The impact of climate change and human activities to vegetation carbon sequestration variation in Sichuan and Chongqing
气候变化和人类活动对川渝植被固碳变化的影响

2.1. Dataset

Normalized difference vegetation index (NDVI) was obtained from MOD13Q1 provided by NASA for 2010–2020 with 250 m horizontal spatial resolution and 16 d temporal resolution (https://modis.gsfc.nasa.gov). The NDVI was merged, cropped, and reprojected, and the monthly data in the study area were synthesized with the Maximum Value Composites (MVC).

The meteorological data from China Meteorological Data Network (https://data.cma.cn), including 192 ground-based meteorological stations in the study area (temperature and sunshine hours) and had been quality-controlled and checked. The temperature from the stations was interpolated using Anusplin to obtain the monthly average temperature dataset with the same resolution as NDVI. Due to the few radiation stations, the error generated in the interpolation process may be more significant; We obtained the solar radiation using an empirical formula (Chen et al., 2004; Glover and McCulloch, 1958), calculated from the sunshine hours as well as interpolated and resampled. Precipitation was obtained from the 1 km monthly precipitation dataset, provided by National Tibetan Plateau Data Center (http://data.tpdc.ac.cn/) (Ding and Peng, 2020; Peng et al., 2017, 2018, 2019; Peng, 2020).

Elevation from the Resource and Environmental Science and Data Center (www.resdc.cn), based on radar topographic mapping SRTM data from the U.S. Space Shuttle Endeavour. The land cover is GlobeLand30 global land cover data with 30 m horizontal spatial resolution, which is obtained from the National Geomatics Center of China Global Land Cover Data Product Service website (DOI: 10.11769) with high accuracy, the overall accuracy of V2010 is 83.50% and the Kappa coefficient is 0.78.

2.2. Model for the calculation of NPP

The CASA model based on light energy utilization theory was used to estimate the NPP. J. L. Monteith (1972) proposed that plant productivity can be calculated by multiplying the photosynthetic utilization efficiency with the effective solar radiation absorbed by the vegetation. NPP can be obtained from equation (1):(1)$NPP\left(x,t\right)=APAR\left(x,t\right)\times \epsilon \left(x,t\right)$where NPP (x,t) is the net primary productivity of vegetation (gC/m²) at time t in x grid points. APAR (MJ/m²/t) is the photosynthetically active radiation absorbed by the vegetation at time t, which mainly relies on the total solar radiation as well as the characteristics of the vegetation (Wang et al., 2017b). ε (gC/MJ) indicates the actual light energy utilization of the vegetation. Factors such as temperature, precipitation, and soil can influence productivity by affecting the photosynthesis of vegetation. APAR and ε can be calculated from equations (2), (3)):(2)$APAR\left(x,t\right)=FPAR\left(x,t\right)\times SOL\left(x,t\right)\times 0.5$(3)$\epsilon \left(x,t\right)=T_1\left(x,t\right)\times T_2\left(x,t\right)\times W\left(x,t\right)\times {\epsilon }_{\mathit{max}}$where FPAR is the proportion of incident photosynthetically active radiation absorbed by the vegetation; SOL (MJ/m²/t) is the total solar radiation at the time t; 0.5 is the coefficient, which is the proportion of effective solar radiation that can be used by vegetation. T is the temperature stress factor, which is the stress effect of high (T₁) and low (T₂) temperature, respectively. W is the water stress factor. ${\epsilon }_{\mathit{max}}$ is the maximum light energy utilization of the vegetation. Some studies have delved into the CASA model and provided more detailed computational details (Li et al., 2021; Potter et al., 1993; Zhu et al., 2007).
基于光能利用理论的CASA模型用于估算NPP。J.L.Monteith（1972）提出，植物生产力可以通过光合利用效率乘以植被吸收的有效太阳辐射来计算。NPP可以由等式（1）获得： (1)$NPP\left(x,t\right)=APAR\left(x,t\right)\times \epsilon \left(x,t\right)$ 其中NPP（x，t）是x个网格点中t时刻植被的净初级生产力（gC/m ² ）。APAR（MJ/m ² /t）是植被在时间t吸收的光合有效辐射，主要取决于太阳总辐射以及植被的特性（Wang等人，2017b）。ε（gC/MJ）表示植被的实际光能利用。温度、降水和土壤等因素可以通过影响植被的光合作用来影响生产力。APAR和ε可由等式(2)、(3)计算： (2)$APAR\left(x,t\right)=FPAR\left(x,t\right)\times SOL\left(x,t\right)\times 0.5$ (3)$\epsilon \left(x,t\right)=T_1\left(x,t\right)\times T_2\left(x,t\right)\times W\left(x,t\right)\times {\epsilon }_{\mathit{max}}$ 其中FPAR是植被吸收的入射光合有效辐射的比例；SOL（MJ/m ² /t）为t时刻的太阳辐射总量；0.5为系数，是可被植被利用的有效太阳辐射的比例。T为温度应力因子，分别为高温（T ₁ ）和低温（T ₂ ）的应力效应。W为水分胁迫因子。 ${\epsilon }_{\mathit{max}}$ 是植被的最大光能利用率。一些研究深入研究了CASA模型，并提供了更详细的计算细节（李等，2021；波特等人，1993；朱等，2007）。

The NPP was calculated with the CASA model in the Sichuan-Chongqing, and to further connect the NPP with the amount of carbon sequestration. Using the CASA model may not be very precise and the calculation process of NPP is more complicated (Li et al., 2021), while factors including the interpolation of meteorological data and the calculation of total solar radiation may also affect the NPP. Piao et al. (2001) estimated the NPP using the CASA model in China, and the findings indicated that the NPP could reach 400 gC/m²/a in the Chinese subtropics, which is compatible with the results of this study. Liu et al. (2013) estimated NPP in Sichuan based on the CASA model and found that their estimates of NPP were slightly lower compared to other studies, while the present study's estimates of NPP were more consistent with the results of other studies mentioned in Liu et al. (2013). The estimation of NPP in this study was also comparable with the results from Li et al. (2022).
用CASA模型计算了川渝地区的NPP，并进一步将NPP与固碳量联系起来。使用CASA模型可能不是很精确，NPP的计算过程更复杂（Li et al.，2021），而包括气象数据插值和太阳总辐射计算在内的因素也可能影响NPP。Piao等人。（2001）使用CASA模型估算了中国的NPP，结果表明，中国亚热带的NPP可达400 gC/m ² /a，这与本研究的结果一致。刘等人。（2013）基于CASA模型估计了四川的NPP，发现与其他研究相比，他们对NPP的估计略低，而本研究对NPP的估计与Liu等人提到的其他研究的结果更一致。（2013）.本研究中NPP的估计也与Li等人的结果相当。（2022）.

Nevertheless, the CASA model calculates the results for the actual NPP, considering not only meteorological elements but also human factors. To further investigate the relationship between meteorology and anthropogeny with vegetation carbon sequestration, separate calculations were needed for different impact factors. Therefore, the TM model was applied to calculate the NPP under the influence of meteorological factors (PNPP) (Lieth, 1975). Since the TM model only considered meteorological factors, the NPP caused by human activities can be obtained by calculating the difference between the CASA and TM model results. The TM model can be expressed as following equations (4), (5), (6)):(4)$L=300+25T+0.05T^3$(5)$V=\frac{1.05N}{\sqrt{1+{\left(1.05\frac{N}{L}\right)}^2}}$(6)$PNPP=3000\left[1-e^{-0.0009695\left(V-20\right)}\right]$where T represents the average annual temperature and the maximum annual evapotranspiration that can be calculated L (mm). N represents the total annual precipitation (mm), with the maximum evapotranspiration is used to calculate the average annual actual evapotranspiration V (mm), which is further calculated to obtain PNPP.
然而，CASA模型计算实际核电厂的结果，不仅考虑了气象因素，还考虑了人为因素。为了进一步研究气象和人为与植被固碳之间的关系，需要对不同的影响因子进行单独计算。因此，TM模型被应用于计算气象因素影响下的NPP（PNPP）（Lieth，1975）。由于TM模型只考虑了气象因素，人类活动引起的NPP可以通过计算CASA和TM模型结果之间的差异来获得。TM模型可以表示为以下等式（4）、（5）、（6））： (4)$L=300+25T+0.05T^3$ (5)$V=\frac{1.05N}{\sqrt{1+{\left(1.05\frac{N}{L}\right)}^2}}$ (6)$PNPP=3000\left[1-e^{-0.0009695\left(V-20\right)}\right]$ 其中T表示年平均温度和可计算的最大年蒸散量L（mm）。N代表年总降水量（mm），最大蒸散量用于计算年平均实际蒸散量V（mm），进一步计算得到PNPP。

Referring to Chen et al. (2014) and Teng et al. (2020) on the calculation of NPP under the influence of human activities (HNPP), which has been applied in multiple regions (Chen et al., 2019b), it can be expressed as:$HNPP=PNPP-NPP$
参考陈等人。（2014）和滕等人。（2020）关于人类活动影响下的NPP（HNPP）的计算，已在多个地区应用（Chen等人，2019b），可表示为： $HNPP=PNPP-NPP$

However, the calculation methodology separated human activities and climate change as two relatively independent drivers, ignoring the interactions between the two (Yin et al., 2020); in addition, short-term uncertainties and time lag effects were also inevitable for both (Teng et al., 2020). To further assess the bias between the two models, the national nature reserves in the western of Sichuan, which had lower population densities and relatively weaker human activities (relatively independent regions of the climate change and human activities), were selected to discuss the deviations and stability between the two models (More details can be found in SI).
然而，计算方法将人类活动和气候变化作为两个相对独立的驱动因素分开，忽略了两者之间的相互作用（Yin et al.,2020）；此外，短期不确定性和时滞效应对两者来说也是不可避免的（Teng等人，2020）。为了进一步评估两个模型之间的偏差，选择了人口密度较低、人类活动相对较弱（气候变化和人类活动相对独立的区域）的四川西部国家级自然保护区来讨论两个模型之间的偏差和稳定性（更多细节可在SI中找到）。

Vegetation in an ecosystem absorbed CO₂ and produced organic matter through photosynthesis, so a link was made between the amount of CO₂ sequestered and organic matter by calculating the mass of organic matter produced by vegetation over time. Previous studies have used this method to conduct an in-depth exploration of carbon (Chen et al., 2020; Liang et al., 2023). The chemical formula of photosynthesis in vegetation: 6CO₂+ 6H₂O–C₆H₁₂O₆+6O₂, the detailed carbon sequestration capacity (CSC) and the total carbon sequestered (TCS) were shown in equations (8), (9)):(8)$\mathit{CSC}=1.62\frac{NPP}{0.45}$(9)$TCS=\mathit{CSC}*250*250$where CSC (g/m²) is the value of the carbon dioxide sequestered by the vegetation, represented the carbon sequestration capacity of a particular image element, and the unit of average carbon sequestration capacity per unit time is g/m²/t (Chen et al., 2020). The TCS (Tg) of a pixel was obtained by multiplying CSC with the area of the pixel.
生态系统中的植被吸收CO ₂ 并通过光合作用产生有机物，因此通过计算植被随时间产生的有机物质量，在CO ₂ 封存量和有机物之间建立了联系。先前的研究已经使用这种方法对碳进行了深入的探索（陈等，2020；梁等，2023）。植被光合作用的化学式：6CO ₂ +6H ₂ O-C ₆ H ₁₂ O ₆ +6O ₂ ，详细的碳封存能力（CSC）和总碳封存量（TCS）如式（8）、（9）所示： (8)$\mathit{CSC}=1.62\frac{NPP}{0.45}$ (9)$TCS=\mathit{CSC}*250*250$ 其中CSC（g/m ² ）是植被封存的二氧化碳的值，代表特定图像元素的碳封存能力，单位时间平均碳封存能力的单位为g/m ² /t（Chen等人，2020）。通过将CSC乘以像素的面积来获得像素的TCS(Tg)。

2.3. Statistical analysis
2.3.统计分析

Through applying the one-dimensional linear regression method, we analyzed the CSC to study the spatial and temporal trends.(10)$Slope=\frac{n\times \sum _{i=1}^n\left(i\times {Var}_i\right)-{\sum }_{i=1}^{n}i\times {\sum }_{i=1}^n{Var}_i}{\mathrm{n}\times {\sum }_{i=1}^n{i}^2-{\left({\sum }_{i=1}^{n}i\right)}^2}$where Slope is the trend of the relevant variable and its positive or negative value indicates the degree of variation of the variable over time; n is the study period; Var_i is the variable at time i.
通过应用一维线性回归方法，我们分析了CSC的时空变化趋势。 (10)$Slope=\frac{n\times \sum _{i=1}^n\left(i\times {Var}_i\right)-{\sum }_{i=1}^{n}i\times {\sum }_{i=1}^n{Var}_i}{\mathrm{n}\times {\sum }_{i=1}^n{i}^2-{\left({\sum }_{i=1}^{n}i\right)}^2}$ 其中斜率是相关变量的趋势，其正值或负值表示变量随时间的变化程度；n为研究周期；Var是i时刻的变量。

The partial correlation coefficient is a statistical measure of the correlation between variables based on the correlation coefficient. When investigating the association between multiple variables, the partial correlation coefficient effectively excludes the influence of redundant variables, thus reflecting the relationship between two variables more accurately. The relationship between climatic factors and vegetation was also commonly investigated by partial correlation coefficients (Yan et al., 2019).(11)$R_{xy,z}=\frac{R_{xy}-R_{xz}{R}_{yz}}{\sqrt{\left(1-R_{xz}^2\right)}\times \sqrt{\left(1-R_{yz}^2\right)}}$
偏相关系数是基于相关系数的变量之间相关性的统计度量。在考察多个变量之间的关联时，偏相关系数有效地排除了冗余变量的影响，从而更准确地反映了两个变量之间的关系。气候因素和植被之间的关系也通常通过偏相关系数进行研究（Yan等人，2019）。 (11)$R_{xy,z}=\frac{R_{xy}-R_{xz}{R}_{yz}}{\sqrt{\left(1-R_{xz}^2\right)}\times \sqrt{\left(1-R_{yz}^2\right)}}$

The correlation coefficients between the three factors x, y, and z indicate the correlation between these three factors two by two (Rxy、Rxz、Ryz), respectively. Rxy,z represents the partial correlation coefficient between the X-factor and the Y-factor without considering the Z-factor.
三个因素x、y和z之间的相关系数分别表示这三个因素之间的两两(Rxy、Rxz、Ryz)的相关性。Rxy，z表示X因子和Y因子之间的偏相关系数，而不考虑z因子。

3.1. The spatial distribution and variation of carbon sequestration capacity
3.1.固碳能力的空间分布与变化

The annual TCS within the study area during the last 11 years ranged from 740.86 to 907.89 Tg. The variation of TCS in the study area fluctuated considerably from 2010 to 2015, after which it gradually stabilized at around 850 Tg; The proportion of TCS was relatively stable across seasons and did not show substantial inter-annual variation. Whereas there were significant differences in the TCS of different land types, forest (46.26%), cropland (27.68%) and grassland (22.72%) accounted for 96.66% of the TCS in the study area, while the other six land types accounted for only 3.34% (More details about the TCS can be found in SI). In general, TCS was not only affected by the size/area of the corresponding land types but also by the CSC. Thus, the following study will focus on the spatial distribution and variation of CSC and its influence factors.
在过去11年中，研究区域内的年TCS范围为740.86至907.89 Tg。2010年至2015年，研究区TCS的变化波动较大，之后逐渐稳定在850 Tg左右；TCS的比例在不同季节相对稳定，没有显示出显著的年际变化。不同土地类型的TC存在显著差异，森林（46.26%）、农田（27.68%）和草地（22.72%）占研究区TC的96.66%，而其他六种土地类型仅占3.34%（关于TC的更多细节可在SI中找到）。一般来说，TCS不仅受到相应土地类型的大小/面积的影响，还受到CSC的影响。因此，以下研究将重点关注CSC的空间分布和变化及其影响因素。

The regional distribution of CSC in the study area was distinct, with higher urbanized areas had lower CSC and higher CSC concentrated in mountainous areas. As one of the most densely populated regions in China, the development process in Sichuan-Chongqing has been accelerating in recent years, especially in the basin hinterland, which was highly urbanized (Li and Lin, 2022). Most human activities were concentrated in the eastern part of the study area, especially in the two mega-cities of Chengdu and Chongqing. Therefore, the CSC in the dense urban areas in the basin was relatively low, specifically in the urban areas of each city and the two mega-cities. Unlike the basin urban agglomeration, the west Sichuan plateau had relatively low CSC due to relatively high altitudes (Fig. 1 (a)). In contrast, the high value of CSC was concentrated in mountainous areas with lush vegetation and low impact of human activities, such as the mountainous areas along the basin and southern Sichuan. According to the administrative divisions (SI Fig. S-5), the average multi-year carbon sequestration in Guangyuan, Bazhong, and Dazhou, which were in the Qinba mountainous region in the north of the study area, was 1670.78 ± 108.41 g/m²/a (Fig. 1 (b)); the annual average CSC in the mountainous region of southern Sichuan (Liangshan and Panzhihua) was 1629.27 ± 82.23 g/m²/a, and the above were the five cities with the highest CSC (Ordered by: Bazhong 1746.1 g/m²/a, Panzhihua 1670.78 g/m²/a, Guangyuan 1645.49 g/m²/a, Dazhou 1620.76 g/m²/a, Liangshan 1588.56 g/m²/a) in the study area respectively. The three cities with the lowest average CSC in the study area were Chengdu, Deyang, and Ganzi, with CSC of 1094.23 g/m²/a, 1100.21 g/m²/a, and 1142.20 g/m²/a, respectively. The CSC of different land types showed significant discrepancies (Fig. 2). The maximum average CSC was 1700.0 ± 252 g/m²/a in the forest grid points, followed by 1496.23 ± 218.35 g/m²/a (wetland), 1482.44 ± 267.68 g/m²/a (cultivated land); while only 668.22 ± 112.6 g/m²/a (bare land) and 409.23 ± 117.53 g/m²/a (permanent snow and ice), respectively.
研究区CSC的区域分布不同，城市化程度较高的地区CSC较低，而较高的CSC集中在山区。作为中国人口最稠密的地区之一，川渝地区近年来的发展进程不断加快，尤其是高度城市化的盆地腹地（李和林，2022）。大部分人类活动集中在研究区的东部，尤其是成都和重庆两个特大城市。因此，流域密集城区的CSC相对较低，特别是各城市和两个特大城市的城区。与盆地城市群不同，川西高原由于海拔相对较高，CSC相对较低（图1（a））。相比之下，CSC的高值集中在植被茂盛、人类活动影响较小的山区，如盆地沿岸山区和四川南部。根据行政区划（SI图S-5），位于研究区北部秦巴山区的广元、巴中和达州的多年平均固碳量为1670.78±108.41 g/m ² /a（图1（b））；川南山区（凉山、攀枝花）年平均CSC为1629.27±82.23 g/m ² /a，以上为研究区CSC最高的5个城市（依次为：巴中1746.1 g/m ² /a、攀枝花1670.78 g/m ² /a、广元1645.49 g/m ² /a、达州1620.76 g/m ² /a、凉山1588.56 g/m ² /a）。研究区平均CSC最低的三个城市是成都、德阳和甘孜，CSC分别为1094.23 g/m ² /a、1100.21 g/m ² /a和1142。9#/a时分别为20 g/m。不同土地类型的CSC显示出显著差异（图2）。森林网格点平均CSC最大值为1700.0±252 g/m ² /a，其次为1496.23±218.35 g/m ² /a（湿地），1482.44±267.68 g/m ² /a（耕地）；而13#/a（裸地）和14#/a（永久冰雪）分别仅为668.22±112.6 g/m和409.23±117.53 g/m。

1. Download: Download high-res image (2MB)
2. Download: Download full-size image

Fig. 1. The mean CSC spatial distribution (a) and mean CSC in different cities (b) in Sichuan-Chongqing from 2010 to 2020 (In (b), the yellow background was the average altitude of the city from 0 to 700 m, the red background was 700–3500 m, and the blue was above 3500 m).

1. Download: Download high-res image (270KB)
2. Download: Download full-size image

Fig. 2. CSC of different land types in Sichuan-Chongqing from 2010 to 2020.

The trends of CSC and TCS with increased altitude were disparate (Fig. 3 (a) and SI Fig. S-4). The topography of Sichuan-Chongqing was complicated and undulating, with a variety of topographical features in the study area: Sichuan Basin, Qinba Mountains, Hengduan Mountains, western Sichuan Plateau, etc. Spatial differentiation of vegetation from climate and topography was apparent (Shi et al., 2023). The amount of carbon sequestration at different altitudes was summarized in units of 50 m above sea level to obtain the characteristics of carbon sequestration with altitude. The results showed that the CSC increased rapidly until 700 m; Then at altitude of 700–3500 m, the CSC was relatively stable between 1700 and 1900 g/m²/a (average 1817.24 ± 89.16 g/m²/a); After the altitude rise to 3500 m, the CSC showed a gradual decrease and convergence to 0 at around 5000 m. The percentage of each land type was also counted according to the altitudes, and the results showed that at altitude 0–700 m, cultivated land was dominant; from 700 to 3500 m, the highest percentage of forest area was occupied; above 3500 m, the highest percentage area was grassland (Fig. 3(b,c,d)). The TCS was influenced by the number of grid points as well as the CSC, and its increase with altitude was specified as follows: the TCS increased rapidly to reach a maximum of around 400m, and the multi-year average TCS can reach 35.29 Tg, then declined, but increased to about 10 Tg at 3400m–4400m, and then decreased rapidly to 0 (SI Fig. S-4).

1. Download: Download high-res image (469KB)
2. Download: Download full-size image

Fig. 3. The CSC at different altitudes (a) and the percentage of land types (b. 0–700 m; c.700–3500 m; d. Above 3500 m) in Sichuan-Chongqing from 2010 to 2020.

The variation trend of CSC during 2010–2020 exhibited distinct regional features and manifests as increased trend in the east (Sichuan basin and surrounding mountainous areas) and decreased trend in the west (western Sichuan plateau). As shown in Fig. 4, the most prominent increased trend was found in the western mountains of the basin (Longmen Mountain, Qionglai Mountain); the cities of Leshan and Yibin, which were located in the southern basin, also had a prominent increased trend. As for the mountains in southern Sichuan, the overall trend increased in the north and decreased in the south. The regions with an increased trend of CSC account for about 55.55% of the total area, with about 10.66% of the zones exceeding +50 g/m²/a. In addition to the western and southern sides of the basin, Wushan in Chongqing and the eastern part of the Sichuan Ridge Valley showed an increase in CSC of more than +50 g/m²/a, but this did not show a widespread regional concentration like the former. Only 3.16% has a decreased trend of CSC less than −50 g/m²/a, mainly in the western Sichuan Plateau (especially Ganzi and Aba). From the multi-year average rate of change of each city, only four cities in the plateau of western Sichuan (Ganzi, Aba) and the mountains of southern Sichuan (Panzhihua, Liangshan) showed a decreased trend of CSC.

1. Download: Download high-res image (1MB)
2. Download: Download full-size image

Fig. 4. Variation trend of annual CSC (a) and average trend of annual CSC by cities (b) in Sichuan-Chongqing from 2010 to 2020 (In (b), the yellow background was the average altitude of the city from 0 to 700 m, the red background was 700–3500 m, and the blue was above 3500 m).

3.2. The relative contributions of climate change and human activities to CSC variation

The spatial distribution of distinct drivers was calculated using the trends of CSC and impact factors, and the relative contribution of different dominant impact factors to CSC was investigated. The specific driving factors can be divided into six categories (Fig. 5)(Yuan et al., 2021). The variation of CSC caused by different drivers in various regions of Sichuan-Chongqing in the past 11 years showed significant differences. Relatively significant impacts of climate change were observed in the hinterland of the Sichuan basin; however, in the west of the area studied, with a relatively significant impact of human activities on CSC, especially in Ganzi; while the combined effect of climate change and human activities was found in the basin margins and the mountainous along it. It seemed that climate change usually leads to changes in CSC at relatively low elevations, and human activities gradually dominate and influence CSC when elevation reaches a certain altitude.

1. Download: Download high-res image (1MB)
2. Download: Download full-size image

Fig. 5. Spatial distribution of different impact factors and the relative contribution on CSC from 2010 to 2020.

For the eastern basin and its margins, the increase in CSC was mainly manifested by the combination of climate change and human activities. This combined effect of the two on CSC accounts for about 31.59%. About 29.72% where both together contribute to the increase of CSC, and only 1.87% exhibit the contribution to the decrease of CSC (Fig. 5 (b)). As shown in Fig. 6, the relative contributions of climate change and human activities driven increases in CSC were further calculated, where higher values indicated a more significant contribution of climate change to CSC and lower values indicated a more significant contribution of human activities to CSC. The degree of contribution of different influencing factors showed that human activities have promoted the increase of CSC in the mountainous region along the basin, and its influence gradually diminished with the movement towards the center of the basin, and climate change as the main contribution to the increase of CSC. Besides, in the northeastern part of Liangshan, climate change dominated the increase of CSC.

1. Download: Download high-res image (846KB)
2. Download: Download full-size image

Fig. 6. The degree of contribution to CSC from a combination of climate change and human activities.

As for human activities as the main influence factor, CSC reduction was caused by human activities as the primary contribution in approximately 41.92% and was found mainly on the west and south sides of western Sichuan plateau and southern Sichuan mountains (Fig. 5(c)). Only 4.49% of the area exhibited an increase in CSC facilitated by human activities, mostly occurring at the edges. Human activities mainly affect the western Sichuan plateau and the southern Sichuan mountains, a region with relatively high altitude and fragile ecosystems, while the territory was mostly dominated by animal husbandry. Based on the national environmental policy, people have taken measures such as afforestation and reforestation help to improve the NPP of vegetation; while the rapid urban expansion, industrialization, overgrazing, and other human activities in recent years have brought a certain level of challenge to the ecological environment (Huang et al., 2020; Ning et al., 2022), and the impacts of human activities will be further discussed in Section 3.3.

Climate change mostly contributed to the increase of CSC (21.34% of the area where climate change had a positive effect), while only 0.66% of the area contributed negatively to CSC. Most studies on the response of NPP to climate change mainly were concentrated on temperature and precipitation, which had a significant effect on vegetation compared to other climate factors (Fatima et al., 2020). As shown in Fig. 7, the CSC and annual mean temperature exhibited a positive correlation in the east but a negative correlation in the west. CSC was positively related to temperature in about 60.31% of the areas, mainly in the eastern part; in the area between 700 and 3500m, CSC was positively correlated with temperature in approximately 68.79% of the cases, while 31.21% were negatively correlated; the CSC in the western part of the study area (Ganzi, Aba, Liangshan, Panzhihua), where the average elevation reached 3000 m, was negatively correlated with temperature in most areas. Approximately 13.14% of these areas were significantly correlated with the average annual temperature (P < 0.05). The spatial distribution of the correlation between CSC and precipitation differed significantly from that of the correlation with temperature. The correlation between CSC and precipitation was negative in about 60.19% of the areas, mainly in most of the western Sichuan plateau and the center of the Sichuan basin, while positive correlations were found in the southern part of the Sichuan basin, most of Chongqing and the mountainous areas along the basin. Only 7% of these areas were significantly correlated with the total annual precipitation (P < 0.05).

1. Download: Download high-res image (2MB)
2. Download: Download full-size image

Fig. 7. Spatial distribution of skewed correlation coefficients and significance of mean annual temperature, total annual precipitation, and CSC.

3.3. Discussion

Instead of focusing on different ecological regions, this study investigated the CSC in Sichuan-Chongqing from the perspective of administrative divisions. From a more fine-grained perspective, the changes in carbon sequestration in recent years and the influencing factors were analyzed in the context of vegetation studies. There was a large vertical gradient of topography in the study area, and elevation was an essential factor influencing vegetation, which had an influential role in determining the growth of vegetation as well as ecological restoration (Du et al., 2020; Shi et al., 2023; Zheng et al., 2020b). The two iconic heights were 700 m and 3500 m. The trend of CSC increased significantly with the elevation until reaching 700 m height and decreased after 3500 m. A similar conclusion was reached in the study by Wang et al., 2017a, after 3500 m, NPP diminished with increasing elevation. The research area can be roughly classified into three classes (SI Fig. S-5) according to altitude: the Sichuan basin (area percentage 28.57%), which was mostly below 700 m in elevation; the mountainous areas along the basin and southern Sichuan (36.53%), which are between 700 and 3500 m in altitude; and the western Sichuan plateau (34.93%), which is above 3500 m. The general distribution of NPP (carbon sequestration) of the area studied was similar to that of Liu et al. (2013), but NPP (carbon sequestration) increased in different regions. Compared to previous studies, they focused more on the trend of carbon sequestration/NPP was concentrated in the western Sichuan region, and the studied time period was mostly around 2000–2015. Li et al. (2022) used the Factor Analysis Backpropagation neural network model to estimate the NPP in the western Sichuan region between 2000 and 2016, and the area with an increased trend of NPP in the corresponding study period was about 81.42% and a decreased area only accounted for 18.58% (while the area that passed the significance test was only 8.14%). Xu et al. (2023) investigated the characteristics of NPP spatial variation in Southwest China, which showed that in eastern Sichuan and western Chongqing, NPP was significantly increased; while the NPP variation trend in western Sichuan was not significant. Encouraged by the policy, the carbon sequestration in densely populated urban areas in the eastern part of the area studied has shown an increased trend in recent years, which was consistent with Xu et al. (2023); however, in the western part of Sichuan, there was a distinct trend of weakening.

As for the spatial distribution and variation of CSC, the area with an increased trend of CSC was approximately 55.55%, of which climate change was the primary contributing factor, with 21.34%, human activities accounted for 4.49%, and the combination of the both together contributed to the increase in CSC accounting for approximately 29.72% of the region. But in the regions with a decreased CSC trend (44.45%), the role of human activities accounted for 41.92% and the part of climate change and both together account for only 2.53%. At the global scale, climatic factors explain about 54% of the vegetation activity trends, while the remaining part is a collection of other factors, including human activities (de Jong et al., 2013). The negative impact of human activities in recent years was focused on western Sichuan, among which the area of CSC reduction due to human activities in Aba, Ganzi, and Liangshan was about 55.97, 122.07, and 32.24 thousand square kilometers, respectively. Compared with previous studies, we found that the CSC in western Sichuan, especially in Ganzi, had tended to decrease in recent years. In western Sichuan, human activities were the main influencing factor, and Yin et al. (2020) also indicated that in the Hengduan Mountains, human activities contributed 66.11% to the change of NPP. Previous studies have mainly focused on the period 2000–2015 (Yin et al., 2020; Chen et al., 2019a), while this research selected the relatively closer period 2010–2020, and the difference in the study period was also an important reason for the divergence of the results. However, the results of Chen et al. (2017) showed a decreased trend of NPP in about 67.2% of the Transverse Mountain region during 2001–2014. The western Sichuan plateau had a large grassland area (SI Fig. S-2). Although Aba and Ganzi accounted for about 1.08% and 1.37% of the province's population (SI Fig. S-6), the number of large livestock accounted for 24.05% and 27.53% of Sichuan Province (SI Fig. S-7), respectively. The main form of grassland utilization in arid and semi-arid regions was grazing, and different livestock types and numbers had different intensities of grazing, which in turn directly affected the change of grassland (Cao et al., 2004; Zhang et al., 2018). As a result, it makes sense that the CSC of the western Sichuan plateau would be reduced. The southern part of Sichuan was a little different, with the number of sheep and goats in Liangshan accounted for about 32.49% of the province (average for 2014–2020). The excessive development of unnatural urban economies such as mining, agriculture, and animal husbandry in Panzhihua had led to the deterioration of its ecological environment (Dai et al., 2021).

Unlike human activities, climate change mostly assisted in the increase of CSC, with 21.34% of areas had a positive effect of climate change (excluding the combined effect of human activities and climate) and only a minority of regions where it had a negative effect on CSC. In recent decades, extreme precipitation, drought, and heat caused by global climate change have become more frequent, and the global scale potential NPP increased by about 13% from 1901 to 2001 (Del Grosso et al., 2008). Natural disasters were also an important factor which could affect vegetation carbon sequestration, such as floods caused by extreme precipitation (Caruso et al., 2013), extreme climate-induced droughts, fires, and low-temperature freezes affected the physiological structure of vegetation and its ecosystem, making vegetation mortality higher (Eggen et al., 2019; Reyer et al., 2013). Maintaining a specific range of temperature and precipitation is beneficial for vegetation; excessively high or low will damage vegetation growth (He et al., 2022; Shi et al., 2023). The unique topography of Sichuan-Chongqing has resulted in large differences in climate characteristics between regions. The spatial distribution of precipitation was uneven, with high in eastern and low in western, and high in basin and low in plateau, with similar characteristics of temperature. In recent years, the annual variation of temperature compared to precipitation showed little difference, but the trend of precipitation increased year by year (SI Fig. S-8). Spatially, the Sichuan basin urban agglomeration was more appropriate for vegetation growth than other regions in terms of both precipitation and temperature, but CSC was not higher, especially in the hinterland of the basin. However, with a dense population and high urbanization, benefiting from policies and people's increased environmental awareness, the CSC, in general, has shown an increased trend within the basin, and in the mountainous areas along the basin, the CSC had increased significantly in the past few years. For high altitude areas in western Sichuan, alpine vegetation systems were most sensitive to temperature increase, but some studies showed that higher temperature increased water vapor pressure difference, which partially closed stomata and reduced photosynthetic efficiency (Chen et al., 2021). Also, precipitation may affect the organic matter of the soil, which in turn affected the NPP of the grass (Xu et al., 2016). The effect of human activity was much more significant in western Sichuan, but precipitation in the region did not change significantly, and the relatively warm and dry environment was not beneficial to the vegetation growth (Yan et al., 2019). In the majority of the Sichuan basin, CSC was positively associated with temperature; however, the spatial distribution of the correlation of CSC with precipitation was relatively complex, with a negative correlation in the Chengdu Plain. The change in precipitation showed a spatial distribution trend of increased in the south and decreased in the north. Li et al. (2023) showed that sufficient precipitation facilitated the CSC of forested areas in the Sichuan basin. The variation of precipitation in the Chengdu plain area was smaller than that in the cities of southern Sichuan. In several cities in the southern Sichuan basin, precipitation increased by more than 20 mm, and these regions were associated significantly and positively with precipitation.