BY 4.0 license Open Access Published by De Gruyter Open Access November 19, 2020

Analysis and prediction of LUCC change in Huang-Huai-Hai river basin

Jianwei Wang, Kun Wang, Tianling Qin, Hanjiang Nie, Zhenyu Lv, Fang Liu, Xiaoqing Shi and Yong Hu
From the journal Open Geosciences

Abstract

Land use/cover change plays an important role in human development and environmental health and stability. Markov chain and a future land use simulation model were used to predict future change and simulate the spatial distribution of land use in the Huang-Huai-Hai river basin. The results show that cultivated land and grassland are the main land-use types in the basin, accounting for about 40% and 30%, respectively. The area of cultivated land decreased and artificial surfaces increased from 1980 to 2010. The degree of dynamic change of land use after the 1990s was greater than that before the 1990s. There is a high probability of exchange among cultivate land, forest and grassland. The area of forest decreased before 2000 and increased after 2000. Under the three emission scenarios (RCP2.6, RCP4.5, and RCP8.5) of IPSL-CM5A-LR climate model, the area of cultivated land will decrease and that of grassland will increase in the upstream area while it will decrease in the downstream area. The above methods and rules will be of great help to future land use planning.

1 Introduction

Land use and land cover change (LUCC) research plays an important part in the global change research [1,2,3]. Human alterations in land cover as a result of the use of land-based natural resources not only have local and regional impacts but can also have important effects at the global level [4]. Under the adjustment of institutional factors, people’s response to economic opportunities promoted changes in land cover [5]. Inundation and resettlement will cause LUCC [6]. Economic, technological, institutional, and location factors are often considered as potential drivers of LUCC [7]. The accelerated industrialization and urbanization following economic reforms and population increases have greatly affected LUCC through the increase of built-up areas and urban sprawl [8]. The fragmentation of cultivated land can be observed due to construction on the land of the countryside [9]. In addition to human activities, climate change also affects LUCC [10]. Climate change will cause changes in temperature and precipitation, affecting vegetation growth and the temporal and spatial distribution of water resources. Grasslands are most responsive to extreme drought events followed by forests and desert vegetation [11]. After experiencing 8 years of extreme drought events in the San Juan National Forest, Colorado, USA, the average dieback rate of Populus tremuloide gradually increased from 15% to 27% in July 2011 and to 38% in July 2012 [12]. In the long-term, repeated drought-induced mortality events, along with unsuccessful recruitment of the dominant tree species in forests may lead to a vegetation succession process where trees would be replaced by understory species (such as drought-tolerant shrub) [13]. Comprehensively consider the impact of human activities and climate change on LUCC and predict future LUCC to provide a basis for our future land planning.

There are many methods for LUCC research. The spatial calculating analysis model is used to analyze LUCC [14]. Remote sensing and Geographic Information Systems (GIS) are also used in LUCC research [15]. Change-vector analysis is a valuable technique for LUCC detection [16]. A system dynamic model can be used to model LUCC [17]. CA (Cellular Automata)-Markov model is commonly used for LUCC prediction [18,19]. Land use simulation model Conversion of Land Use and its Effect at Small regional extent utilizes the spatial dominant dynamic simulation model established by Wageningen University [20,21]. The future land use simulation model (FLUS [The FLUS model could be downloaded from http://www.geosimulation.cn/flus.html]) developed by Liu et al. [22] could be used for multiple LUCC dynamic simulation. The FLUS model considers the impact of climate conditions, geography, and socioeconomic factors on LUCC and simulates the spatial distribution of various land-use type under different land-use conversion targets. The LUCC can provides information for our better decision on the development and utilization of natural resources [23,24]. According to research goals of this article, we choose GIS spatial analysis ang FLUS model as research tools.

The Huang-Huai-Hai river basin accounts for 15% of China’s total land area. It is an important economic center and food production base in China and has an important strategic position in China’s development [25]. The plains in the eastern part of the basin are China’s agricultural production bases. The area of arable land and grain output account for 20.4% and 23.6% of the country’s total [26]. Since the twenty-first century, the economic development of the Huang-Huai-Hai river basin has driven the process of urbanization and great changes have taken place in land-use types. The Huang-Huai-Hai region is located in China’s three major climatic regions, including monsoon climate, dry climate, and highland climate. The climate change in recent years has an impact on the region, especially for the vegetation. LUCC research is of great significance to the future sustainable development of the Huang-Huai-Hai river basin. Hence, the law of LUCC and the spatial distribution of LUCC under future climate conditions will be studied in this paper.

2 Models and methods

Based on historical land use data, the Raster Calculator in the ArcGIS spatial analysis tool is used to calculate relevant indicators and transfer matrices of LUCC to analyze the laws of historical changes.

The other part is the spatial distribution of LUCC under future climate conditions. Historical land-use data, socioeconomic data, and climate data are input into the FLUS model for verification. The future climate conditions are input into the verified model to simulate the spatial distribution of LUCC under future climate conditions. The article process and structure are shown in Figure 1.

Figure 1 Article structure diagram.

Figure 1

Article structure diagram.

2.1 Metric indexes of LUCC

Land use of a dynamic degree is used to analyze the dynamic change of land use. The single land-use dynamic degree refers to the quantity change of a certain land-use type [27] within a certain period. The formula [28] is as follows:

(1)K=UbUaUa×1T×100%

where K is the dynamic degree of one land-use type during the study period; Ua and Ub are the area of a certain land-use type at the beginning and end time of the study period, respectively; and T is the length of the study period. When T is set to the year, K is the annual change rate of a land-use type during the study period.

The comprehensive land-use dynamics degree indicates the overall dynamics of land use in a certain area. The formula [28] is as follows:

(2)LC=i=1nLCi=j2i=1nLUi×1T×100%

where LC is the rate of change in land use and reflects the degree of change in the amount of land use; LUi is the area of land-use type i at the beginning of the period; LCij is the absolute value of the area converted from land use of type i to that of non-type i during the period; and T is the length of the study period. When T is set to the year, LC is the annual change rate of a land-use type during the study period.

2.2 Markov model

The Markov model determines the changing trend of each state of the system through the initial probability of different states of the systems and the transition probability between states, to achieve the purpose of predicting the future. Different land-use types in a certain area are mutually convertible, and the average conversion state change of land-use structure is relatively stable over a period of time. Therefore, the dynamic evolution of land use has the property of the Markov model. The Markov model has been widely used in the LUCC research [29]. With the support of the ArcGIS software, the land use raster of two periods are overlapped and the corresponding property in the overlay map database PAT.DBF is extracted. According to the probability distribution matrix of initial S(0) and the transition probability of the nth stage P(n), the future area of land use could be calculated by computer simulation. The Markov simulation model S(n) [30] is as follows:

(3)S(n)=S(n1)×P(1)=S(0)×P(n)

2.3 FLUS model and verification

Land-use change models are usually used to detect where the change occurred or will potentially occur [31]. Traditional methods can provide static analysis of the LUCC for the fixed beginning and end date [32]. And the LUCC model also can help people better understand the mechanism of social, economic, and physical variables’ impacts on LUCC [33]. Through knowing the factors contributing to the change, the model provided a probabilistic prediction of where the change may occur [34,35]. LUCC models are also used to assess the cumulative impact of land-use change and develop future scenarios [36]. It helps and supports land-use planning and decision-making [37]. The FLUS model (Figure 2) that interactively integrates top-down system dynamics (SD) with a bottom-up CA model could be used for multiple LUCC dynamic simulation [22]. The SD model is used to project the land-use scenario demands under various socio-economic and environmental driving factors. A self-adaptive inertia and competition mechanism is developed within the CA model to process the complex competitions and interactions among the different land-use types. This model has been successfully applied in the simulation of land use and land cover change in China, as well as on the global scale [38]. In the model simulation process, not all land-use types can be transformed into each other. According to the natural evolution of land resources and the actual situation of human activities, we set up a cost matrix (Table 1). For example, artificial surface cannot be converted into other types of land, cultivated land can be converted into other types of land, and water bodies can be converted into other types of land. In this way, the simulation is more realistic.

Figure 2 FLUS model frame diagram.

Figure 2

FLUS model frame diagram.

Table 1

Cost matrix in the FLUS model

Cultivated landForestGrasslandWater bodiesArtificial surfacesUnutilized land
Cultivated land111111
Forest110010
Grassland111010
Water bodies000100
Artificial surfaces000010
Unutilized land111111

Note: 1 means conversion is allowed; 0 means conversion is not allowed.

Receiver operation characteristic (ROC) curve analysis can obtain multiple pairs of true (false) positive rate values [39]. With the false positive rate on the abscissa and the true positive rate on the ordinate, the drawn curve is called the ROC curve, and the area under the ROC curve is the AUC value (area under the curve), which reflects the value of the diagnostic test. It is generally believed that when the AUC value is 0.5–0.7, the diagnostic value is low; when the AUC value is 0.7–0.9, the diagnostic value is moderate; when it is greater than 0.9, the diagnostic value is good. Cohen’s kappa coefficient [40,41] is a measure of agreement between two variables. The ROC curve and kappa are used to test the simulation effect of the FLUS model.

3 Study area and datasets

3.1 Study area

The geographical range of the Huang-Huai-Hai area is 95°53′∼122°60′E, 32°10′∼43°N, with a total area of about 1.43 million km2. The Huang-Huai-Hai region is located in China’s three major climatic regions, including monsoon climate, dry climate, and highland climate. And it includes three first-level water resource regions, namely, the Yellow river basin, the Huai river basin, and the Hai river basin. The geographical location of the Huang-Huai-Hai area is shown in Figure 3.

Figure 3 Location of the Huang-Huai-Hai river basin.

Figure 3

Location of the Huang-Huai-Hai river basin.

3.2 Data datasets

The data used in this article are land use data, socioeconomic data, terrain data, and historical climate data. The time resolution of the data is years, and the spatial resolution is 1 km × 1 km. Future climate data (liner interpolated and bias-corrected) obtained from Global Climate Model (IPSL-CM5A-LR) provided by ISI-MIP (The Inter-Sectoral Impact Model Intercomparison Project) [42]. The details are listed in Table 2.

Table 2

List of data used in this study

CategoryDataYearData resourceDescription
Land useLand use1980\1990\2000\2010http://www.resdc.cnDownload from the data resource and the size is 1 km × 1 km
Socioeconomic dataPopulation2005
GDP
TerrainDEM2000http://www.gscloud.cnSame as above
AspectCalculated from DEM
SlopeCalculated from DEM
RoadsRoadhttp://www.openstreetmap.orgConvert shp file to 1 km × 1 km raster file
ClimateTemperature2001–2010http://www.resdc.cnDownload from the data resource and the size is 1 km × 1 km
Precipitation
Future climateTemperature2011–2030http://www.isi-mip.orgResampled 0.5° × 0.5° raster into 1 km × 1 km

4 Results

4.1 Law of land use conversion

4.1.1 Dynamic change of land use

Land cover proportions in the Huang-Huai-Hai river basin during 1980–2010 were calculated as listed in Table 3. The land use types in the Huang-Huai-Hai river basin were mainly composed of cultivated land, forest, and grassland in the past 20 years. And the area of the three land types accounted for more than 85% of the total area. From the perspective of the dynamics of land use change, the cultivated land has been decreasing, the artificial surfaces have been increasing, and the forest area has been reduced before increasing in the three periods (year 1980–1990, year 1990–2000, year 2000–2010). Overall, the degree land-use change after the 1990s was significantly higher than that before the 1990s.

Table 3

Land cover proportion and dynamics in Huang-Huai-Hai river basin during 1980–2010 unit: %

Land-use typeCover proportionDynamics
19801990200020101980–19901990–20002000–2010
Cultivated land41.9741.9441.7941.10−0.06−0.36−1.67
Forest13.1013.0813.0613.25−0.15−0.141.46
Grassland31.9831.9931.7731.620.04−0.67−0.49
Water bodies2.572.402.442.49−6.451.611.96
Artificial surfaces5.175.395.836.404.118.319.66
Unutilized land5.225.205.105.15−0.29−1.991.01
Total1001001001000.35*0.83*0.87*

Note: * indicates a comprehensive land-use dynamics degree.

4.1.2 Net change, exchange change, and the total change in land-use types

Based on the analysis of statistical results in Table 4, during the three periods, artificial land was dominated by expansion and cultivated land was dominated by the conversion of spatial location. Water bodies and grassland were degraded and were accompanied by some spatial conversion. In particular, the forest area increased during 2000–2010 and its net change accounted for 76% of the total change.

Table 4

Summary of land-use change in the Huang-Huai-Hai river basin during 1980–2010 unit: %

Statistical variablesYear rangeLand-use typesTotal
Cultivated landForestGrasslandWater bodiesArtificial surfacesUnutilized land
Increase1980–19900.190.020.110.090.210.070.70
1990–20000.470.120.320.150.450.151.66
2000–20100.220.250.320.160.580.211.73
Loss1980–19900.210.040.100.260.000.090.70
1990–20000.620.130.530.110.010.261.66
2000–20100.920.060.470.110.010.161.73
Net change1980–19900.020.020.010.170.210.020.45
1990–20000.150.010.210.040.440.110.96
2000–20100.70.190.150.050.570.051.71
Exchange change1980–19900.380.040.20.1800.140.94
1990–20000.940.240.640.220.020.32.36
2000–20100.440.120.640.220.020.321.76
Total change1980–19900.40.060.210.350.210.161.39
1990–20001.090.250.850.260.460.413.32
2000–20101.140.310.790.270.590.373.47

4.1.3 Land-use conversion

The land-use changes during 1980–2010 were analyzed from the perspective of quantitative. In order to analyze the mutual transformation of land use, the land use conversion matrix was calculated based on the ArcGIS platform. The results from 1980 to 1990 are listed in Table 5. The loss of cultivated land was mainly converted into artificial surfaces, and the degradation of water bodies was mainly converted into cultivated land. The land-use conversion results from 1990 to 2000 are listed in Table 6. The loss of cultivated land was mainly converted into artificial surfaces and also converted into water bodies and grassland. At the same time, the water bodies and grassland were mainly converted into cultivated land. The land-use conversion results from 2000 to 2010 are listed in Table 7. The law of cultivated land conversion was still the same, but the change of forest was more obvious than the previous two periods.

Table 5

The conversion matrix of land use from 1980 to 1990

Year 1980 (km2)Year 1990 (km2)
Cultivated landForestGrasslandWater bodiesArtificial surfacesUnutilized landTotal
Cultivated land6,00,639108888511,966496,03,701
Forest1151,87,77844792211,88,372
Grassland6351204,58,5382241672744,59,958
Water bodies1,5374874633,24665673436,967
Artificial surfaces1803274,382074,405
Unutilized land4073730925027373,81075,086
Total6,03,3511,88,0914,60,13134,58277,46674,868
Table 6

The conversion matrix of land use from 1990 to 2000

Year 1990 (km2)Year 2000 (km2)
Cultivated landForestGrasslandWater bodiesArtificial surfacesUnutilized landTotal
Cultivated land5,94,4204511,2021,0225,8823746,03,351
Forest6351,86,1591,1222892551,88,091
Grassland4,2501,0324,52,4663933301,6604,60,131
Water bodies1,1455218233,0325511634,582
Artificial surfaces59221377,390077,466
Unutilized land7001232,06965015371,17374,868
Total6,01,2091,87,8194,57,04335,13883,90273,378
Table 7

The conversion matrix of land use from 2000 to 2010

Year 2000 (km2)Year 2010 (km2)
Cultivated landForestGrasslandWater bodiesArtificial surfacesUnutilized landTotal
Cultivated land5,87,9771,8082,7471,4656,8593536,01,209
Forest1221,86,935312101298511,87,819
Grassland1,7541,6334,50,2724546312,2994,57,043
Water bodies7616022533,52028828435,138
Artificial surfaces6818455583,708883,902
Unutilized land4791071,21323122771,12173,378
Total5,91,1611,90,5614,54,81435,82692,01174,116

4.2 Prediction and spatial distribution simulation of land-use change

4.2.1 Prediction of land-use change

The prediction steps are as follows: the first step is the generation of the initial state matrix. Select the land use types in 2000 as the initial matrix as follows (unit: km2).

(4)S(0)=CultivatedlandforestGrasslandwaterbodiesartificialsurfacesuntilizedland=6,01,2091,87,8194,57,04335,138 83,09273,378

The land-use type area transition matrix of the initial state (Table 6) was obtained from 2000 to 2010. Thus, the initial state transition probability matrix (unit: %) from 2000 to 2010 (Table 8) and the annual average transition probability matrix were calculated to predict the land-use areas in 2030.

Table 8

The initial state transition probability matrix from 2000 to 2010 (n = 0)

20002010
Cultivated landForestGrasslandWater bodiesArtificial surfacesUnutilized land
Cultivated land97.800.300.460.241.140.06
Forest0.0699.530.170.050.160.03
Grassland0.380.3698.520.100.140.50
Water bodies2.170.170.6495.400.820.81
Artificial surfaces0.080.020.050.0799.770.01
Unutilized land0.650.151.650.310.3196.92

The future land-use area of change trend was calculated with the formula (3), formula (4), and Table 7. The results of 2,030 are listed in Table 9.

Table 9

Land use prediction results in 2030

YearArea (km2)
Cultivated landForestGrasslandWater bodiesArtificial surfacesUnutilized land
20305,71,7801,95,9054,50,39437,0571,07,85975,495

4.2.2 Spatial distribution simulation of land-use change

4.2.2.1 Model establishment and verification

Based on the land use of 2000, the geographical and socio-economic factors are considered to simulate the land use distribution in 2010 with the FLUS model. And the result was verified with actual data of 2010.

The simulation result of spatial land use and the actual one of 2010 are shown in Figure 4, with three small regions to display more details of the simulated results. The figure shows that the simulated pattern is well correlated with the actual. To quantitatively validate the simulated result, the confusion matrix of simulated results and the actual land use was calculated (Table 10). Then, the overall accuracy and Cohen’s Kappa coefficient for all land-use types were calculated. In addition, ROC curves of different land-use types in the simulation results were drawn, as shown in Figure 5. The results showed that, except for the poor simulation of unutilized land, the simulation results of the other five types were very good. Therefore, the model can be used for later predation.

Figure 4 The simulated land-use pattern and the actual land-use pattern in 2010.

Figure 4

The simulated land-use pattern and the actual land-use pattern in 2010.

Table 10

Confusion matrix of the predicted land-use pattern versus the actual pattern in 2010

Land use typesActual land use in 2010Total
Cultivated landForestGrasslandWater bodiesArtificial surfacesUnutilized land
Cultivated land55,87580488260764913258,949
Forest76217,39871046332218,971
Grassland90662343,736896049245,906
Water bodies71347432,71424453,586
Artificial surfaces6562644328,26969,033
Unutilized land489436122176,7717,403
Total58,96018,90745,8513,6109,0527,4681,43,848

Note: Kappa coefficient = 0.91, overall accuracy = 0.93.

Figure 5 ROC curves of different land-use types based on simulation results in 2010.

Figure 5

ROC curves of different land-use types based on simulation results in 2010.

4.2.2.2 Prediction of land use in future climate scenarios

Based on the previous step, the result of Markov’s prediction was used as the target for future land use and was as the input into the model. IPSL-CM5A-LR is comparatively better among all five climate models in this study area [41]. The spatial distribution of land use in 2030 was simulated under different climate scenarios (RCP2.6, RCP4.5, and RCP8.5). RCP2.6, RCP4.5, and RCP8.5 represent three scenarios of low, medium, and high greenhouse gas emissions, respectively, and represent possible future climate changes. The analysis under the conditions of the three climatic models makes our research on land-use changes under the future climatic conditions more comprehensive. The results are shown in Figure 6.

Figure 6 (a) Simulation result under RCP2.6, (b) simulation result under RCP4.5, and (c) simulation result under RCP8.5.

Figure 6

(a) Simulation result under RCP2.6, (b) simulation result under RCP4.5, and (c) simulation result under RCP8.5.

According to the prediction results shown in Table 9, the area of cultivated land will decrease largely in 2030 and that of the artificial surface will increase correspondingly. Based on the prediction results of 2030, the spatial distribution of different land-use types in three different scenarios is shown in Figure 4. In the three scenarios, the spatial distribution of artificial surfaces and water bodies are the same. However, there are some differences in the spatial distribution of cultivated land, grassland, and forest. Huang-Huai-Hai river basin includes three first-level sub-watershed (Figure 7) and seventeen second-level sub-watershed (Figure 8). The areas with a great difference in the spatial distribution of land use are shown in Figure 9. The specific results are shown in Figure 10 and the meaning of basin number is listed in Table 11.

Figure 7 First-level sub-watershed.

Figure 7

First-level sub-watershed.

Figure 8 Second-level sub-watershed.

Figure 8

Second-level sub-watershed.

Figure 9 Areas with different land use space simulation results.

Figure 9

Areas with different land use space simulation results.

Figure 10 Areas with different land use spatial distribution in different climate condition.

Figure 10

Areas with different land use spatial distribution in different climate condition.

Table 11

The meaning of the basin number

Basin numberMeaning
1_3Catchment area from Longyangxia to Lanzhou
1_5Catchment area above Longyangxia
1_6Catchment area from Sanmenxia to Huayuankou
2_1Catchment area of coastal rivers of Shangdong peninsula
2_5Yishusi river basin
3_3South of Haihe river basin

5 Discussion

5.1 Impact of historical climate and human activities on land use/cover

The comprehensive land-use dynamics degree of the 1980s, 1990s, and 2000s were 0.35, 0.83, and 0.87, respectively, as shown in Table 2. After the 1990s, under the influence of Economic Reform and open up, China’s economy developed rapidly. The transformation of the land by humans was huge. In particular, the acceleration of urbanization and rapid population growth have turned cultivated land, woodland, and grassland into resident land, and the water area had degraded. From a climate perspective, the temperature has increased by 1°C (1980–2010), and the increase in temperature is helpful for vegetation growth (photosynthesis) [42]. But the total amount of vegetation was still decreased. There may be two reasons for this. In the 1990s, people only paid attention to economic development and neglected the protection of vegetation. During this period, forest land and grassland decreased. In the twenty-first century, our country not only pays attention to social development but also pays attention to the protection of vegetation. Therefore, during the period from 2000 to 2010, vegetation was partially restored. The other is precipitation. Since 1980, the total amount of precipitation has not changed dramatically, but the temporal and spatial distribution of precipitation has been quite different. Especially after the 1990s, the occurrence of extreme hydrological events (droughts, floods) had an impact on the change of land cover [43,44]. The survival rate of trees is affected by extreme floods and the closer to the river, the lower the survival rate [45]. Drought reduces soil moisture content, and prolonged drought causes trees to die due to long-term lack of moisture [46]. The impact of repeated extreme precipitation events, such as repeated extreme drought events, even several years apart, may also have a greater negative impact on vegetation mortality [12]. As the North China Plain is strongly affected by the temperate monsoon, floods caused by heavy rains have had a huge negative impact on the Loess Plateau in the Yellow River Basin. More and more extreme precipitation has caused many environmental problems and ecological crises, such as soil erosion and land desertification [47]; for the Huai River Basin, since 1950, small floods have occurred every 2 years, and major floods have occurred every 5 years [48]; for the Haihe River Basin, although the annual precipitation is not large, most of the rainfall in the past 50 years has been reduced in the form of heavy rain, and the intensity of the heavy rain has been increasing [49]. In recent years, extreme hydrological events in all regions show an increasing trend. Floods mainly reduce vegetation through physical damage, and droughts reduce vegetation mainly through physiological damage. Based on the results of the net land-use change and exchange change and the land-use conversion matrix from 1980 to 2010, it is possible to know the conversion probability and direction of land use. Cultivated land, forest and grassland are frequently transformed into each other, while artificial surface is not easily transformed into other land types, and have strong expansion [50,51]. Under natural conditions, the probability of land use conversion is also one of the reasons for the difference in the spatial distribution of land use.

5.2 Impact of future climate on land use/cover

Compared with 2010, the spatial pattern of land use has changed in the future (Year 2030). We used the secondary watershed to make spatial statistics on the land-use pattern, and the results show that the distribution patterns of a forest, grassland, and cultivated land significantly change in the source and downstream region. The spatial distribution of areas with significant changes in the secondary watershed, shown in Figure 10. In the RCP 2.6 scenario, the area of cultivated land use in the upstream is reduced by more than 30%, while that is less than 10% in the RCP 4.5 and RCP 8.5 scenario. Whereas the area of grassland in downstream is reduced by more than 30% under the RCP 2.6 scenario but less than 10% under 10% under the RCP 4.5 and RCP 8.5 scenarios. According to the analysis of climatic conditions, the major factor for the different land-use patterns in upstream and downstream is precipitation. In the future, human activities will aim to minimize the disturbance of the natural environment and carry out green development [52]. Precipitation and temperature play the important roles in the growth of vegetation [53]. Therefore, precipitation and temperature will become the main factors affecting vegetation in the future. The risk of drought disaster risk in the Huang-Huai-Hai river basin is mainly reflected in the growing season precipitation in the entire region, the frequency of drought events, and the regulation and storage effect of the underlying surface on precipitation [54]. The higher risk areas are concentrated in the upper reaches of the basin. The risk of flood disaster risk in the Huang-Huai-Hai river basin is mainly reflected in the extreme precipitation in the entire basin and the regulation and storage effect of the underlying surface on precipitation [55]. The higher risk areas are concentrated in the lower reaches of the basin. The risk of drought and flood events in the basin in the future will have an impact on land use. At the same time, the pattern of land-use change in the upstream was different from that in the downstream. The upstream is dominated by the reduction of cultivated land, and the downstream is dominated by the reduction of grassland. That is because precipitation in the upstream will decrease 30 mm and that will increase about 40 mm in the downstream in the future. There is a large area of farmland upstream of the whole basin, which is affected not only by rainwater but also by artificial irrigation from the Yellow River. Therefore, with the participation of human activities, the area of cultivated land in the future may not be reduced much.

6 Conclusions

LUCC is affected by humans and climate change. By analyzing the historical changes in land use, the land-use changes after the 1990s had been dramatic, mainly reflected in the reduction of cultivated land and the expansion of artificial land. The social reason was mainly due to the increase in population, and the climatic reason was the increase in temperature. At the same time, this paper also studied the spatial distribution of land use in future climate scenarios. The changes in the upstream and downstream regions were the most obvious. The obvious changes in land use types were cultivated tland, grassland, and forest. The major factor was precipitation.

This article analyzes the causes of LUCC from the perspective of climatic conditions. The spatial distribution of land use in future climate conditions was also predicated. This result can also provide data support for future research on land use under future climate conditions.

Acknowledgments

This study was supported by the National Key Research and Development Project (Grant No. 2017YFA0605004; 2016YFA0601503) and the National Science Fund for Distinguished Young Scholars (Grant No. 51725905 and 51879275) and the National Natural Science Foundation of China (No. 51809002).

References

[1] Veldkamp A, Lambin EF. Predicting land-use change. Agric Ecosyst Environ. 2001;85(1):1–6. 10.1016/S0167-8809(01)00199-2. Search in Google Scholar

[2] Weinzettel J, Hertwich EG, Peters GP, Steen-Olsen K, Galli A. Affluence drives the global displacement of land use. Global Environ Change. 2013;23(2):433–8. 10.1016/j.gloenvcha.2012.12.010. Search in Google Scholar

[3] Jia K. Forest cover changes in the three-north shelter forest region of China during 1990 to 2005. J Environ Informat. 2015;26(2):126–37. 10.3808/jei.201400268. Search in Google Scholar

[4] Jamal M, Baba VV. Land-use and land-cover change. Science/research plan. Global Change Rep. 1995;43:669–79. 10.1177/001872679104401105. Search in Google Scholar

[5] Lambin EF, Turner BL, Geist HJ, Agbola SB, Angelsen A, Bruce JW, et al. The causes of land-use and land-cover change: moving beyond the myths. Global Environ Change. 2001;11(4):261–9. 10.1016/S0959-3780(01)00007-3. Search in Google Scholar

[6] Seeber C, Hartmann H, Xiang W, King L. Land use change and causes in the Xiangxi catchment, three Gorges area derived from multispectral data. J Earth Sci, 2010;21(6):846–55. 10.1007/s12583-010-0136-7. Search in Google Scholar

[7] van Vliet J, de Groot HLF, Rietveld P, Verburg PH. Manifestations and underlying drivers of agricultural land use change in Europe. Landscape Urban Plan, 2015;133:24–36. 10.1016/j.landurbplan.2014.09.001. Search in Google Scholar

[8] Long H, Tang G, Li X, Heilig GK. Socio-economic driving forces of land-use change in Kunshan, the Yangtze River Delta economic area of China. J Environ Manage, 2007;83(3):351–64. 10.1016/j.jenvman.2006.04.003. Search in Google Scholar

[9] Long HL, Li XB. Rural housing land transition in transect of the Yangtze River [in Chinese with English abstract]. Acta Geograph Sin. 2005;60:179–88. 10.3321/j.issn:0375-5444.2005.02.001. Search in Google Scholar

[10] Vadrevu K, Heinimann A, Gutman G, Justice C. Remote sensing of land use/cover changes in South and Southeast Asian Countries. Int J Dig Earth. 2019;12(10):1099–102. 10.1080/17538947.2019.1654274. Search in Google Scholar

[11] Xu HJ, Wang XP, Zhang XX. Decreased vegetation growth in response to summer drought in Central Asia from 2000 to 2012. Int J Appl Earth Observ Geoinform. 2016;52:390–402. 10.1016/j.jag.2016.07.010. Search in Google Scholar

[12] Anderegg WRL, Plavcová L, Anderegg LDL, Hacke UG, Berry JA, Field CB. Drought’s legacy: multiyear hydraulic deterioration underlies widespread aspen forest die-off and portends increased future risk. Global Change Biol. 2013;19(4):1188–96. 10.1111/gcb.12100. Search in Google Scholar

[13] Ibáñez B, Gómez-Aparicio L, Stoll P, Ávila JM, Pérez-Ramos IM, Marañón T. A neighborhood analysis of the consequences of Quercus suber decline for regeneration dynamics in Mediterranean forests. PLoS One. 2015;10(2):1–18. 10.1371/journal.pone.0117827. Search in Google Scholar

[14] Arowolo AO, Deng X. Land use/land cover change and statistical modelling of cultivated land change drivers in Nigeria. Region Environ Change. 2018;18(1):247–59. 10.1007/s10113-017-1186-5. Search in Google Scholar

[15] Reis S. Analyzing land use/land cover changes using remote sensing and GIS in Rize, North-East Turkey. Sensors. 2008;8(10):6188–202. 10.3390/s8106188. Search in Google Scholar

[16] Chen J, Gong P, He C, Pu R, Shi P. Land-use/land-cover change detection using improved change-vector analysis. Photogramm Eng Remote Sens. 2003;69(4):369–79. 10.14358/PERS.69.4.369. Search in Google Scholar

[17] Yu W, Zang S, Wu C, Liu W, Na X. Analyzing and modeling land use land cover change (LUCC) in the Daqing City, China. Appl Geograph. 2011;31(2):600–8. 10.1016/j.apgeog.2010.11.019. Search in Google Scholar

[18] Cao J, Wu J, Li C. The forecast of land use and land cover change tendency based on Markov process in Suzhou District Jiuquan City Gansu Province [in Chinese with English abstract]. Territory Nat Resour Study. 2008;1:45–7. Search in Google Scholar

[19] Yulianto F, Maulana T, Khomarudin MR. Analysis of the dynamics of land use change and its prediction based on the integration of remotely sensed data and CA-Markov model, in the upstream Citarum Watershed, West Java, Indonesia. Int J Dig Earth. 2019;12(10):1151–76. 10.1080/17538947.2018.1497098. Search in Google Scholar

[20] Verburg PH, Soepboer W, Veldkamp A, Limpiada R, Espaldon V, Mastura SSA. Modeling the spatial dynamics of regional land use: The CLUE-S model. Environ Manage. 2003;30(3):391–405. 10.1007/s00267-002-2630-x. Search in Google Scholar

[21] Verburg PH, Overmars KP. Combining top-down and bottom-up dynamics in land use modeling: exploring the future of abandoned farmlands in Europe with the Dyna-CLUE model. Landscape Ecol. 2009;24(9):1167–81. 10.1007/s10980-009-9355-7. Search in Google Scholar

[22] Liu X, Liang X, Li X, Xu X, Ou J, Chen Y, et al. A future land use simulation model (FLUS) for simulating multiple land use scenarios by coupling human and natural effects. Landscape Urban Plan. 2017;168:94–116. 10.1016/j.landurbplan.2017.09.019. Search in Google Scholar

[23] Pancost RD. Cities lead on climate change. Nat Geosci. 2016;9(4):264–6. 10.1038/ngeo2690. Search in Google Scholar

[24] Lu D, Mausel P, Brondízio E, Moran E. Change detection techniques. Int J Remote Sens. 2004;25(12):2365–401. 10.1080/0143116031000139863. Search in Google Scholar

[25] Shen F, Geng L, Qin F, Xu P. Analysis of water saving in received area in Huang-Huai-Hai watersheds and Eastern and Middle Lines of Water Transferring Project from South to North, China [in Chinese with English abstract]. Adv Water Sci. 2002;13(6):768–74. Search in Google Scholar

[26] Liu M, Wu J, Lv A, Zhao L, He B. The water stress of winter wheat in Huang-Huai-Hai Plain of China under rain-fed condition [in Chinese with English abstract]. Progr Geograph. 2010;29(4):427–32. 10.11820/dlkxjz.2010.04.006. Search in Google Scholar

[27] National Standards of the People’s Republic of China (GBT_21010_2017). Current and use classification, Beijing, Standards Press of China, 2017. (in Chinese with English abstract). Search in Google Scholar

[28] Huang C, Teng M, Zeng L, Zhou Z, Xiao W, Zhu J, et al. Long-term changes of land use/cover in the Three Gorges Reservoir Area of the Yangtze River, China [in Chinese with English abstract]. J Appl Ecol. 2018;29(05):1585–96. Search in Google Scholar

[29] Yang X, Zheng X, Chen R. A land use change model: integrating landscape pattern indexes and Markov-CA. Ecol Modell. 2014;283:1–7. 10.1016/j.ecolmodel.2014.03.011. Search in Google Scholar

[30] Zhang R, Tang C, Ma S, Yuan H, Gao L, Fan W. Using Markov chains to analyze changes in wetland trends in arid Yinchuan Plain, China. Math Comput Modell. 2011;54(3–4):924–30. 10.1016/j.mcm.2010.11.017. Search in Google Scholar

[31] Prestele R, Alexander P, Rounsevell MDA, Arneth A, Calvin K, Doelman J, et al. Hotspots of uncertainty in land-use and land-cover change projections: a global-scale model comparison. Global Change Biol. 2016;22(12):3967–83. 10.1111/gcb.13337. Search in Google Scholar

[32] Mondal MS, Sharma N, Garg PK, Kappas M. Statistical independence test and validation of CA Markov land use land cover (LULC) prediction results. Egyptian J Remote Sens Space Sci. 2016;19(2):259–72. 10.1016/j.ejrs.2016.08.001. Search in Google Scholar

[33] Zheng HW, Shen GQ, Wang H, Hong J. Simulating land use change in urban renewal areas: a case study in Hong Kong. Habitat Int, 2015;46:23–34. 10.1016/j.habitatint.2014.10.008. Search in Google Scholar

[34] Overmars KP, Koning DGHJ. Spatial autocorrelation in multiscale land use models. Ecol Modell. 2003;164:257–70. 10.1016/S0304-3800(03)00070-X. Search in Google Scholar

[35] Dan W, Wei H, Shuwen Z, Kun B, Bao X, Yi W, et al. Processes and prediction of land use/land cover changes (LUCC) driven by farm construction: the case of Naoli River Basin in Sanjiang Plain. Environ Earth Sci. 2015;73(8):4841–51. 10.1007/s12665-014-3765-9. Search in Google Scholar

[36] Verburg PH, Rounsevell MDA, Veldkamp A. Scenario-based studies of future land use in Europe. Agric Ecosyst Environ. 2006;114(1):1–6. 10.1016/j.agee.2005.11.023. Search in Google Scholar

[37] Guan D, Li H, Inohae T, Su W, Nagaie T, Hokao K. Modeling urban land use change by the integration of cellular automaton and Markov model. Ecol Modell. 2011;222(20):3761–72. 10.1016/j.ecolmodel.2011.09.009. Search in Google Scholar

[38] Li X, Chen G, Liu X, Liang X, Wang S, Chen Y, et al. A new global land-use and land-cover change product at a 1-km resolution for 2010 to 2100 based on human-environment interactions. Anna Am Assoc Geograph. 2017;107(5):1040–59. 10.1080/24694452.2017.1303357. Search in Google Scholar

[39] Pencina M, D’agostino RBJ, Vasan R. Evaluating the added predictive ability of a new marker: From area under the ROC curve to reclassification and beyond. Stat Med. 2008;27(2): 112–57. 10.1002/sim.2929. Search in Google Scholar

[40] Kyakuno T. Prediction of land use changes with Bayesian spatial modeling from the perspective of urban climate. Urban Climate. 2020;31:100569. 10.1016/j.uclim.2019.100569. Search in Google Scholar

[41] Pontius RG, Boersma W, Castella JC, Clarke K, de Nijs T, Dietzel C, et al. Comparing the input, output, and validation maps for several models of land change. Ann Region Sci. 2008;42(1):11–37. 10.1007/s00168-007-0138-2. Search in Google Scholar

[42] Wu X, Liu H, Li X, Ciais P, Babst F, Guo W, et al. Differentiating drought legacy effects on vegetation growth over the temperate northern hemisphere. Glob Chang Biol. 2017;24(1):504–16. 10.1111/gcb.13920. Search in Google Scholar

[43] Zeppel MJB, Wilks JV, Lewis JD. Impacts of extreme precipitation and seasonal changes in precipitation on plants. Biogeoences. 2014;11:3083–93. 10.5194/bg-11-3083-2014. Search in Google Scholar

[44] Matusick G, Ruthrof KX, Kala J, Brouwers NC, Breshears DD, Hardy GESJ. Chronic historical drought legacy exacerbates tree mortality and crown dieback during acute heat-wave-compounded drought. Environ Res Lett. 2018;13(9):1–14. 10.1088/1748-9326/aad8cb. Search in Google Scholar

[45] Tzeng HY, Wang W, Tseng YH, Chiu CA, Tsai ST. Tree mortality in response to typhoon-induced floods and mudslides is determined by tree species, size, and position in a riparian formosan gum forest in subtropical taiwan. PLoS One. 2018;13(1):e0190832. 10.1371/journal.pone.0190832. Search in Google Scholar

[46] Harper CW, Blair JM, Fay PA, Knapp AK, Carlisle JD. Increased rainfall variability and reduced rainfall amount decreases soil co2 flux in a grassland ecosystem. Global Change Biol. 2005;11(2):322–34. 10.1111/j.1365-2486.2005.00899.x. Search in Google Scholar

[47] He Z, He J. Temporal and spatial variation of extreme precipitation in the Yellow River Basin from 1960 to 2012 [in Chinese with English abstract]. Resour Sci. 2014;36(3):490–501. Search in Google Scholar

[48] She D, Xia J, Zhang Y, Du H. The trend analysis and statistical distribution of extreme rainfall events in the Huaihe River Basin in the past 50 years [in Chinese with English abstract]. Acta Geograph Sin. 2011;66(9):1200–10. Search in Google Scholar

[49] Pei Y, Wang J, Luo L. Analysis of effect of South-to-North Water Transfer Project on aquatic ecosystems of Haihe River Basin [in Chinese with English abstract]. Acta Ecol Sin. 2004;24(10):2115–23. Search in Google Scholar

[50] Pontius RG, Millones M. Death to Kappa: birth of quantity disagreement and allocation disagreement for accuracy assessment. Int J Remote Sens. 2011;32(15):4407–29. 10.1080/01431161.2011.552923. Search in Google Scholar

[51] Piani C, Weedon GP, Best M, Gomes SM, Viterbo P, Hagemann S, et al. Statistical bias correction of global simulated daily precipitation and temperature for the application of hydrological models. J Hydrol. 2010;395(3–4):199–215. 10.1016/j.jhydrol.2010.10.024. Search in Google Scholar

[52] Jung SH, Feng T. Government subsidies for green technology development under uncertainty. Eur J Operat Res. 2020;286(2):726–39. 10.1016/j.ejor.2020.03.047. Search in Google Scholar

[53] Li W, Du J, Li S, Zhou X, Duan Z, Li R, et al. The variation of vegetation productivity and its relationship to temperature and precipitation based on the GLASS-LAI of different African ecosystems from 1982 to 2013. Int J Biometeorol 2019;63(7):847–60. 10.1007/s00484-019-01698-x. Search in Google Scholar

[54] Wilby RL. Greenhouse hydrology. Progr Phys Geograph. 1995;19(3):351–69. 10.1177/030913339501900304. Search in Google Scholar

[55] Gázquez F, Calaforra JM, Fernández-Cortés Á. Flash flood events recorded by air temperature changes in caves: a case study in Covadura Cave (SE Spain). J Hydrol. 2016;541:136–45. 10.1016/j.jhydrol.2015.10.059. Search in Google Scholar

Received: 2020-03-28
Revised: 2020-10-22
Accepted: 2020-10-29
Published Online: 2020-11-19

© 2020 Jianwei Wang et al., published by De Gruyter

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