Open Access Published by De Gruyter Open Access June 9, 2021

Analysis on vegetation cover changes and the driving factors in the mid-lower reaches of Hanjiang River Basin between 2001 and 2015

Yunjun Zhan, Jiemeng Fan, Tingting Meng, Zhongwu Li, Yan Yan, Jiejun Huang, Di Chen and Lintong Sui
From the journal Open Geosciences

Abstract

The mid-lower reaches of the Hanjiang River Basin, located in the core of economic development in Hubei Province, is an integral part of the Yangtze River Economic Belt. In recent years, the watershed ecosystem has become more sensitive to climate changes and human activities, thus affecting the regional vegetation cover. To maintain a stable watershed ecosystem, it is critical to analyze and evaluate the vegetation change and its response to temperature, precipitation, and human activities in this region. This study, based on the trend analysis, partial correlation analysis, and residual analysis, evaluated the change characteristics of vegetation cover as well as the corresponding driving factors in the basin from 2001 to 2015. The results showed that (1) the overall spatial pattern of vegetation cover in the study area was “high in the west and north, lower on both sides of Hanjiang River, and lowest in the center and southeast,” and the pattern changed parabolically with the increasing elevation. (2) Over the 15 years, vegetation cover in the basin showed an increasing trend, and the increased and decreased areas were 90.72 and 9.23%, respectively. (3) The response of vegetation cover to climatic factors varies greatly depending on the increasing elevation. That is, the lag effect under the impact of temperature disappeared gradually, while it became more evident under the impact of precipitation. (4) On the whole, human activities had a positive effect on the regional vegetation cover. The negative effect in the areas around the Nanyang Basin and the positive effect in most parts of the Jianghan Plain were gradually decreased.

1 Introduction

Hanjiang River, the primary tributary of Yangtze River, is the water source for the middle route of the South-to-North Water Transfer Project. The Hanjiang River Basin is one of China’s major commodity grain bases. Its eco-environmental changes significantly affect the Yangtze River Economic Belt and the ecological and grain safety of regions along with the South-to-North Water Transfer Project. Vegetation cover change, one of the important indexes for monitoring regional eco-environmental quality [1,2,3], can effectively reflect the changes in climate, hydrology, and human activities in basins. Against the backdrop of hydraulic engineering construction and modern agricultural development, studying vegetation cover change in the Hanjiang River Basin and the corresponding driving factors is of great practical importance.

Climatic factors and human activities are the two most common indicators in vegetation cover studies. Especially as the global climate is radically changing, vegetation response to such a situation has become a central concern for many scholars [4,5,6]. A lot researches have showed that higher annual average temperatures are the leading cause of vegetation change [7,8,9]. In addition, the vegetation growth condition was susceptible to precipitation variation [10,11,12], especially in arid and semi-arid areas [13,14]. Besides climate changes, human activities such as urban construction, environmental governance, and ecological engineering are also crucial to vegetation cover [15,16,17], which has gained much attention [18,19,20]. Other researchers concentrated more on the correlation between vegetation cover and topographic factors, including altitude, slop, and aspect [21,22,23].

Currently, studies on the vegetation cover in the Hanjiang River Basin mostly center on Southern Shaanxi, where the water comes from, and the upper reaches [24,25,26,27]. Also, the upper and mid-lower reaches of the Hanjiang River differ in landform and hydrologic patterns. They are facing different ecological problems, especially under the influence of hydraulic engineering construction. So far, many researchers have analyzed from different perspectives such as river water quality, industrial development, and climate (temperature, humidity, and relative humidity) [2830]. However, there is still lack of analysis of vegetation cover changes. Therefore, taking the mid-lower reaches of the Hanjiang River Basin as the study object, this study analyzed the change characteristics of the regional vegetation cover as well as its driving factors over the 15 years, with the elevation elements taken into consideration.

The specific approaches of this study were as follows: (1) Based on the Normalized Difference Vegetation Index (NDVI) and meteorological data from 2001 to 2015, this study disclosed the change characteristics of vegetation cover in different elevations over the 15 years using the pixel-based linear fitting approach. (2) With partial correlation, this study further analyzed the effect of temperature and precipitation in different periods on vegetation in various elevations. (3) This study established a “climate-NDVI” model to probe into impact of human activities on vegetation under the pretext of hydraulic engineering construction and modern agricultural development using the residual analysis method. The study’s significance is to provide a relevant theoretical basis and realistic reference for the formulation of policy on vegetation restoration and ecological protection in the water basin.

2 Methods and data

2.1 Study area

Hanjiang River, the Yangtze River’s primary tributary, has a trunk stream of 1,577 km in length. The segment from Danjiangkou to Hankou is the mid-lower reach, which is as long as 652 km. The Hanjiang River Basin, with a control area of 64,000 km2, locates in the subtropical monsoon region, where the climate is mild and humid, and the annual precipitation is 800–1,300 mm. The water is abundant, but is unevenly distributed throughout the year. The runoff from May to October approximately accounts for 75% of the whole year. With a great interannual variation, the Hanjiang River is the most changeable one among all tributaries of the Yangtze River. Thereinto, the middle reach is about 270 km, extending from Danjiangkou to Zhongxiang. It is mainly covered by mountain lands (distributed in the west and the north) and hills (distributed in the Nanyang Basin) and primarily used for croplands and forest lands. The segment below Zhongxiang is the lower reach, which is about 382 km. Here, the topographic relief is not significant, and the river branches crisscross this reach. There are many flat alluvial plains and small lakes (distributed in the Jianghan Plain), and the reach is mainly used for croplands.

2.2 Study data

The NDVI data from 2001 to 2015 were derived from the MOD13Q1 product on the NASA MODIS Terrestrial Science Team (http://landval.gsfc.nasa.gov). The spatial and temporal resolutions of this product are 250 m and 16 days. Now, it has been widely used for the dynamic monitoring of regional vegetation change. Considering the honological phase and climate characteristics in the study area, the NDVI data from April to October during these 15 years were accumulated to get NDVIg (the cumulative growing-season NDVI).

The meteorological data were obtained from the China Meteorological Data Service Center (http://data.cma.cn/), and observations from a total of 22 meteorological stations in and around the study area were selected. Since there are many methods to interpolate meteorological data, it is important to choose a suitable one according to the natural geographical characteristics of the study area. Thin plate spline (TPS) interpolation is a surface extension of the spline function, which uses smooth parameters to achieve an optimal balance between data fidelity and the smoothness of the fitted surface, ensuring a smooth and continuous interpolation surface with reliable accuracy [31]. Meanwhile, TPS can also introduce linear covariate submodels, such as temperature and elevation, which are suitable for areas with large topographic relief. Considering the topographic characteristics of the study area, this article used TPS for spatial interpolation of the temperature data. Ordinary Kriging (OK) interpolation can not only estimate the spatial distribution of the measured parameters but also estimate the variance distribution of the estimated parameters, which is one of the most common methods for interpolation of the precipitation data [31]. Liu and Chen have pointed out that the highest accuracy could be obtained by OK interpolation for precipitation data in the Yangtze River Basin [32]. Therefore, this article chose to use OK interpolation for spatial interpolation of precipitation data.

The DEM data and vegetation type data were obtained from the Resource Environment Data Cloud Platform of CAS (http://www.resdc.cn/) and both were resampled to 250 m.

2.3 Study methods

2.3.1 Unary linear regression trend analysis based on the pixel

The growing-season NDVI g from 2001 to 2015 was calculated, and the least square method was employed to fit the change rate of NDVI g pixel by pixel. The variation trend of a single pixel was used to comprehensively reflect the spatial-temporal variation characteristics of vegetation cover in the study area. The computational formula can be expressed as follows:

(1) B slope = n i =1 n NDVI g t i i =1 n NDVI g i =1 n t i n i =1 n t i 2 i =1 n t i 2
where NDVI g refers to the cumulative growing-season NDVI that can reflect the vegetation growth condition throughout the year. Besides, t i represents the ith year of the research ( n = 15), and B slope is the slope of NDVI g ’s variation trend line. The situation that B slope > 0 means the NDVI g keeps increasing over the 15 years. The higher the value, the more obvious the increase of vegetation coverage. Conversely, the coverage will decrease.

The F test was used to test the significance of the change trend, and is calculated as follows:

(2) F = U × n 2 Q ,
where U = i =1 n ( y ˆ i y ¯ ) 2 is the regression square sum, Q = i =1 n ( y i y ˆ i ) 2 is the residual sum of square, y i is the i th year of NDVI g, y ˆ i is the regressive value, y ¯ is the average value of 15 years, and n = 15.

2.3.2 Calculation method about the influence of climate changes on the vegetation cover based on partial correlation analysis

The influence of climatic factors on the vegetation is mainly manifested as temperature and precipitation. By controlling the linear effect of the temperature (precipitation) in this article, the partial correlation method was employed to analyze each factor’s impact on the vegetation separately. The computational formula is

(3) r x y , z = r x y r x z r y z ( 1 r x z 2 ) ( 1 r y z 2 ) ,
where r x y , z refers to the partial correlation coefficient between variables x and y after the variable z is fixed. r x y , r x z , and r y z represent the correlation coefficients between x and y, x and z, as well as y and z, respectively. Thereinto, x, y, and z represent the NDVI g , temperature, and precipitation, respectively, and vice versa.

2.3.3 Calculation method about the influence of human activities on the vegetation cover based on residual analysis

In the context of climate changes, this study revealed the influence of human activities on the vegetation growth through the residual analysis. The specific methods are as follows:

  1. (a)

    The Danjiangkou Reservoir Area of the Middle Route of the South-to-North Water Transfer Project had started a 5-year immigration plan at the end of 2008, which caused a huge disturbance to the regional vegetation cover. Thus, this article took 2001–2008 as the baseline period and 2009–2015 as the influence period. Besides, it was assumed that there was a balanced state among the vegetation system, climate changes, and human activities from 2001 to 2008.

  2. (b)

    The “climate-NDVI” model based on climatic factors was obtained by means of the multiple regression analysis, and its simulation effect was evaluated with an adjusted determination coefficient. The computational formula can be expressed as follows:

    (4) R a 2 = 1 SSE / ( n k 1 ) SST / ( n 1 ) ,
    where R a 2 is the adjusted determination coefficient, SSE is the residual sum of squares, and SST is the sum of squares of total deviations. Besides, nk − 1 and n − 1 represent the degrees of freedom for squares’ error sum and the total sum of squares, respectively.

  3. (c)

    The “climate-NDVI” model was used to predict the NDVI g from 2009 to 2015 in the mid-lower reaches of the Hanjiang River Basin.

  4. (d)

    The residual between real and simulated NDVI g values was calculated, and its variation trend with the year was analyzed. The computational formula of this residual is expressed as follows:

(5) σ = NDVI g NDVI g-sim ,
where NDVI g-sim represents the simulated NDVI g , and σ represents the residual. If σ is positive, human activities will promote the vegetation cover. Conversely, they will impede the vegetation cover.

3 Results

3.1 Spatial-temporal variation characteristics of the vegetation cover

An overlay analysis was carried out based on the distribution of vegetation types (Figure 1(b)) and DEM data in the Basin to obtain their distribution at different elevations through statistics (Table 1). The concrete results were as follows: First, crops and most coniferous forests that were more frequently affected by human activities were primarily found below 300 m. Second, broad-leaved forests, shrub forests, and few crops were mixed between 300 and 1,900 m. These areas were mainly planted with economic trees, such as walnut, chestnut, and oil plants. Third, the shrub, coniferous, and broad-leaved forests were sparsely distributed above 1,900 m.

Figure 1 Location and vegetation types of the study area: (a) distribution of DEM and weather stations and (b) distribution of vegetation types.

Figure 1

Location and vegetation types of the study area: (a) distribution of DEM and weather stations and (b) distribution of vegetation types.

Table 1

Vegetation distribution at different elevations

DEM Vegetation distribution Regions Description
≤300 m Crops: 93.363%; coniferous forests: 71.876%a Nanyang Basin Most areas in the middle reaches of the Hanjiang River
Jianghan Plain Most areas in the lower reaches of the Hanjiang River
300–1,900 m Broad-leaved forests: 77.901%; shrub forests: 57.575%; crops: 6.363% Western and northern mountainous areas (areas around Nanyang Basin) Fangxian, Gucheng, Baokang, Nanzhang, Neixiang, Nanzhao
>1,900 m Shrub forests: 4.226%; coniferous forests: 5.208%; broad-leaved forests: 3.237% Shennongjia Forestry District Shennongjia

    a

    The coniferous forests distributed in the mid-lower reaches of the Hanjiang River Basin are mostly the secondary forests or the semi-natural forests formed through artificial cultivation [33,34].

3.1.1 Temporal variation characteristics

From the whole basin (Figure 2(a)), the annual average NDVIg was 4.01 over the 15 years, and the vegetation cover fluctuated and increased. From different elevations, the annual vegetation NDVIg was averagely 3.70 at altitudes below 300 m and increased year by year in a similar way as that in the whole basin. The vegetation cover also showed an increased trend between 300 and 1,900 m. However, there was no significant change above 1,900 m. It is worth noting that the vegetation cover plummeted at altitudes above 1,900 m in 2010 for the Hanjiang River witnessed a once-in-a-century torrential rain that led to flash floods and landslides. Consequently, the mountainous vegetation was destructed extensively [35].

Figure 2 Temporal variation characteristics of the vegetation cover in the mid-lower reaches of the Hanjiang River Basin from 2001 to 2015: (a) annual variation of the NDVIg at different elevations and (b) monthly variation of the NDVIg at different elevations.

Figure 2

Temporal variation characteristics of the vegetation cover in the mid-lower reaches of the Hanjiang River Basin from 2001 to 2015: (a) annual variation of the NDVIg at different elevations and (b) monthly variation of the NDVIg at different elevations.

According to the month-by-month analysis, how the monthly average values of the growing-season NDVI changed with the elevation is shown in Figure 2(b), indicating that the vegetation NDVI’s intraannual variation at different elevations had obvious seasonal characteristics in the Basin. The details were as follows: The vegetation was in a regreening stage from April to June when the NDVI value increased gradually. In July and August, the vegetation grew vigorously, and its NDVI surged. September and October were the final stage of the growing season, in which the value decreased gradually. On the whole, the vegetation NDVI at altitudes below 300 m peaked in July, and that above 300 m all peaked in June. According to the research conducted by Shen and Li and the spatial distribution of vegetation types (Figure 1(b)), this is mainly because the areas below 300 m are primarily covered by crops whose growing season begins later than that of other vegetation types [36]. Therefore, the crops’ NDVI reaches its peak relatively late.

3.1.2 Spatial variation characteristics

In the mid-lower reaches of the Hanjiang River Basin, the vegetation coverage was “high in the west and the north, lower on both sides of Hanjiang River, and lowest in the center and the southeast,” showing significant intraregional differences (Figure 3(a)). Thereinto, the average growing-season NDVIg value exceeded 4.5 in the Shennongjia Forestry District, as well as western and northern mountainous areas. However, the value was 4–4.5 on both sides of the Hanjiang River and less than 4 in the Nanyang Basin and the Jianghan Plain. The NDVIg value was relatively low in the center of the basin because the vegetation may be destructed due to hydraulic engineering construction. However, the vegetation in the southeast was more likely affected by the rapid urbanization.

Figure 3 (a) Spatial distribution of the average NDVIg in the basin from 2001 to 2015 and (b) NDVIg variation with increasing elevation.

Figure 3

(a) Spatial distribution of the average NDVIg in the basin from 2001 to 2015 and (b) NDVIg variation with increasing elevation.

By calculating the average NDVIg values at different elevations in combination with the DEM data, it was found that the growing-season NDVIg changed with the increasing elevation “in a parabolic shape” (Figure 3(b)). The concrete manifestations are as follows: the NDVIg value (3.71 on average) was lowest below 300 m and ascended rapidly with the increasing elevation. The value (5.08 on average) was highest between 300 and 1,900 m and increased slowly with the increasing elevation. The value was relatively low (5.07 on average) above 1,900 m and decreased steadily with the increasing elevation.

The variation trend for the vegetation cover in the basin from 2001 to 2015 was calculated based on the unitary linear regression model (Figure 4(a) and Table 2). In the study area, the vegetation cover, accounting for 83.02% of the total area, showed an increasing trend. The extremely significantly increased area was 41.8%, mainly distributed in the eastern part of the Nanyang Basin (Sheqi, Biyang, Tongbai, Suizhou, etc.) and its surrounding mountainous areas (Gucheng, Nanzhao, etc.). The significantly increased and weakly significantly increased areas were 17.12 and 31.80%, respectively, and they covered almost the whole study area. The unchanged area was 0.05%. The decreased area was 9.23%, which was sporadicly distributed in the south of the Nanyang Basin (Xiangyang, Yicheng, etc.) and the southeastern of the Jianghan Plain (Hanchuan, Wuhan, Hanyang, etc.). The extremely significantly decreased area was 1.26%.

Figure 4 (a) Spatial distribution of Bslope in the basin from 2001 to 2015 and (b) Bslope variation with increasing elevation.

Figure 4

(a) Spatial distribution of Bslope in the basin from 2001 to 2015 and (b) Bslope variation with increasing elevation.

Table 2

Area and proportion of different vegetation change types in the mid-lower reaches of Hanjiang River

Type Significant test Area (km2) Percentage
Extremely significantly increase Bslope > 0, p < 0.01 26381.13 41.80
Significantly increase Bslope > 0, 0.01 < p < 0.05 10805.92 17.12
Weakly significantly increase Bslope > 0, p > 0.05 20073.81 31.80
Unchanged Bslope = 0 33.86 0.05
Weakly significantly decrease Bslope < 0, p > 0.05 4461.15 7.07
Significantly decrease Bslope < 0, 0.01 < p < 0.05 568.58 0.90
Extremely significantly decrease Bslope < 0, p < 0.01 791.55 1.26

The vegetation change at different elevations was analyzed based on the DEM data, namely, the relation between Bslope and DEM (Figure 4(b)). During the 15 years, the Bslope value was higher than 0 below 300 m and increased gradually with the increasing elevation. The value was also above 0 between 300 and 1,900 m, but it decreased gradually with increasing elevation. The reason may be that with the increased elevation, the accessibility and frequency of human activities gradually decreased, which weakened the effect on vegetation. Finally, the NDVIg’s Bslope value fluctuated around 0 above 1,900 m. Topographically, the vegetation cover showed an increasing trend below 1,900 m, while it changed little above 1,900 m.

3.2 Analysis of the driving factors of vegetation cover changes

To explore the vegetation’s driving factors in different regions, the NDVIg time-series data from 2001 to 2015 were used to analyze the correlation between vegetation cover and climatic factors at different elevations in the basin. In addition, the residual analysis method was employed to analyze the effect of human activities on the vegetation change.

3.2.1 Climatic factors

In combination with the DEM data, the partial correlation analysis was used to calculate the partial correlation coefficient between NDVIg and temperature (precipitation) at different elevations from interannual and intraannual two-time scales. Figure 5 shows the spatial distribution of the correlation between the multi-year mean NDVIg and the multi-year mean temperature and precipitation in the growing season on the annual scale.

Figure 5 Correlation between NDVIg and temperature and precipitation on the annual scale: (a) NDVIg and temperature and (b) NDVIg and precipitation.

Figure 5

Correlation between NDVIg and temperature and precipitation on the annual scale: (a) NDVIg and temperature and (b) NDVIg and precipitation.

From the perspective of the whole basin, the significantly negatively correlated area between NDVI and temperature was 8.58%, mainly distributed in the western mountainous areas of the Nanyang Basin. The significantly positively correlated area was 18.52%, mainly distributed in the east of the Nanyang Basin (Figure 5(a) and Table 3). The significantly negatively correlated area between NDVI and precipitation accounted for 3.96%, mainly distributed in the northern mountainous area of the Nanyang Basin. The significantly positively correlated area was 27.65%, mainly distributed in the Nanyang Basin and the Jianghan Plain (Figure 5(b) and Table 3). As the elevation increased, correlation of NDVIg with temperature increased gradually, while that with the precipitation decreased gradually (Table 2).

Table 3

Area proportion of correlation between NDVIg and climatic factors at different elevations in the mid-lower reaches of Hanjiang River Basin from 2001 to 2015

DEM Climatic factors Negative (%) Positive (%) Not significant (%)
<300 m Temperature 10.39 14.37 75.24
Precipitation 2.35 30.42 67.23
300–1,900 m Temperature 1.86 34.00 64.14
Precipitation 9.67 18.31 72.02
>1,900 m Temperature 12.31 79.11 8.57
Precipitation 5.39 16.27 78.34
The basin Temperature 8.58 18.52 72.90
Precipitation 3.96 27.65 68.39

For the change of hydrothermal conditions that have a certain lag effect on the vegetation growth, this article analyzed the correlation between the vegetation NDVI and temperature, precipitation on the monthly scale. This article calculated the partial correlation coefficients between the monthly NDVI and the temperature and precipitation in the current month, 1 month ago, and 2 months ago, and it was found that the vegetative response to temperature and precipitation showed significantly different lag effects in different regions.

From the perspective of different elevations, the vegetative response to temperature lagged by 2 months below 300 m, and the significant positive and negative correlated areas were 25.97 and 10.37%, respectively. By comparison, the response to precipitation had no obvious lag effect, and the significant positive and negative correlated areas were 16.57 and 15.41%, respectively (Table 4 and Figure 6(a and f)). Between 300 and 1,900 m, the response to temperature lagged by 2 months, and the response to precipitation lagged by 1 month. The significant positive and negative correlated areas were 23.81 and 29.91%, respectively (Table 4 and Figure 6(a and e)). There was no obvious lag effect in the response to th temperature above 1,900 m, and the response to precipitation lagged by 1 month. The significant negative and positive correlated areas were 31.58 and 24.73%, respectively (Table 4 and Figure 6(c and e)).

Table 4

Area proportion of correlation between the average monthly NDVI and climatic factors at different elevations in the mid-lower reaches of Hanjiang River Basin from 2001 to 2015

Climatic factors Time scale DEM Negative (%) Positive (%) Not significant (%)
Temperature 2 months ago <300 m 10.37 25.97 63.66
300–1,900 m 7.82 23.81 68.37
>1,900 m 27.49 3.09 69.42
1 month ago <300 m 11.63 10.91 77.46
300–1,900 m 13.48 7.27 79.25
>1,900 m 18.14 2.25 79.62
In the month <300 m 12.57 11.36 76.07
300–1,900 m 12.87 13.00 74.13
>1,900 m 31.58 0.96 67.46
Precipitation 2 months ago <300 m 4.03 22.28 73.69
300–1,900 m 9.34 9.83 80.83
>1,900 m 2.83 23.40 73.78
1 month ago <300 m 15.98 12.8 71.22
300–1,900 m 5.79 29.91 64.30
>1,900 m 10.92 24.73 64.35
In the month <300 m 15.41 16.57 68.02
300–1,900 m 30.23 5.22 64.56
>1,900 m 33.65 1.41 64.94
Figure 6 Partial correlation coefficient between the monthly NDVI and temperature and precipitation on the monthly scale: (a) NDVI and the temperature 2 months ago, (b) NDVI and the temperature 1 month ago, (c) NDVI and the temperature in the month, (d) NDVI and the precipitation 2 months ago, (e) NDVI and the precipitation 1 month ago, and (f) NDVI and the precipitation in the month.

Figure 6

Partial correlation coefficient between the monthly NDVI and temperature and precipitation on the monthly scale: (a) NDVI and the temperature 2 months ago, (b) NDVI and the temperature 1 month ago, (c) NDVI and the temperature in the month, (d) NDVI and the precipitation 2 months ago, (e) NDVI and the precipitation 1 month ago, and (f) NDVI and the precipitation in the month.

To sum up, as the elevation increased, the lag effect disappeared gradually under the impact of temperature, while it became more significant under the impact of precipitation. This is because the vegetation in croplands at low altitudes is mainly affected by human activities, and its growth is not completely synchronous with the temperature due to agricultural management measures such as mulching films and plastic houses. With the increasing elevation, the vegetation is affected by both temperature and precipitation. However, for the areas at high altitudes that are closer to the Danjiangkou Reservoir, where the water is abundant, the vegetation becomes more sensitive to temperature than to precipitation. The smaller the influence of precipitation is, the longer the lag time will be ref. [37].

By further analyzing the Nanyang Basin and the Jianghan Plain at low altitudes, it was found that the two regions were significantly different even if their altitudes were similar (Figure 6(a, c, e, and f) and Table 5). In the Nanyang Basin, the vegetation NDVI’s response to temperature and precipitation lagged by 2 and 1 months, respectively, and the significant positive correlated areas were 28.09 and 32.37%, respectively. However, in the Jianghan Plain, the vegetative response to these two factors showed no obvious lag effect, and the significant positive correlated areas were 39.09 and 23.18%, respectively.

Table 5

Area proportion of correlation between monthly NDVI and climatic factors below 300 m

Climatic factors Time scale Regions Negative (%) Positive (%) Not significant (%)
Temperature 2 months ago Nanyang Basin 4.70 28.09 67.21
Jianghan Plain 17.48 13.69 68.84
1 month ago Nanyang Basin 17.55 2.44 80.00
Jianghan Plain 4.98 16.53 78.49
In the month Nanyang Basin 15.82 2.98 81.20
Jianghan Plain 4.64 39.09 56.27
Precipitation 2 months ago Nanyang Basin 9.40 6.50 84.10
Jianghan Plain 6.56 15.34 78.11
1 month ago Nanyang Basin 4.73 32.37 62.90
Jianghan Plain 12.58 13.84 73.58
In the month Nanyang Basin 31.63 4.76 63.62
Jianghan Plain 13.57 23.18 63.24

This may relate to the different crops in these two regions. The Nanyang Basin locates in the transition zone from subtropical one to temperate one. It is mainly planted with wheat, corn, and peanut. By comparison, Jianghan Plain, located in the subtropical zone, is mainly planted with rice and cotton. The crops’ phenological characteristics directly affect the vegetation response to climatic factors in the two regions. Moreover, the lag effect is not synchronous because of the agricultural management measures for different crops. This has been verified in the research conducted by Zhao et al. on the response of crop phenology to climate changes in North China [38].

3.2.2 Human activities

With the acceleration of urbanization, human activities have profoundly affected the vegetation cover change. In this article, the “climate-NDVI” model was built based on the data between 2001 and 2008 to predict the NDVIg value from 2009 to 2015. Then, this predicted value was compared to the measured value between 2009 and 2015 for the analysis. Then, the residual analysis method was adopted to calculate the impact of human activities on the vegetation cover, thus analyzing the variation trend of the influence of human activities.

First, after the correlation between NDVIg and climatic factors was analyzed, a “climate-NDVI” model was built based on the temperature 2 months ago (L1), the precipitation 1 month ago (L2), the average growing-season temperature (L3), and the average growing-season precipitation (L4), which were highly correlated with NDVIg. The regression equation for the NDVIg at different elevations in the mid-lower reaches of the Hanjiang River Basin is expressed as follows:

(6) NDVI g = 3.992 L 1 0.307 L 2 3.925 L 3 0.537 L 4 ( DEM 300 m ) 1.188 L 1 0.396 L 2 1.398 L 3 1.083 L 4 ( 300 < DEM 1 , 900 m ) 0.194 L 1 0.859 L 2 1.024 L 3 + 1.282 L 4 ( DEM > 1 , 900 m ) ,

In the above three equations, the adjusted determination coefficients were all above 0.5. The F statistical value was over 150, and its corresponding p value was less than 0.01. The multiple regression model passed the significance test.

The “climate-NDVI” model was used to simulate the NDVI in the mid-lower reaches of the Hanjiang River Basin from 2009 to 2015. Then, the simulated multi-year mean NDVIg was calculated (Figure 7(a)). According to the simulation results, the basin’s vegetation cover decreased from northwest to southeast, which was similar to the measured NDVIg value. Thereinto, the vegetation growth condition was optimal in the Shennongjia Forestry District, as well as western and northern mountainous areas, showing a simulated mean NDVIg of over 4.5. The second was the Nanyang Basin, showing a simulated mean NDVIg of 3.5–4.5. The vegetation condition was worst in Jianghan Plain, with a simulated mean NDVIg of less than 3.5.

Figure 7 (a) Spatial distribution of the average NDVIg-sim, (b) residuals spatial distribution, and (c) residuals variation trend.

Figure 7

(a) Spatial distribution of the average NDVIg-sim, (b) residuals spatial distribution, and (c) residuals variation trend.

A residual analysis was carried out based on the measured and simulated NDVIg values, thus analyzing the effect of human activities on the vegetation cover in the basin (Figure 7(b)). The NDVIg residual was 0.26 on average, indicating that human activities promoted the regional vegetation growth. The concrete manifestations were as follows: the NDVIg residual was generally less than 0 in areas around the Nanyang Basin, which meant that such activities inhibited the vegetation growth slightly. However, in the central areas of the Nanyang Basin, the residual fluctuated around 0, indicating that the vegetation growth was not affected by these activities. The NDVIg of the Jianghan Plain generally exceeded 0, especially in Jingmen, Zhongxiang, Jingshan, Tianmen, etc. Here, human activities showed a more significant promoting effect.

Further calculation of the residual variation trend is shwn in Figure 7(c), which showed that the slope for the change of the influence of human activities is 0.07/15a on average. In this figure, the positive value indicated that the influence of human activities continued to increase, while the negative value meant the influence weakened. It can be known by combining with the residual spatial distribution in Figure 7(b) that human activities had an inhibited effect on the vegetation growth in areas around the Nanyang Basin, but this effect decreased gradually. However, in most areas of the Jianghan Plain, such activities promoted the vegetation growth, and this promoting effect also decreased gradually.

4 Discussion

4.1 Vegetation cover change trend

During the study period, the vegetation cover had a fluctuated increasing trend in the basin, which was consistent with the variation trend in the whole Hanjaing River Basin [37]. The vegetation was improved most significantly in the Nanyang Basin and its surrounding areas. Relevant research indicated that this was associated with the strategy for the harmonious development of “new industrialization, new urbanization and agricultural modernization” as well as the project of returning farmland to forest and grassland in Henan [39,40]. In addition, vegetation cover in the southeastern of the Jianghan Plain showed a decreased trend, which may be related to the urban economic development, for it has been proved that the rapid GDP growth would reduce the vegetation cover [41].

4.2 Response of vegetation cover to climatic factors

On the interannual scale, the vegetation NDVIg was significantly negatively correlated with the average growing-season temperature in the western mountainous area of the Nanyang Basin, while it was positively correlated with the temperature in the east of the Nanyang Basin (Figure 5a). According to the previous study [42], the reason may be that higher temperatures accelerated the transpiration and evaporation of mountainous vegetation. Thus, the vegetation growth was restricted due to water shortage. However, in eastern agricultural regions, the increase of the temperature may prolong the frost-free season [43], thus extending the growing season and increasing the vegetation cover. In the Jianghan Plain, the NDVIg showed a weak correlation with the average growing-season temperature but showed a strong one with the mean precipitation (Figure 5(a and b)). However, Kong CF proposed that the vegetation growth was more strongly correlated with the annual average temperature than with the temperature [44]. Different temporal scales of studies may cause a discrepancy between the two conclusions. This article only calculated the annual growing-season data, while the previous results were obtained based on the full-year data.

On the intraannual scale, the vegetative response to temperature and precipitation lagged by 2 and 1 months, respectively, in the Nanyang Basin. There was no noticeable lag effect in the vegetative response to the two factors in the Jianghan Plain. However, the research carried out by Li showed that the vegetation response to the temperature had no lag effect in the Nanyang Basin and the Jianghan Plain, and the response to precipitation lagged by 20 days [45]. Different spatial-temporal scales of studies may cause the discrepancy.

4.3 Response of vegetation cover to human activities

Through investigation and evaluation, the vegetation growth in the Nanyang Basin and its surrounding areas at different stages may be mainly affected by hydraulic engineering construction, and the direct impact was from the water pipeline construction [46]. During the construction process, the surface vegetation would be directly eradicated in the operation area, resulting in the reduction of forest and grassland. But after the construction, the operation area gradually turned into bare land, and vegetation could be restored after a period of natural succession. Meanwhile, the water storage was increased due to the Danjiangkou Hydraulic Project after October 2005, which submerged the surrounding croplands, forest lands, and grasslands and had a significant effect on the vegetation cover [47,48,49]. Moreover, due to the ecological benefit brought by the hydraulic engineering construction [50,51], the regional and local climate formed, and the growth conditions of forest lands and grasslands became better [49]. Besides, the implementation of ecological engineering projects such as returning farmland to the forest reduced the negative effect that human activities had on the vegetation cover, and the vegetation cover increased significantly [52,53]. Therefore, the negative effect of human activities on the vegetation cover weakened gradually, and the vegetation cover increased significantly.

After 2009, the water source for the middle route of South-to-North Water Transfer witnessed mass immigration, which had an important impact on the regional vegetation cover changes [54]. On the one hand, in the migrating-out areas, due to the policy of returning farmland to forest, a large amount of farmland reclaimed from wasteland had been turned into forest again [55]. On the other hand, in the resettlement areas, large amounts of wood were used in house construction, resulting in a decrease in the forest area [56]. In addition, immigration had changed people’s production and lifestyles, especially the farming system and farming methods, which also had an impact on the vegetation cover changes [57].

Furthermore, the NDVIg residual generally exceeded 0 in the Jianghan Plain because the modern agricultural development, especially mechanization, had a positive effect on the regional vegetation. According to the data, the agricultural area decreased by 1451.58 km2 between 2000 and 2015 in the Jianghan Plain [58]. However, in the same period, the total annual NPP of regional farmlands tended to be stable on the whole [59], which, to a certain extent, indicated that the realistic productivity of farmlands was improved. Compared to the data about the mechanization level in this period, the total power of agricultural machinery increased from 5.72 × 106 to 1.76 × 107 kW, and the grain yield increased from 6.6 × 106 to 8.0 × 106 t [60]. Without the interference of planting structure adjustment and crop variety change, mechanization may be the key in promoting the vegetation growth in the Jianghan Plain.

5 Conclusion

With the mid-lower reaches of the Hanjiang River Basin as the study object, this article, based on the MODIS NDVI time-series data from 2001 to 2015, discussed the response of the vegetation cover change to climatic factors and human activities at different elevations under the pretext of hydraulic engineering construction and modern agricultural development. The following are the major conclusions:

  1. (1)

    The vegetation cover in the study area was “high in the west and north, lower on both sides of Hanjiang River, and lowest in the center and southeast.” It changed parabolically with increasing elevation: Vegetation cover increased below 1,900 m and then decreased steadily over 1,900 m.

  2. (2)

    Over the 15 years, vegetation cover in the basin showed an increasing trend, and the increased and decreased areas were 90.72 and 9.23%, respectively. The extremely significantly increased and decreased areas were 41.8 and 1.26%, respectively.

  3. (3)

    The response of vegetation cover to climatic factors varies greatly depending on increasing elevation. That is, the lag effect under the impact of temperature disappeared gradually, while it became more evident under the impact of precipitation.

  4. (4)

    Human activities promoted vegetation cover on the whole, with an average slope of 0.07/15a. Around the Nanyang Basin and the most parts of the Jianghan Plain, human activities played an inhibited and a promoted role, respectively, but both roles were gradually weakening.

Acknowledgments

The authors are grateful for the data provider of MOD13Q1 NDVI, National Meteorological Information Center and Chinese Academy of Sciences Resource and Environment Data Cloud Platform. This research was funded by CRSRI Open Research Program (Program SN: CKWV2016403/KY) and supported by the National Key Research Program (Program SN:2018YFC0506905).

    Author contributions: Y. Z. and J. F.: conceptualization; Z. L.: methodology; J. F., D. C., and L. S.: software; Y. Z., J. F., and D. C.: validation; J. H.: formal analysis; J. F.: data curation; Y. Z., T. M.: writing – original draft preparation; Y. Z. and J. F.: writing – review and editing; Y. Y.: funding acquisition.

    Conflict of interest: Authors state no conflict of interest.

References

[1] Zhang BY, Yu ZZ, Liu LY, Zhang ZY, Sun TT. Study on the vegetation coverage changes monitoring of Beijing city by remote sensing [in Chinese with English abstract]. Areal Res Dev. 2008;27(1):108–12. Search in Google Scholar

[2] Xia J, Yao WQ, Wang JH, Song XY. A study on the relationship between dynamic change of vegetation coverage and precipitation in Beijing’s mountainous areas during the last 20 years. Math Comput Model. 2011;54(3):1079–85. Search in Google Scholar

[3] Sun YL, Shan M, Pei XR, Zhang XK, Yang YL. Assessment of the impacts of climate change and human activities on vegetation cover change in the Haihe River basin, China. Phys Chem Earth. 2020;115(prepublish):102834. Search in Google Scholar

[4] Bao G, Qin Z, Bao Y, Zhou Y, Li W, Sanjjav A. NDVI-based long-term vegetation dynamics and its response to climatic change in the Mongolian Plateau. Remote Sens. 2014;6(9):8337–58. Search in Google Scholar

[5] Palmate SS, Pandey A, Kumar D, Pandey RP, Mishra SK. Climate change impact on forest cover and vegetation in Betwa Basin, India. Appl Water Sci. 2017;7(1):103–14. Search in Google Scholar

[6] Gu Z, Duan X, Shi Y, Li Y, Pan X. Spatiotemporal variation in vegetation coverage and its response to climatic factors in the Red River Basin, China. Ecol Indic. 2018;93:54–64. Search in Google Scholar

[7] Karlsen SR, Tolvanen A, Kubin E, Poikolainen J, Høgda KA, Johansen B, et al. MODIS-NDVI based mapping of the length of the growing season in northern Fennoscandia. Int J Appl Earth Obs. 2008;10:253–66. Search in Google Scholar

[8] Wang X, Piao S, Ciais P, Li J, Friedlingstein P, Koven C, et al. Spring temperature change and its implication in the change of vegetation growth in North America from 1982 to 2006. Proc Natl Acad Sci USA. 2011;108(4):1240–5. Search in Google Scholar

[9] Zhang YD, Zhang XH, Liu SR. Correlation analysis on normalized difference vegetation index (NDVI) of different vegetations and climatic factors in Southwest China. Chin J Appl Ecol. 2011;22(2):323–30. Search in Google Scholar

[10] Liu Y, Wang X, Guo M, Tani H, Matsuoka N, Matsumura S. Spatial and temporal relationships among NDVI, climate factors, and land cover changes in Northeast Asia from 1982 to 2009. GISci Remote Sens. 2011;48(3):371–93. Search in Google Scholar

[11] Hu MQ, Mao F, Sun H, Hou YY. Study of normalized difference vegetation index variation and its correlation with climate factors in the three-river-source region. Int J Appl Earth Obs. 2011;13(1):24–33. Search in Google Scholar

[12] Zhang Y, Liang W, Liao Z, Han Z, Xu X, Jiao R, et al. Effects of climate change on lake area and vegetation cover over the past 55 years in Northeast Inner Mongolia grassland, China. Theor Appl Climatol. 2019;138(1–2):13–25. Search in Google Scholar

[13] Tong S, Zhang J, Bao Y, Wurina, Terigele, Weilisi, et al. Spatial and temporal variations of vegetation cover and the relationships with climate factors in Inner Mongolia based on GIMMS NDVI3g data. J Arid Land. 2017;9(3):394–407. Search in Google Scholar

[14] Lu Q, Zhao D, Wu S, Dai E, Gao J. Using the NDVI to analyze trends and stability of grassland vegetation cover in Inner Mongolia. Theor Appl Climatol. 2019;135(3):1629–40. Search in Google Scholar

[15] Li S, Yang S, Liu X, Liu Y, Shi M. NDVI-based analysis on the influence of climate change and human activities on vegetation restoration in the Shaanxi-Gansu-Ningxia Region, Central China. Remote Sens. 2015;7(9):11163–82. Search in Google Scholar

[16] Xu KJ, Zeng HD, Ren J, Xie JS, Yang YS. Spatial and temporal variations in vegetation cover in an eroded region of subtropical red soil and its relationship with the impact of human activity [in Chinese with English abstract]. Acta Ecol Sin. 2016;36(21):6960–8. Search in Google Scholar

[17] Ma Q, Long Y, Jia X, Wang H, Li Y. Vegetation response to climatic variation and human activities on the Ordos Plateau from 2000 to 2016. Env Earth Sci. 2019;78(12):709. Search in Google Scholar

[18] Wei ZF, Wang DG, Zhang C, Liu X, Zhang H. Response of vegetation cover to climate change and human activities in Northwest China during 1999–2010 [in Chinese with English abstract]. J Desert Res. 2014;34(6):1665–70. Search in Google Scholar

[19] Chen T, Bao AM, He DM. Analyzing dynamic vegetation change and its driving factors in Panzhihua City using MODIS NDVI data [in Chinese with English abstract]. Resour Environ Yangtze Basin. 2018;27(8):1847–57. Search in Google Scholar

[20] Deng CH, Bai HY, Gao S, Liu RJ, Ma XP, Huang XY, et al. Spatial-temporal variation of the vegetation coverage in Qinling Mountains and its dual response to climate change and human activities [in Chinese with English abstract]. J Nat Resour. 2018;33(3):425–38. Search in Google Scholar

[21] Peng J, Liu Y, Shen H, Han Y, Pan Y. Vegetation coverage change and associated driving forces in mountain areas of Northwestern Yunnan, China using RS and GIS. Env Monit Assess. 2012;184(8):4787–98. Search in Google Scholar

[22] Shen LN, Jing Y, Sun YL, Yan YL, Lv HP. Spectrum characteristics of vegetation coverage change and analysis of the difference of topographic gradient in Haihe River Basin [in Chinese with English abstract]. J Tianjin Norm Univ, Nat Sci Ed. 2017;37(6):43–9. Search in Google Scholar

[23] Zhao T, Bai HY, Deng CH, Meng Q, Guo SZ, Qi GZ. Topographic differentiation effect on vegetation cover in the Qinling Mountains from 2000 to 2016 [in Chinese with English abstract]. Acta Ecol Sin. 2019;39(12):4499–4509. Search in Google Scholar

[24] Ma CH. Multi-scale vegetation cover changes in Shaanxi Hanjiang river basin [in Chinese with English abstract]. J Shaanxi Univ Technol (Nat Sci Ed). 2013;29:66–70. Search in Google Scholar

[25] Bin X, Cheng ZG, Wang JF, Jin LY. Seasonal variation of NDVI and altitude dependent characteristics in Qinling-Daba Mountains in recent 17 years [in Chinese with English abstract]. J Chengdu Univ Inf Technol. 2019;34(3):302–7. Search in Google Scholar

[26] Luan JK, Liu DF, Liu H, Huang Q, Lin M, Li GB. Analysis of the affecting factors of vegetation index change in the upper reach of Hanjiang River basin [in Chinese with English abstract]. J North China Univ Water Resour Electr Power (Nat Sci Ed). 2019;40(1):46–54. Search in Google Scholar

[27] Yang Q, Liu DF, Meng XM, Huang Q, Lin M. Vegetation index change in the upper reaches of Han River and its attribution analysis [in Chinese with English abstract]. South–North Water Transf Water Sci Technol. 2019;17(4):138–48. Search in Google Scholar

[28] Du Y, Zhao Y, Cai SM, Ren XH. Effect on industrial development in middle-and-down reaches of Hanjiang River owing to middl-line project of transferring water from south to north [in Chinese with English abstract]. Resour Environ Yangtze Basin. 1999;8(2):135–9. Search in Google Scholar

[29] Wang YM, Zhou XD, Ning DT. Effect on climate in middle-and-down reaches of Hanjiang River owing to middl-line project of transferring water from south to north [in Chinese with English abstract]. J Soil Water Conserv. 2001;21(6):40–2. Search in Google Scholar

[30] Zeng Q. An analysis of water environment dynamics and causes in middle-and-down reaches of Hanjiang River [in Chinese with English abstract]. Resour Environ Dev. 2008;1:9–13. Search in Google Scholar

[31] Zheng XB, Luo YX, Yu F, Wei-Min KA, Juan CH. Comparisons of spatial interpolation methods for agro-climate factors in complex mountain areas of Southwest China [in Chinese with English abstract]. Chin J Agrometeorol. 2008;29(4):458–62. Search in Google Scholar

[32] Liu YT, Chen AB. Optimal spatial interpolation method for precipitation data in China’s river basins [in Chinese with English abstract]. Yangtze River. 2019;50(4):100–5. Search in Google Scholar

[33] Wang YM. On the vegetation regionalization of Hubei Province (A) [in Chinese with English abstract]. J Wuhan Bot Res. 1985;3(1):61–73. Search in Google Scholar

[34] Wang YM. On the vegetation regionalization of Hubei Province (B) [in Chinese with English abstract]. J Wuhan Bot Res. 1985;3(5):165–76. Search in Google Scholar

[35] Shen GH, Li JS. Analysis of “2010.7” extraordinary rainstorm and flood in the upper Hanjiang River [in Chinese with English abstract]. Sci Tech Inf Gansu. 2011;40:66–7, 70. Search in Google Scholar

[36] Staudinger MD, Mills KE, Stamieszkin K, Record NR, Hudak CA, Allyn A, et al. Phenological spatial characteristics of Hubei Province based on time series vegetation index [in Chinese with English abstract]. Resour Environ Yangtze Basin. 2019;28(7):1583–9. Search in Google Scholar

[37] Li XY. Research on spatial and temporal changes of vegetation cover and response to climate factors in Hanjiang Basin [in Chinese with English abstract]. J Northwest Norm Univ (Nat Sci). 2013;49(2):102–6. Search in Google Scholar

[38] Zhao YX, Xiao DP, Bai HZ, Tao F. Research progress on the response and adaptation of crop phenology to climate change in China [in Chinese with English abstract]. Prog Geogr. 2019;38(2):224–35. Search in Google Scholar

[39] Wang XC, Liu WK, Yang HJ, Wang SD, Zhang HB. Spatiotemporal change of fractional vegetation coverage and its landscape pattern in Henan Province [in Chinese with English abstract]. Bulletin of Soil and Water Conservation. 2015;35(6):241–7, 254. Search in Google Scholar

[40] Yu YY, Wang YP, Qin F, Guo PF. Vegetation index change and its response to climate in Henan Province [in Chinese with English abstract]. J For Environ. 2019;39(3):280–6. Search in Google Scholar

[41] Han FY, Wang TM, Sun Y. Monitoring and analyzing dynamic change of vegetation coverage in Wuhan area based on Landsat remote sensing data [in Chinese with English abstract]. Geomat Spat Inf Technol. 2019;42(4):90–2. Search in Google Scholar

[42] Mao D, Wang Z, Luo L, Ren C. Integrating AVHRR and MODIS data to monitor NDVI changes and their relationship with climatic parameters in Northeast China. Int J Appl Earth Obs. 2012;18:528–36. Search in Google Scholar

[43] Yan JH, Fu H, Zhou HS, Xu XC. Variations of frost-free period and its impact on grain yields in Henan Province during 1961–2013. Agric Sci Technol. 2015;16(8):1588–91. Search in Google Scholar

[44] Kong CF, Yue YC, Xu K. Spatial-temporal variation of vegetation net primary productivity and its response to climate change in Jianghan Plain from 2000 to 2015. Resour [in Chinese with English abstract]. Environ Yangtze Basin. 2019;28(6):1460–9. Search in Google Scholar

[45] Li XY. Time-delayed correlation analysis between vegetation cover changes and climatic factors in Hanjiang River Basin [in Chinese with English abstract]. Bull Soil Water Conserv. 2013;33:268–70. Search in Google Scholar

[46] Dong MT, Tang MF, Li SY, Cao H, Deng H, Wu G. Effects of water transfer project construction on vegetation patterns [in Chinese with English abstract]. Acta Ecol Sin. 2016;36(20):6656–63. Search in Google Scholar

[47] Liu H, Yin J, Chen J, Chen XL. Dynamic change and cause of water area in Danjiangkou reservoir based on remote sensing image [in Chinese with English abstract]. Resour Environ Yangtze Basin. 2016;25:1759–66. Search in Google Scholar

[48] Yue H, Liu H. Water area extraction and dynamic monitoring of the Danjiangkou reservoir using MNDWI. Env Sustain Dev. 2016;41(5):201–4. Search in Google Scholar

[49] Qin YL, Wen L, Wei PF, Li QG, Li XM. Analysis of dynamic change of landscape pattern in the surrounding areas of Danjiangkou reservoir [in Chinese with English abstract]. Yellow River. 2019;41(4):69–73. Search in Google Scholar

[50] Guan GM, Zhang LS, Wu ZY. The role of Danjiangkou dam heightening works in flood control planning of the Hanjiang river [in Chinese with English abstract]. Yangtze River. 2006;37:53–4. Search in Google Scholar

[51] Liu S. Bringing Danjiangkou reservoir into full play in flood control and drought defying [in Chinese with English abstract]. Hubei Water Power. 2008;5:47–50, 54. Search in Google Scholar

[52] Yang YY. A study on China’s south-to-north water transfers and harmonious social economic development in central region [in Chinese with English abstract]. J Zhongnan Univ Econ Law. 2007;3:3–9, 142. Search in Google Scholar

[53] Wang GH. The thoughts of establishing the Danjiangkou reservoir water source conservation and local fiscal problems [in Chinese with English abstract]. Financ Dev. 2008;10:30–3. Search in Google Scholar

[54] Gao WW, Zeng Y, Liu Y, Yi H, Wu B, Ju H. Forest dynamic monitoring by remote sensing from 2000 to 2015 in the water source area of the south-to-north water diversion project [in Chinese with English abstract]. Sci Silvae Sin. 2019;55(4):97–107. Search in Google Scholar

[55] Lu B, Xie HQ, Wang QL, Li PX, Sun JY. Research on vegetation change in Nanyang based on Landsat data [in Chinese with English abstract]. Sci Technol Inf. 2020;18(1):71–2. Search in Google Scholar

[56] Qin KL. The impact of the south-to-north water transfer project on the regional ecological environment [in Chinese with English abstract]. Water Resour South North Water Diversion. 2010;8:8–9. Search in Google Scholar

[57] Tu RM. Building a new city on flat land: a basic model for the development of large resettlement sites for external relocation and centralized resettlement – taking the Hubei Province resettlement of the middle route project of south-to-north water transfer project as an example [in Chinese with English abstract]. Jianghan Tribune. 2013;11:70–6. Search in Google Scholar

[58] China National Knowledge Infrastructure. Available online: http://data.cnki.net/ (accessed on 15 August 2019). Search in Google Scholar

[59] Duan H, Hui-Min Y, Hong C, Xiao-meng G, Qi-hui S. Research on spatiotemporal characteristics of farmland ecosystem NPP in Jianghan Plain from 2000 to 2015 [in Chinese with English abstract]. J Nat Resour. 2020;35(4):845–56. Search in Google Scholar

[60] Yu RL, Ma ZL. Spatio-temporal patterns and influencing factors of grain production in Jianghan Plain during 2000–2015 [in Chinese with English abstract]. World Reg Stud. 2018;27(3):99–108. Search in Google Scholar

Received: 2020-12-11
Revised: 2021-04-27
Accepted: 2021-04-29
Published Online: 2021-06-09

© 2021 Yunjun Zhan et al., published by De Gruyter

This work is licensed under the Creative Commons Attribution 4.0 International License.