NEP displayed an overall spatial pattern of high in the southeast and low in the northwest from 2000 to 2020. Its regional NEP generally fluctuates. The carbon sink region has a larger area than the carbon source region, and plants’ carbon sink ability keeps enhancing.
(2) From 2000 to 2020, the YRD’s regional NEP variation trend shows significant spatial heterogeneity. Areas with decreasing trends only occupy 5.17% of the total research area. The trend is generally constant, but mostly shifting upward since the carbon sink ability will increase as the ecosystem stabilize. Afforestation, closing land for reforestation, and other ecological projects will significantly improve the regional NEP.
(3) In terms of future variation trends, most forestation areas of the YRD will keep a decreasing trend in NEP. Regions where the Hurst index is smaller than 0.5 occupy 74% of the region area. Regions where the Hurst index is greater than 0.5 occupy only 26% of the area. The future area ratio of the variation trend rank from top to bottom is: persistent decline (63.69%) > persistent rise (19.46%) > decline to rise (10.85%) > increase to decline (6.00%).
(4) Climate change and human activity all have double-edged effects on the YRD’s regional ecosystem carbon sink. NEP change is more sensitive to rainfall than temperature. Increased rainfall contributes to local NEP increases. Human activity has a positive influence on NEP in the middle north of Anhui and northern Jiangsu’s coastal areas. Negative influential zones are concentrated along the Yangtze River coastal line, around the Shanghai Metropolitan Area, and Hangzhou Bay city cluster. During ecosystem structural evolution, the unchanged area contributes more to the research area’s NEP variation than the changed area; therefore, maintaining stabilization of ecosystem structure plays an apparent promoting role in the enhancement of regional ecosystem carbon sinks.
A D, Zhao W J, Qua X Y et al., 2016. Spatio-temporal variation of vegetation coverage and its response to climate change in North China Plain in the last 33 years. International Journal of Applied Earth Observation and Geoinformation, 53: 103-117.


Ahlstrom A, Raupach M R, Schurgers G et al., 2015. The dominant role of semi-arid ecosystems in the trend and variability of the land CO2 sink. Science, 348(6237): 895-899.


Cao S P, Yi H, Zhang L F et al., 2023. Spatiotemporal dynamics of vegetation net ecosystem productivity and its response to drought in Northwest China. GIScience Remote Sensing, 60(1): 2194597.


Carvalhais N, Forkel M, Khomik M et al., 2014. Global covariation of carbon turnover times with climate in terrestrial ecosystems. Nature, 514 (7521): 213-217.


Cheng F S, Tian J X, He J Y et al., 2023. The spatial and temporal distribution of China’s forest carbon. Frontier in Ecology and Evolution, 11: 1110594.

Cui Y P, Li N, Fu, Y et al., 2021. Carbon neutrality and mitigating contribution of terrestrial carbon sink on anthropogenic climate warming in China, the United States, Russia and Canada. Journal of Geographical Sciences, 31(7): 925-937.


Dai E F, Huang Y, Wu Z et al., 2016. Analysis of spatio-temporal features of a carbon source/sink and its relationship to climatic factors in the Inner Mongolia grassland ecosystem. Journal of Geographical Sciences, 26(3): 297-312.


Dong G, Zhao F Y, Chen J Q et al., 2020. Non-climatic component provoked substantial spatiotemporal changes of carbon and water use efficiency on the Mongolian Plateau. Environmental Research Letters, 15(9): 095009.


Duffy K A, Schwalm C R, Arcus V L et al., 2021. How close are we to the temperature tipping point of the terrestrial biosphere? Science Advances, 7(3): eaay1052.


Evans R D, Koyama A, Sonderegger D L et al., 2014. Greater ecosystem carbon in the Mojave Desert after ten years exposure to elevated CO2. Nature Climate Change, 4(5): 394-397.


Fankhauser S, Smith S M, Allen M et al., 2022. The meaning of net zero and how to get it right. Nature Climate Change, 12(1): 15-21.


Gao X, Huang X X, Lo K et al., 2021. Vegetation responses to climate change in the Qilian Mountain Nature Reserve, Northwest China. Global Ecology and Conservation, 28: e01698.


Guo X M, Fang C L, 2021. Integrated land use change related carbon source/sink examination in Jiangsu province. Land, 10: 1310.


Huang B B, Lu F, Wang X K et al., 2022. Ecological restoration and rising CO2 enhance the carbon sink, counteracting climate change in northeastern China. Environmental Research Letters, 17(1): 014002.


Huang Y T, Wang F, Zhang L J et al., 2023. Changes and net ecosystem productivity of terrestrial ecosystems and their influencing factors in China from 2000 to 2019. Frontiers in Plant Science, 14: 1120064.


Jian J S, Bailey V, Dorheim K et al., 2022. Historically inconsistent productivity and respiration fluxes in the global terrestrial carbon cycle. Nature Communication, 13(1): 1733.


Jiang W G, Yuan L H, Wang W J et al., 2015. Spatio-temporal analysis of vegetation variation in the Yellow River Basin. Ecological Indicators, 51: 117-126.


Jiapaer G, Liang S L, Yi Q X et al., 2015. Vegetation dynamics and responses to recent climate change in Xinjiang using leaf area index as an indicator. Ecological Indicators, 58: 64-76.


Kauppi P E, Mielikainen K, Kuusela K, 1992. Biomass and carbon budget of European forests, 1971 to 1990. Science, 256 (5053): 70-74.


Keenan T F, Prentice I C, Canadell J G et al., 2016. Recent pause in the growth rate of atmospheric CO2 due to enhanced terrestrial carbon uptake. Nature Communication, 7: 13428.


Keller C K, 2019. Carbon exports from terrestrial ecosystems: A critical-zone framework. Ecosystem, 22(8): 1691-1705.


Li J S, Guo X M, Chuai X W et al., 2021. Reexamine China’s terrestrial ecosystem carbon balance under land use-type and climate change. Land Use Policy, 102: 105275.


Liu J Y, Shao Q Q, Yan X D et al., 2016. The climatic impacts of land use and land cover change compared among countries. Journal of Geographical Sciences, 26(7): 889-903.


Liu Q Q, Wang S J, Zhang W Z et al., 2019. Examining the effects of income inequality on CO2 emissions: Evidence from non-spatial and spatial perspectives. Applied Energy, 236: 163-171.


Liu X C, Wang S, Zhuang Q L et al., 2022. A review on carbon source and sink in arable land ecosystems. Land, 11(4): 580.


Liu Y, Dang C Y, Yue H et al., 2021. Enhanced drought detection and monitoring using sun-induced chlorophyll fluorescence over Hulun Buir Grassland, China. Science of the Total Environment, 770: 145271.


Liu Y C, Yu G R, Wang Q F et al., 2014. Carbon carry capacity and carbon sequestration potential in China based on an integrated analysis of mature forest biomass. Science China-Life Sciences, 57(12): 1218-1229.


Liu Y L, Wang Q, Bi J Z et al., 2010. The analysis of NDVI trends in the coastal zone based on Mann-Kendall test: A case in the Jiaodong Peninsula. Acta Oceanologica Sinica, 32(3): 79-87.

Mitchard E T A, 2018. The tropical forest carbon cycle and climate change. Nature, 559(7715): 527-534.


Pan S F, Yang J, Tian H Q et al., 2020. Climate extreme versus carbon extreme: responses of terrestrial carbon fluxes to temperature and precipitation. Journal of Geophysical Research: Biogeosciences, 125: e2019JG005252.

Pathak K, Malhi Y, Sileshi G W et al., 2018. Net ecosystem productivity and carbon dynamics of the traditionally managed Imperata grasslands of North East India. Science of the Total Environment, 635: 1124-1131.


Pei Z Y, Ouyang H, Zhou C P et al., 2009. Carbon balance in an alpine steppe in the Qinghai-Tibet Plateau. Journal of Integrative Plant Biology, 51(5): 521-526.


Piao S L, Fang J Y, Zhu B et al., 2005. Forest biomass carbon stocks in China over the past 2 decades: Estimation based on integrated inventory and satellite data. Journal of Geophysical Research: Biogeosciences, 110(G1): G01006.

Post E, Steinman B A, Mann M E, 2018. Acceleration of phenological advance and warming with latitude over the past century. Scientific Reports, 8: 3927.


Qin Q T, Chen J J, Yang Y P et al., 2021. Spatiotemporal variations of vegetation and its response to topography and climate in the source region of the Yellow River. China Environmental Science, 41(8): 3832-3841. (in Chinese)

Rogelj J, Huppmann D, Krey V et al., 2019. A new scenario logic for the Paris Agreement long-term temperature goal. Nature, 573: 357-363.


Salemdeeb R, Saint R, Clark W et al., 2021. A pragmatic and industry-oriented framework for data quality assessment of environmental footprint tools. Resources, Environment and Sustainability, 3: 100019.

Tang X L, Fan S H, Qi L H et al., 2016. Soil respiration and net ecosystem production in relation to intensive management in Moso bamboo forests. Catena, 137: 219-228.


Tans P P, Fung I Y, Takahashi T, 1990. Observational constraints on the global atmospheric CO2 budget. Science, 247(4949): 1431-1438.


Tokoro T, Hosokawa S, Miyoshi E et al., 2014. Net uptake of atmospheric CO2 by coastal submerged aquatic vegetation. Global Change Biology, 20(6): 1873-1884.


Wang R, Qi R, Cheng J H et al., 2020. The behavior and cognition of ecological civilization among Chinese university students. Journal of Cleaner Production, 243: 118464.


Wei Y M, Chen K Y, Kang J N et al., 2022. Policy and management of carbon peaking and carbon neutrality: A literature review. Engineering, 14: 52-63.


Xu Q, Dong Y X, Yang R, 2018. Influence of different geographical factors on carbon sink functions in the Pearl River Delta. Scientific Reports, 7: 110.


Ye X, Chuai X W, 2022. Carbon sinks/sources? Spatiotemporal evolution in China and its response to built-up land expansion. Journal of Environmental Management, 321: 115863.


Yu G R, Chen Z, Piao S L et al., 2014. High carbon dioxide uptake by subtropical forest ecosystems in the East Asian monsoon region. Proceedings of the National Academy of Science of the United States of America, 111: 4910-4915.

Zhang A X, Deng R R, 2022. Spatial-temporal evolution and influencing factors of net carbon sink efficiency in Chinese cities under the background of carbon neutrality. Journal of Cleaner Production, 365: 132547.


Zhang L, Guo H D, Jia G S et al., 2014. Net ecosystem productivity of temperate grasslands in northern China: An upscaling study. Agricultural and Forest Meteorology, 184: 71-81.


Zhang L, Ren X L, Wang J B et al., 2019. Interannual variability of terrestrial net ecosystem productivity over China: Regional contributions and climate attribution. Environmental Research Letters, 14(1): 014003.


Zhang M, Cheng Y Y, Bao Y et al., 2022. Seasonal to decadal spatiotemporal variations of the global ocean carbon sink. Global Change Biology, 28(5): 1786-1797.


Zhao H X, Gu B J, Fan J D et al., 2023. Socioeconomic factors influence the spatial and temporal distribution of blue-green infrastructure demand: A case of Nanjing city. International Journal of Environmental Research and Public Health, 20(5): 3979.


Zou C, Li H, Chen D H et al., 2022. Spatial-temporal changes of carbon source/sink in terrestrial vegetation ecosystem and response to meteorological factors in Yangtze River Delta Region (China). Sustainability, 14: 10051.