全 文 ：第 18 卷 第 1 期
Vol. 18 No. 1
草地学报
ACTA AGRESTIA SINICA
2010 年 1 月
Jan. 2010
Study on Land Use and Land Cover Change in Tianzhu County,
Northeast End of Qingha-i Tibet Plateau
WU Jing1, 2 , L I Chun-bin1, 2 , ZHANG De- gang 1* , PU Xiao-peng1
( 1. Pratacultural College, Gansu Ag ricultural University, Lanzhou, Gansu Province 730070, China; 2. C ol lege of Resources
and Environm ental S ciences, Gansu Agricu ltural U niver sity, Lanzh ou, Gan su Pr ovin ce 730070, C hina)
Abstract: T ianzhu County, located on the northeastern end of Q ilian Mountain and upriver of Shiyang Riv-
er, is a par t o f the three plateau convergence zone ( Qingha-i T ibet P lateau, Inner Mongolia Plateau, and
Loess Plateau) and a f rag ile ecolog ical environment. The changes of its ecosystem w ould impact on not on-
ly this county itself, but also the Shiy ang River w atershed. In addit ion to the pressing environmental prob-
lem, T ianzhu County is also a state- level poverty-st ricken county. T herefo re, the contradict ion betw een
development and env ir onmental pro tect ion in this area is prom inent . Based on Remote Sensing, Geographic
Informat ion System, and Global Positio n System ( 3S) , this paper studied the land use and land cover
change ( LUCC) in Tianzhu County f rom the ear ly 1990s to late 1990s. T he r esults ar e as fol low s: F irst ly,
farm land increased at the rate of 2. 34% per year w ith the main w ay of exploitat ion of w ood-g rass land and
unused land, and the per centage w as 80% and 17% respectiv ely . At the same t ime, the phenomenon of a-
bandoning farmland w as serious, in w hich 74% changed to unused land. Secondly, the 91% of increased
w ood-gr ass lands w as f rom unused land which happened mainly in the mountain areas above 4, 000 meter.
In conclusion, human agricultural act ivity disturbed the study area mo re intensiv ely ; the conversion o f un-
used lands ( fo rmed in the cold climate in high lat itude) in high mountains to w ood-g rass lands could be the
response to the w arming of the study area; the substant ial increase in cultiv ated lands in study area w ould
increase the w ater consumpt ion and then agg ravate the eco system collapse of dow nst ream M inqin Oases.
Key words: Land use and land cover change ( LUCC) ; T ianzhu County; 3S techno log y
青藏高原东北缘天祝县土地利用/土地覆盖变化研究
吴静1, 2, 李纯斌1, 2 , 张德罡1* , 蒲小鹏1
( 1. 甘肃农业大学草业科学学院, 甘肃 兰州 730070; 2. 甘肃农业大学资源与环境学院, 甘肃 兰州 730070)
摘要: 青藏高原东北缘的天祝县位于青藏高原、内蒙古高原和黄土高原三大高原交错带, 生态环境脆弱。因其处于
石羊河上游地区,所以其环境变化还可能影响整个石羊河流域。此外,天祝县还是个国家级贫困县, 经济发展与生
态保护的矛盾也比较突出。文章基于 3S 技术,研究了 20 世纪 90 年代土地利用/覆盖变化格局。结果表明 :耕地
以每年 2. 34%的速率递增,主要增长方式是开发林草地和未利用土地,同时耕地摞荒现象比较严重, 74%的耕地减
少方向是转为未利用土地; 91%的新增林草地来自未利用地, 而且, 大部分以分散形式发生在 4000 m 以上的高山
上部。因此,人类农耕活动对该区域的扰动比较剧烈; 高山上部未利用地向林草地的转化可能是对研究区域气候
变暖的响应;研究区域持续的耕地增长带来的水消耗增加可能会加剧下游民勤绿洲的生态环境恶化。
关键词:土地利用/土地覆盖变化;天祝县; 3S 技术
中图分类号: S159. 2 文献标识码: A 文章编号: 1007-0435( 2010) 01-0016-05
The global env ironment is changing due to hu-
man and natural causes. T he chal lenges of under-
standing global environmental change and of prov-i
ding the po licy-relevant know ledge base requir ed
by societ ies to deal w ith this change ar e acr oss
many aspects of scient if ic endeavors[ 1] . Land- use
收稿日期: 2009-06-08;修回日期: 2009- 10-19
基金项目:世界银行/全球环境基金项目( 052456CHA)
作者简介:吴静( 1973- ) ,女,四川道浮人,博士研究生,研究方向为草地生态, E-m ail: w ujing@ gsau. edu. cn; * 通讯作者 Au thor for corre-
spond ence, E- mail : z han gdg@ gsau. edu. cn
第 1期 吴静等:青藏高原东北缘天祝县土地利用/土地覆盖变化研究
and land-cover change ( LUCC) is the result of the
complex inter act ions of the physical environment
such as topogr aphy, so ils , and climate and human
driving fo rces including populat ion, technolog y, e-
conomic development , ur banizat ion, po lit ical
st ructure, social o rganizat ion, and culture, so it
has received much at tent ion in r ecent years. Sever-
al international o rganizat ions have called for com-
parativ e studies of LU CC in dif ferent reg ions.
LU CC has been an indispensable part of g lobal en-
v ir onmental change, and the cor e issues of g lobal
change study[ 1~ 4] .
The Qingha-i T ibet P lateau ( QT P) is a sens-i
t ive region of global environmental change, and the
essential study area for revealing the mechanism o f
global change[ 5, 6] . Warming o f the plateau is ev-i
denced by many obser vations, researches
[ 7~ 11]
show that w arming tendency has been higher than
that o f Northern H em isphere and reg ions at the
same lat itudes since the 1950s, and earlier than
that of east China. Especially the upw ar d t rend o f
temperature r ise af ter 1980s, the plateau becomes
more sensit ive in reflect ing global change, mean-
while, as vegetation coverage decreases, land cult-i
v at ion and desert ificat ion are exacer bated.
T ianzhu County is located on the nor theaster n
end o f QT P, w here the main terr ain is dominated
by Qilian M ountain, and a part o f the plateau. It
invo lves 6 o rigin branches of Shiyang River and is
the main part o f Shiyang valley ecosystem. T he
LU CC study o f this area is signif icant not only to
the r esearch o f plateau environmental change, but
also to the explanation of the desert ificat ion of Shiyang
River valley. But the relevant study is rare.
T his study surv eyed the LU CC o f T ianzhu
County during the 1990s, through integ rated tech-
niques of remote sensing, g lobal po sit ion sy stem
and geographic info rmat ion systems. We hope to
provide a case study o f the margin reg ion of the
QT P.
1 Materials and Methods
1. 1Study area
T ianzhu Tibetan Autonomous County, lo cated
on the eastern Gansu Corr idor of the Qilian M oun-
tain ( 10207E-10346E, 3631N-3755N ) , has
an ar ea of 7149. 8 km
2
, administer ing 8 tow ns, and
11 tow nships ( 2004 ) . T he study ar ea includes
T ianzhu County except fo r Danma T ownship and
Qilian Township.
T he study ar ea is a part of the three plateau
conver gence zone ( Qingha-i T ibet P lateau, Inner
M ongolia Plateau, and Loess Plateau) and belong s
to QTP according to natural div ision. The climate
is complicated, the annual sunshine hours varies
betw een 2500 to 2700 h, the annual mean tempe-
rature varies betw een - 8 and 4 , and the aver-
age annual precipitat ion varies between 265 mm
and 632 mm.
T ianzhu County is an important t radit ional an-
imal husbandry county o f China, as w el l as a state-
level poverty-str icken county, w hich increases the
dif f iculty of solv ing the environmental pr oblem.
Fig. 1 Location of study area
17
草地学报 第 18卷
1. 2 Data
Scanned mult-i temporal T M dig ital images,
taken by Landsat o f U SA in the June, 1991 and
August 1999, w ere the main data sources used for
ident ify ing land cover classes. T M images w er e
gained fr eely f rom the U niv ersity o f M ary land.
Other valuable materials applied in image in-
terpretat ion included the aerial pho tography cover-
ing Dacaigou ( a unit of study area) taken in 1976
w ith a nom inal scale of 1 18, 000, M SS dig ital
image with a spat ial r esolut ion of 80 m , topograph-
ic map at scale o f 1 50, 000, and other themat ic
maps, e. g . administr at ive map, geomo rpholo gic
map, climate map, so il map, etc. The Zoning o f
Agricultural and animal husbandry in Tianzhu
County was also used to support this study.
1. 3Methods
T M digital images w er e f irst combined to form
a single image mosaic: one for each year in the t ime
series, for the study ar ea covering tw o scenes and
then a subset by vecto r administ rative county
boundary using the digital image pr ocessing sy stem
ERDAS Imaging so ftw are and ENVI so ftw are. Im-
age interpretat ion and classificat ion w er e carried
out based on land use and land cover categories de-
signed fo r this project . A classif ication system w as
designed according to the criterion of creat ing land
use and land cover classificat ion sy stem of FAO,
taking account of the China land r esource classif-i
cat ion system , U . S. Nat ional Land Cover Data
classif icat ion sy stem, and the spat ial reso lut ion of
materials. The classificat ion definition of this stud-
y area is descr ibed in Table 1.
Table 1 Description for land use/ cover categories in the study area
Land use/ land cover category Descript ion
Farmland ( FL) L on g-term arable land
Wood-Grassland ( W-GL) Land mainly covered by coniferous fores t, b road-leaf for est , s hrub or g ras s
Water ( WL) River s, l ak es, an d ponds for aquacultu re
U nused land ( UL) Land diff icu lt t o use or unused
Con st ruct ion lan d ( CL) Land mainly for resident ial, t ran sportat ion or in dust rial u se
The interpretat ion and classificat ion w ere car-
r ied out by the w ay of interact ive inter pretation.
The classif icat ion results w ere g round trusted in
the field ( Fig . 1) and the average degr ee of inter-
pretat ion accuracy w as 81%.
Consequently , using the map algebra method,
an over lap tw o-period land use themat ic map in
GIS softw are Supermap w as created, and a draf t o f
the change areas and conversion areas w as dis-
played in a land-use conversion matrix , w hich pro-
moted the regional land use change monitoring ana-
lyses. Stat istical analyses w ere then carr ied out
and the land use dynam ics w ere displayed as a vec-
to r map.
Furthermore, the dynamic degr ee of land use
( DD) , repr esent ing the differ ent iat ion of land-use
dynamics, w as calculated to intensiv ely g ener al ize
the land-use pat tern. T he formula is as follow s:
K= ( ( U k- Ua) / Ua) / T 100%
Where K is the deg ree o f land use dynamics;
Ua r epresents areas o f i ( land use type) in the for-
mer period; ( Uk-U a) represents the net change o f
land use areas from the former period a to the
lat ter period k, T is the t ime scale.
2 Results
T he areas o f land use/ cover in dif ferent per-i
ods and the change condit ions are listed in Table 2.
T he quantitat iv e st ructure of land use/ co ver in dif-
ferent periods w ere: FL , W-GL, WL, U L, and
CL accounted for 14%, 63% , 1. 61%, 21%, and
0. 20% respect ively in 1991, 16% , 63% , 1. 60%,
19%, and 0. 15% respect ively in 1999, bo th had
been dominated by W-GL and U L. The land use/
cover quantitat iv e str ucture matched w ell, to a cer-
tain ex tent, w ith the socio-economic status in stud-
y area, animal husbandr y as the base, agr iculture
follow ed, and the urbanizat ion prog ress indist inc-
t ive.
Land use/ cover in the study area had changed
substantially over the study per iod. T able 3 also
shows the changes in the area since 1991. T he o-
18
第 1期 吴静等:青藏高原东北缘天祝县土地利用/土地覆盖变化研究
verall t rend of land use/ cover change in the study
ar ea from 1991 to 1999 w as a conspicuous decrease
of the ar ea of UL and an increase in FL area.
View ed from the dynamic degree of land use/
cover, CL obviously decreased at a rate o f 3. 43%
per year w hile U L decreased at a rate o f 1. 43% .
The area covered by w ater and W-GL remained re-l
at ively stable throughout the period, and if any-
thing , slight ly decr eased its ex tent .
T able 3 shows that signif icant conversion of
land-use categor ies centralized on W-GL, FL, and
U L. A larg e area o f W- GL was converted to FL
and U L during this t ime period and almost the
same large area of UL w as converted to W-GL sim-
ultaneously. The conver sion betw een FL and U L
was also obvious, w ith a larg e area of FL aban-
doned while larg e area of U L reclaimed to FL.
Table 2 Land use structure and dynamic change in study area in the 1990s
Land use/ cover
clas sifi cat ion
1991
Area
( hm2 )
Prop ort ion
( %)
1999
Area
( hm 2)
Proport ion
( % )
Ch ange
Area
( hm2)
Pr oport ion
( % )
FL 85915. 88 13. 67 102023. 17 16. 24 18. 75 2. 34
W-GL 396780. 30 63. 14 396448. 66 63. 09 - 0. 08 - 0. 01
WL 10110. 58 1. 61 10056. 67 1. 60 - 0. 53 - 0. 07
U L 134314. 76 21. 37 118938. 83 18. 93 - 11. 45 - 1. 43
CL 1259. 09 0. 20 913. 28 0. 15 - 27. 46 - 3. 43
Gross 628380. 51 100. 00 628380. 51 100. 00 0. 00 0. 00
Table 3 Land use/ cover conversion matrix in study area in the 1990s ( hm2 )
Land use/ cover classif icat ion
1999
CL FL W-GL WL UL
FL 83174. 58 308. 90 291. 08 2021. 52 119. 79
W-GL 15000. 19 374764. 46 1506. 60 5503. 65 5. 40
1991 WL 394. 67 1478. 07 8006. 97 229. 78 1. 08
U L 3213. 41 19742. 33 244. 18 111084. 33 30. 51
CL 240. 31 154. 90 7. 83 99. 55 756. 50
In creased area 18848. 58 21684. 20 2103. 61 7854. 50 156. 78
Redu ced area - 2741. 29 - 22015. 84 - 2049. 69 - 23230. 43 - 502. 59
Net variation* 16107. 29 - 331. 64 - 53. 91 - 15375. 93 - 345. 81
Total variat ion* 21589. 87 43700. 04 4153. 30 31084. 94 659. 37
* Net variation= in creased area+ reduced area; Total variat ion= increas ed area+ r educed area
3 Discussion and Conclusion
It is obvious that landscape disturbance has
occur red in the study ar ea during the early to late
1990s period. A large proport ion of the land, in-
cluding pr eviously UL and W-GL, w as cult iv ated
and subsequent agricultural abandonment r esulted
in large patches of cleared land suggesting that hu-
man agr icultural act ivity disturbed the study area
more intensiv ely . Data f rom the of f icial w ebsite[ 12]
also demonst rates this: T ianzhu w as tradit ionally
an animal husbandry area w ith litt le farm act ivity,
but since the 1950s, the ar ea of farmland had rap-
idly increased from 129. 33 km
2
in 1950 to 225. 67
km 2 in 1998. The inter change among dif ferent land
catego ries, especially since 1985, w as mo re f re-
quent w ith fast econom ic g row th.
Conversion betw een W-GL and other land
use/ cover types w as ev ident and character ized by
the g ross area o f W- GL remaining stable in the
study per iod, w hile bo th the incr easing area and
decreasing area w er e large. A nalysis show s that
68% of the reduced W-GL area w as converted into
FL, w hile it about 25% to UL; and for the grow th
of W-GL, 91. 04% comed from UL. Fur ther spa-
t ial dist ribut ion analyses of land use/ cover reveal
most of the conversion o f U L to W-GL occurred in
the high mountain areas above 4000 m, e. g . M aya
M ts. , Leig ong M ts. , M azanghuashan M ts. etc.
and in many small patches. Previous studies[ 13~ 14]
19
草地学报 第 18卷
revealed that the w inter temper atures of T ianzhu
rised obv iously since the1980s, and since the 1990s
the temperatures of four seasons all shif t cont inu-
ously, w hich seemed to co incide w ith the contem-
poraneous LU CC. T hus, w arming of study area
might contr ibute to the conversion of U L ( fo rmed
in cold climates at high latitude) to W-GL in the
high mountain areas.
Shiyang River is the main w ater source o f
Minqin Oases and serves about 2 m illion people.
Because Shiyana River takes it s rise in Tianzhu,
the change of land use/ co ver pat tern in study area
simultaneously af fects the ecosystem of M inqin Oa-
ses, dow nst ream o f the Shiy ang River. The M in-
qin Oases ho lds highly impo rtant academic signif-i
cance due to it s desert if icat ion pro cesses. T he sub-
stant ial increase in cult ivated land in Tianzhu
County w ould undoubtedly increase the w ater con-
sumpt ion, w hich aggr avates the ecosy stem co llapse
of Minqin Oases. How ever there w as no signif icant
change in runof f of the Shiyang River betw een
1956 to 1999[ 15] .
Acknowledgments
Authors thank Prof . CHEN Lu-ping ( U. S.
Univer sity of Alaska Fairbanks, School o f Natural
Resources and Ag ricultural Sciences ( SNRA S ) )
fo r his suggest ions of the manuscript and English
cor rect ion.
References
[ 1] GLP. Science Plan and Im plem entation St rategy [ R] . IGBP
Report No. 53/ IH DP Report No. 19. IGBP S ecretariat ,
Stockholm, 2005
[ 2] T urner II B L, S kole D, Sanderson S , et al . Lan d Use and
Land C over Chan ge ( LUCC) Science/ Research Plan[ R] . IG-
BP Repots No. 35, 1995
[ 3] Li X B . T he core area of global change research: a review of
internat ional resear ch on Land u se/ Land cover ch ange[ J] . Ac-
ta Geograph ica Sin ica, 1996, 51( 6) : 553- 557
[ 4] Jiao Y Z, Jiang H , Xu Z. Ef fect s of land u se chang es on soil
phys-i ch emical propert ies in farm-grazing t rans it ion belt of In-
n er Mongolia Au ton om ou s region [ J] . Acta Agrest ia Sin ica,
2009, 17( 2) : 234-238
[ 5] Wang K, Hong F Z, Zong J Y. Resour ces and th eir su stain-
able ut ilit y in the thr ee-river headw aters region [ J] . Acta
Agres tia Sinica, 2005, 13 ( Supplement ) : 28-34
[ 6] Yao T D, Zhu L P. Th e Respon se of environmental changes
on T ibetan Plateau to global chan ges and adaptat ion st rategy
[ J] . Advances in Earth S cien ce, 2006, 21( 5) : 459-464
[ 7] Zheng D, Lin Z Y, Zhang X Q. Progres s in studies of Tib etan
Plateau and global en vir onmental change [ J] . Earth Science
Front iers, 2002, 9( 1) : 95- 102
[ 8] Piao S L , Fang J Y. Dynam ic veg etat ion cover change over the
last 18 years in C hina [ J ] . Quaternary Sciences, 2001, 21
( 4) : 294-302
[ 9] Zh en B H , T ian Z Q, Wang W J, e t al . An aly sis of recent
land u sage and survey in Western China [ J] . Acta Ecologica
Sinica, 2004, 24( 5) : 1078-1085
[ 10] Song L C , H an Y X, Zh ang Q, et al . M on th ly tempora-l spa-
t ial dist ribut ion of s ands torm s in China as w ell as the Origin of
Kosa in Japan an d Korea [ J ] . Chin ese J ou rnal of Atmosph eric
Sciences, 2004, 28( 6) : 820-827
[ 11] Kang S C , Zhang Y J , Qin D H, e t al . Evidence of Ice Core to
intens ive w arming recent ly in the source region of Yangtze
River [ J ] . Ch ines e Science Bul let in, 2007, 52( 4) : 457-462
[ 12] T ianzhu government . In tr odu ct ion of T ianzhu T ibetan Auton o-
m ous County, Gansu Provin ce [ EB/ OL] . ht tp: / / w ww . e56.
com. cn/ minzu/ autonomy/ 9yzh _ detail. asp? Municipalit y _ ID
= 143, 2006-12- 11/ 2009-05- 26
[ 13] Wang J B, Wang Z G. Variat ion characteri st ics of temperature
in northeastern border region of Qin gha-i T ibet Plateau [ J ] .
Variat ion Agricultural Research in the Arid Areas, 2007, 25
( 1) : 176-180
[ 14] Bai X M, Pu X P. T he analyses of climate change of Alpine
Meadow Grassland in Jingqianghe, Gansu province [ J ] . Prat-
acu ltural s cien ce, 2003, 20( 7) : 531- 532
[ 15] Ding H W, Wang G L, H uang X H . Runoff r educt ion into
Hongyash an Reservoir and analysis on water resources crisis
of M ingqin Oases [ J] . Journ al of Desert Research, 2003, 23
( 1) : 84-89
(责任编辑 邵新庆)
20