Available online at www.sciencedirect.

com

Procedia Environmental Sciences 10 (2011) 2709 – 2717

2011 3rd International Conference on Environmental

Science and Information Application Technology (ESIAT 2011)

Analysis of Land Use Dynamic Change and Its Impact on the

Water Environment in Yunnan Plateau Lake Area
— A Case Study of the Dianchi Lake Drainage Area
Zhang Hong1, aˈLi Hailin1ˈChen Zhen1
1
School of Urban Management and Resources Environment, Yunnan University of Finance and Economics, Kunming 650221, China
a
Package_l@126.com

Abstract

With the Dianchi Lake drainage area as the subject of research, the TM remote sensing images of the four time points
taken in 1974, 1988, 1998 and 2008 as data basis and the combination of image processing, geographic information
system and the relevant metrology software, the present paper not only analyzed the land use dynamic change and
tendency, but also studied the mechanism of the impact of land use change on the water environment in the studied area
with Markov model, the results of which showed that: ķ The land use situation in the Dianchi Lake basin in the past 30
years has significantly changed, in which the changes of the arable land, construction-use land and other similar
changes are extremely great, while the changes of the water area and forest are not so obvious. ĸ The results of Markov
model forecast showed that the land for construction in the studied area further increases in 2015-2030 while the
woodland, water area and other types of land further decreases, but the rate of decrease is gradually becoming smaller
and the cultivated land area continues to decrease as well. Ĺ The water quality near the Dianchi Lake drainage area is
affected by the type and change of land use.
© 2011 Published by Elsevier Ltd. Selection and/or peer-review under responsibility of Conference ESIAT2011
© 2011 Published
Organization by Elsevier
Committee. Ltd. Selection and/or peer-review under responsibility of [name organizer]
Open access under CC BY-NC-ND license.

Keywords: land use, ArcGis, Markov, Dianchi Lake basin, water environment

Introduction

Using remote sensing (RS) and geographic information system (GIS) makes it possible to objectively grasp
the structure of land use and the form of driving land use change, understand the feature of land use change
and anticipate the situation of land use structure in the future, which is conducive to guiding people to
rationally and efficiently make use of land and providing an important decision-making for the regional
sustainable development. Furthermore, in the discussion of the reason for the water environmental change,
the change of land use type is an essential factor. If the relationship between land use type and water quality

1878-0296 © 2011 Published by Elsevier Ltd. Selection and/or peer-review under responsibility of Conference ESIAT2011 Organization Committee.
Open access under CC BY-NC-ND license.
doi:10.1016/j.proenv.2011.09.421
2710 Zhang Hong et al. / Procedia Environmental Sciences 10 (2011) 2709 – 2717

is made clear, a rational pollution prevention measure can be determined through an appropriate land use
plan to solve the issue of water environmental pollution and protect water quality.

Studied area and research methodology

Studied area.Located in the central part of Yunnan-Guizhou Plateau and along the divide between the
three river systems of the Yangtze, the Pearl and the Red River, the Dianchi Lake drainage area covers the
four districts of Panlong, Wuhua, Xishan and Guandu in Kunming as well as some city and county areas
like Chenggong, Jinning and Songming. Its geographic coordinate is 102e29Ą~103e01Ą east longitude and 24e
29Ą~25e28Ą north latitude with the matching image measured drainage area of 2,840.7km2. The entire
drainage area is a lake basin long in north and south but narrow in east and west. The landform can be
divided into three sorts of mountainous hill, aggradation plain and the Dianchi Lake water area. The
Dianchi Lake drainage area belongs to north subtropical humid monsoon climate with an annual average
temperature of 14.9ć, a multiple-year average rainfall of 931.8mm and a whole-year frost-free period of 227
days.

Research methodology and data source

Research methodology. Based on the need as well as the form and attribute of land use in the studied area,
the land use is divided into 5 big categories according to Class One Classification: arable land, woodland,
construction-use land, water area and other land type. Both the Edras remote sensing image processing
software and the ArcGis geographic information system software were used to calibrate, match and
enhance the Landsat-5(TM) remote sensing satellite images at four time points taken in 1974, 1988, 1998
and 2008. The resolution of Landsat-5(TM) original data is 30m. Through the GPS field investigation, the
visual interpretation of remote sensing image was obtained and the images of each period were interpreted
with human computer interactive interpretation technology to generate the distribution situation of land use
at four time points. The ArcMap module in the ArcGis was used to conduct a map overlapping analysis, the
transition matrix of land use spatial dynamic change was established according to Markov model and 2
indexes of land use area variation rate and land use dynamic state were used to quantitatively analyze the
land use structural change in the studied area, in which the unit root test and cointegration test were adopted
to analyze the simulation of land use and water environment change in the drainage area.
Data source .The data in the studied area are from the Landsat-5(TM) remote sensing satellite images
taken in 1974, 1988, 1998 and 2008. The TM_Band_4(R), TM_Band_3(G) and TM_Band_2(B) band
resultant images were adopted and the Human computer interactive interpretation was obtained through the
geometric rectification, Edras remote sensing image processing software and ArcGis geographic
information system software.

Land use/cover change

Status quo of land use.On the basis of the previously obtained data, the area statistical function in the
ArcGis geographic information software was used to get the area and occupation percentage of 5 big land
types at the four studying time points (see Table 1).
 Zhang Hong et al. / Procedia Environmental Sciences 10 (2011) 2709 – 2717 2711

Table 1 Status quo of land use structure in 1974-2008

1974 1988 1998 2008

Land type Area Occupation Area Occupation Area Occupation Area Occupation
(km2) (%) (km2) (%) (km2) (%) (km2) (%)
Arable land 492.2 17.3 442.2 15.6 403.2 14.2 265.1 9.3
Land for
86.8 3.1 129.8 4.6 181.9 6.4 397.1 14.0
construction
1272.
Woodland 44.8 1256.7 44.2 1252.6 44.1 1252.8 44.1
3
Water area 320.4 11.3 323.2 11.4 322.9 11.4 320.7 11.3
Other types 669.0 23.6 688.7 24.2 680.1 23.9 605.0 21.3
It can be seen from the table that the woodland at the four studied time points has the highest occupation
percentage with the maximum of 44.8% in 1974 and the next one is water area with no big change, which is
greatly related to the geomorphic feature in the drainage area. The area of arable land and construction-use
land changed greatly: the percentage of arable land was decreased from 17.3% in 1974 to 9.3% in 2008,
while the percentage of land for construction was increased from 3.1% in 1974 to 14.0% in 2008.
Change of the type of land use.The analysis of land use change mainly is mainly to conduct a spatial
overlay with the land use status quo spatial information at the four time points under the ArcGis geographic
information software. With Markov transition matrix and through statistical analysis, calculation and
preparation, the land use status quo map and dynamic evolution transition matrix at each time point are
obtained.
Table 2 Transition matrix of the type of land use in 1974-1988 Unit: km2
Land for
1974\1988 Woodland Arable land Other type Water area Total
construction
Land for
83.18 0.00 0.82 2.71 0.05 86.76
construction
Woodland 0.00 1255.05 0.02 17.24 0.00 1272.30
Arable land 36.78 1.41 430.84 17.71 5.45 492.18
Other type 9.63 0.24 7.70 649.69 1.73 669.00
Water area 0.23 0.05 2.79 1.40 315.99 320.45
Total 129.82 1256.74 442.16 688.74 323.22 2840.68

Table 3 Transition matrix of the type of land use in 1988-1998 Unit: km2
Land for
1988 \1998 Woodland Arable and Other type Water area Total
construction
Land for
106.97 0.00 12.54 9.83 0.49 129.82
onstruction
Woodland 0.95 1252.28 2.30 0.19 1.02 1256.74
Arable land 57.24 0.01 376.61 6.59 1.70 442.16
Other type 16.43 0.09 8.14 660.27 3.82 688.74
Water area 0.34 0.24 3.62 3.18 315.84 323.22
Total 181.94 1252.62 403.20 680.06 322.86 2840.68
2712 Zhang Hong et al. / Procedia Environmental Sciences 10 (2011) 2709 – 2717

Table 4 Transition matrix of the type of land use in 1998-2008 Unit: km2
Land for
1998\2008 Woodland Arable and Other ype Water area Total
construction
Land for
167.94 0.19 8.12 5.22 0.47 181.94
construction
Woodland 0.47 1251.74 0.22 0.16 0.04 1252.62
Arable land 128.98 0.27 224.06 44.99 4.89 403.20
Other type 97.43 0.51 28.36 550.66 3.09 680.06
Water area 2.26 0.10 4.37 3.96 312.16 322.86
Total 397.08 1252.81 265.13 605.01 320.66 2840.68
Taking Table 2 for example in the obtained transition matrix tables, the vertical lines with data represent
the types of land use in 1974 and the horizontal lines represent the types of land use in 1988. The data
represent the area of the types of land use in 1974 changing to various types of land use in 1988. It can be
easily seen from each transition matrix that the area changed from arable land to construction-use land in
1974-2008 is 212.86 km2, of which the area changed in 1998-2008 is 128.98 km2; the area of land for
construction changed from other types of land is also large, i.e., 102.86 km2, and the area of water and
woodland has a tendency of decrease, but the area decreased is very small and the integral change is not
great either.

Analysis of land use dynamic degree and area variation rate

Land use dynamic degree.Land use dynamic degree indicates an annual variation rate of the area of
land use type, which is an important index of land use change. It can quantitatively describe the land use
change rate of a region and its formula is as follows:

Where DC is land use dynamic degree, At1 represents the area of a land use type at t1 and At2 represents
the area of this land use type at t2.
According to the index calculation formula, the dynamic degrees of land use type change in the four
periods were calculated as shown in Table 6.

Table 6 Land use dynamic degree of the Dianchi Lake drainage area (%/year)
Year Arable land Land for construction Woodland Waterarea Othertypes
1974-1988 -0.73 3.54 -0.09 0.06 0.21
1988-1998 -0.88 4.01 -0.03 -0.01 -0.12
1998-2008 -3.43 11.83 0.00 -0.07 -1.10
The results suggest that in each study period of the studied area, the variation direction of land for
construction is always positive with a manifestation of area increasing and the maximum annual variation
rate up to 11.83%. The variation direction of other four types is basically negative with a manifestation of
total area decreasing and the maximum annual variation rate of arable land up to -3.43%.
Land use area variation rate.The area variation rate of land use refers to the specific value of area
differential value to the area in the initial period for a land use type in two different periods. Its calculation
formula is as follows:
 Zhang Hong et al. / Procedia Environmental Sciences 10 (2011) 2709 – 2717 2713

Where C is the area variation rate of land use, At1 represents the area of a land use type at t1 and At2
represents the area of this land use type at t2.
According to the index calculation formula, the area variation rate of land use in the four periods were
calculated as shown in Table 7.
Table 7 The area variation rate of land use in the Dianchi Lake drainage area (%)

Arable Land for

Year Woodland Water area Other types
land construction
1974-1988 -10.16 49.54 -1.23 0.87 2.94

1988-1998 -8.82 40.14 -0.33 -0.09 -1.25

1998-2008 -34.25 118.31 0.02 -0.68 -11.04
The results suggest that in each study period of the studied area, the land use variation rates of arable
land and land for construction are the highest, in which the rate of land for construction in 1998-2008 is
118.31%, other types become second and the variation rates of woodland and water area are the lowest.

Land use forecast based on Markov model

On the basis of Markov model forecast, the transition probability of each period in Matlab was calculated
and the relevant result was forecast compared with the actual value. Based on the actual situation, we
adopted the 1988 in this period of 1988-2008 as the initial situation year and took the transition probability
in this period as the vector of transition probability. By comparing the land use structure of 1998 and 2008
obtained from the forecast with the actual land use structure of 1998 and 2008, we found that the analog
value is very close to the actual value and the relative error average value in the two periods is within 0.05%.
It can be seen that it is feasible to use Markov process to predict the land use structure in the Dianchi Lake
drainage area. The predicted results are as follows:

Table 8 Markov chain prediction of the area of various land use types
˄n=27, 32, 37, 42˅ km2
Arable
Year Land for construction Woodland Water area Other types
land
2015 417.00 1251.54 268.45 584.50 319.19
2020 456.70 1250.60 247.68 567.36 318.34
2025 493.18 1249.67 229.44 550.92 317.48
2030 526.75 1248.74 213.43 535.17 316.59

Land use and water environment change in the drainage area

Land use is a most direct action of human beings to natural environment, so any improper use will cause a
great number of untreated materials such as nitrate and phosphate to discharge into the drainage area and
water body, further polluting the drainage area and water environment. Based on the previous research on
the land use change in the Dianchi Lake drainage area, we hereby made use of the water quality data
obtained from 1994 to 2008 and considered the data of land use found in the relevant years, analyzed and
obtained the time and spatial change of water environment of each year, and further analyzed the impact of
land use change in the Dianchi Lake drainage area on the water environment.
2714 Zhang Hong et al. / Procedia Environmental Sciences 10 (2011) 2709 – 2717

Simulation of land use and water environment change in the drainage area

Water environment is mainly composed of surface water and underground water, but the present project
only analyzed the surface water of the Dianchi Lake. The appraisal standard for the water environment of
the Dianchi Lake refers to the Environmental Quality Standards for Surface Water of the People’s Republic
of China (GB3838-2002) issued by the State Environmental Protection Administration. The routine items
of water quality supervision include: the indexes of temperature, electric conductivity, pH, dissolved
oxygen (DO), total phosphorus (TP), total nitrogen (TN), ammonia nitrogen (NH3-N), permanganate index,
biochemical oxygen demand (BOD), volatile phenol, oil substance, heavy metal, bacteria, etc. Based on the
actual condition of this research, we didn’t analyze all the indexes. By visiting and investigating the
relevant departments, we adopted a time-series method to analyze the TN, TP, BOD and COD obtained in
1994-2008 and study the effect of the land use change on the environment in the Dianchi Lake drainage
area.
Unit root test .The land in the Dianchi Lake drainage area is divided into six big categories—land for
urban, town, village, factory and mining construction, arable land, forest land, grassland, wetland and
gardening land. The proportion of each land type in the drainage area was respectively taken to conduct a
unit root test to the TN, TP, BOD and COD in the drainage area from 1994 to 2008 as well as the time series
of each land type. The present paper adopted the ADF method to conduct a unit root test with the results
shown in Table 9.

Table 9 ADF test of each variable

Original series
Variable
Form ADF value 5% statistical quantity 10% statistical quantity
TP (C,T,3) -3.59 -3.88 -3.39
TN (C,T,3) -4.67 -3.79 -3.34
BOD (C,T,3) -0.90 -3.79 -3.34
COD (C,T,3) -2.61 -3.88 -3.39
GENG (C,T,3) -1.62 -3.79 -3.34
CHENG (C,T,3) -1.85 -3.79 -3.34
LIN (C,T,3) -2.69 -3.93 -3.42
MU (C,T,3) -0.92 -3.79 -3.34
YUAN (C,T,3) -5.60 -3.88 -3.39
SHI (C,T,3) -1.68 -3.79 -3.34

First-order differential sequence

Form ADF value 5% 10% statistical
(C,T,3) -3.30 -3.93 -3.42
(C,T,3)
(C,T,3) -3.34 -3.93 -3.42
(C,T,3) -1.67 -3.93 -3.42
(C,T,3) -3.28 -3.83 -3.36
(C,T,3) -3.28 -3.83 -3.36
(C,T,3) -2.44 -3.83 -3.36
(C,T,3) -2.56 -3.93 -3.42
(C,T,3)
(C,T,3) -3.72 -3.83 -3.36
 Zhang Hong et al. / Procedia Environmental Sciences 10 (2011) 2709 – 2717 2715

Second-order differential sequence Conclusion

Form ADF 5% statistical 10% statistical
(C,T,3) -5.35 -4.11 -3.52 I(2)
(C,T,3) I(0)
(C,T,3) -4.67 -4.11 -3.52 I(2)
(C,T,3) -7.79 -4.11 -3.52 I(2)
(C,T,3) -3.93 -3.93 -3.42 I(2)
(C,T,3) -4.05 -3.93 -3.42 I(2)
(C,T,3) -4.98 -3.88 -3.39 I(2)
(C,T,3) -4.49 -3.38 -3.39 I(2)
(C,T,3) I(0)
(C,T,3) -5.65 -3.88 -3.39 I(2)
It can be seen form the above table that the sequence of total nitrogen and gardening land is 0-order
integration and other sequences are 2-order integration.
Cointegration test.Johansen’s inspection method was used to conduct a cointegration test to get the
cointegration relationship exiting between the total phosphorus, COD and BOD, and land for urban, town,
village, factory and mining construction, arable land, forest land, grassland, wetland and gardening land
with the following relations:

TP=15.31CHENG+10.47GENG-156.74MU-0.66 (1)
(5.06) (7.12) (-2.79) (-0.48)
AdjR2=0.81ˈF=20.37ˈDW=2.21

BOD=543.79CHENG-272.49GENG+160.88LIN-194.24 (2)
(4.40) (-4.34) (2.91) (-3.77)
AdjR2=0.62ˈF=6.96ˈDW=1.88

COD=173.88CHENG-100.75GENG-4920.59SHI+41.88 (3)
(1.40) (-1.45) (1.44) (1.15)
AdjR2=0.39ˈF=3.70ˈDW=2.29
The correlativity existing between TP and BOD under 5% obvious level and BOD under 20% obvious
level can be obtained. It can be concluded that both the land for urban, town, village, factory and mining
construction and the arable land have significantly increased the discharge of total phosphorus
concentration in the drainage area while the grassland has obviously decreased the effect of total
phosphorus concentration in the drainage area; both the land for urban, town, village, factory and mining
construction and the forest land have significantly increased the pollution of BOD in the drainage area
while the arable land has effectively decreased the pollution of BOD in the drainage area; the land for urban,
town, village, factory and mining construction has greatly increased the discharge of COD in the drainage
area while the arable land has effectively decreased the discharge of COD in the drainage area. Since total
nitrogen and gardening land can not form a cointegration relationship with other variables, it is thus not
considered hereby.

Forecast of land use and water environment change in the drainage area
2716 Zhang Hong et al. / Procedia Environmental Sciences 10 (2011) 2709 – 2717

Markov transition matrix was adopted to respectively predict the situation of land use in the four periods of
2015, 2020, 2025 and 2030 and the models (1), (2) and (3) were adopted to separately forecast the TP,
BOD and COD in the Dianchi Lake drainage area with the following results obtained:

Table 10 Forecast of water pollution in the four periods in 2015, 2020, 2025 and 2030 Unit: mg/1

Year TP BOD COD

2015 0.084 2.744 6.25
2020 0.162 3.157 6.08
2025 0.303 2.341 7.02
2030 0.454 3.562 7.96
It can be seen from the above prediction that with the years increasing, the indexes of water pollution
have a tendency of continuous deterioration, for which the main reason is that the proportion of the land for
urban, town, village, factory and mining construction has continuously increased while the proportion of
arable land, forest land, grassland, wetland, etc. in the drainage area has continuously decreased with the
years increment, which has led to the continuous deterioration of environmental condition.

Conclusions

According to the classified results of remote sensing data in the four periods of 1974-2008 as well as under
the support of ArcGis geographic information software, the land use structure was analyzed. And the
indexes of land use dynamic degree and land use area variation rate were used to reveal the time law of land
use/cover change in the Dianchi Lake drainage area in recent 30 years. The results suggest that the type of
land use in the Dianchi Lake drainage area has a remarkable change.
In view of the time span of the entire study, a significant change has taken place in the situation of land
use in the Dianchi Lake drainage area in recent 30 years, in which a great change has taken place to the
arable land, land for construction and other land types, while the change to the water area and woodland is
not obvious.
From the three time intervals of the study there is a greatest change of land use in the studying period of
1998-2008. Similarly, the most obvious is the decrease of arable land and the rapid increase of land for
construction.
The results of Markov model forecast indicate that the area of land for construction is further increased
while the area of woodland, water area and other land types is further decreased, but the rate of decrease is
gradually becoming smaller and the area of arable land is also continuously decreasing in the studied area in
2015-2030.
The change of land use type has an obvious effect on the change of Dianchi Lake water environment. A
conclusion can be drawn from the analysis of the existing data that the water quality of the Dianchi Lake
has a tendency of gradual deterioration in time and it shows in space that the water quality drops rapidly in
the area where human activities disturb most, in which the behavior of the grass sea of the Dianchi Lake is
extremely apparent. Especially at the beginning of the rapid development 1990s, the land for construction
sharply increased, the swift extension of the downtown in Kunming brought about some problems, the
agricultural activities intensified and a great amount of fertilizer was used in the production, which had a
great effect on the deterioration of Dianchi Lake water quality. If no great attention is paid to, the land for
urban, town, village, factory and mining construction is extended randomly and the decrease of wetland,
arable land, grassland, etc. is neglected, the ecology in the drainage area will be unavoidably led to
continuously deteriorate, the ecological capacity will be made unable to adapt to the situation of the
drainage area furthermore and finally the ecosystem in the drainage area will be led to blow up.
Therefore, efforts must be made to raise the consciousness of environmental protection, establish a
concept of environmental protection, strengthen the construction of wetland in the area around the lake,
 Zhang Hong et al. / Procedia Environmental Sciences 10 (2011) 2709 – 2717 2717

rationally control the scale of urban, town, village, factory and mining land use, improve the utilization of
land for construction and enhance the reclamation of farmland. On the basis of keeping the average yield of
land, the land use should be rationally planned to obtain a sustainable development between the land use
and water quality safety in the drainage area.

Acknoledgement

This research was supported by the project of National Natural Science Foundation of China—Study on the
Mode and Managerial Mechanism of Yunnan Plateau Lake Urban Intensive Land Use under Ecological
Perspective (No.: 70863014)

References
[1] DU Na and WANG Jijun. Agriculture Land Use/Land Cover Change of Xiannangou Watershed [J]. Journal of Soil and Water
Conservation Research, 2009,16(6): 171-178.]
[2] HE Ruizhen, ZHANG Jingdong and HU Yanhong. Dynamic change and forecast of land use in Zhengzhou based on RS and
GIS [J]. Journal of Gansu Agricultural University, 2009, 44(4): 126-131]
[3] [ CAO Jingjing, WU Jing and LI Chunbin. The forecast of land use and land cover change tendency based on Markov process
in Suzhou District Jiuquan City Gansu Province [J]. Territory & Natural Resources Study, 2008, (1): 45-47
[4] WANG Haitao XU Zhenwen and ZHANG Guoqing. Land Use/Cover Change—A Case Study of Jidong County [J]. Journal
of Changchun Normal University, 2008, 28 (6): 51-54
[5] GUO Biyun and ZHANG Guangjun. Study on land use and land cover change based on remote sensing and GIS—A case
study of Duolun County in Inner Mongolia [J]. Agricultural Research In The Arid Areas , 2009, 27(4): 240-244
[6] YANG Kun, CHANG Qingrui, GAO Xin, et al. Status analysis of county-level land use based on GIS [J]. Bulletin of Soil and
Water Conservation, 2009, 29 (6): 111-114
[7] MEI Rong and HAI Jun. Markov simulation of urban land use spatial evolution process based on GIS [J]. Inner Mongolia
Science Technology and Economy, 2009, 24 (202): 56-57
[8] XU Xinhui, FU Kaidao and XU Yanhong. Study on land use change of Fadou town in Xichou county based on GIS/RS [J].
Journal of Yunnan Geographic Environment Research, 2009, 21(5): 52-62
[9] PENG Jian and CAI Yunlong. Study of Karst Land Use/Cover Change Under the Framework of LUCC—Present and
Prospect [J]. China Land Science, 2006, 20(5): 48-53
[10] LIU Qiong, OU Minghao and PENG Xiaoying. The forecast of land use structure based on Markov process [J]. Journal of
Nanjing Agricultural University, 2005, 28(3): 107-112