Transcript Lorem Ipsum - Harvard University

Monitoring and Modeling LandUse Change in the Pearl River
Delta, China, Using Satellite
Imagery and Socioeconomic
Data
Robert K. Kaufmann
Harvard University
January 29, 2003
http://www.bu.edu/cees/readmoreRK.html
Modeling and Forecasting Effects of Land-Use
Change in China Based on Socioeconomic Drivers
Boston University
Department of Geography
Principal Investigator: Robert K. Kaufmann
Co-Investigators:
Curtis E. Woodcock
Dennis G. Dye
Karen C. Seto
Chinese Collaborators:
Lu Jinfa, Institute of Geography CAS
Li Xiaowen, IRSA
Wang Tongsan, Economic Forecasting Center
Huang Xiuhua, IRSA
Liang Youcai, State Information Center
Funded by NASA LCLUC-NAG5-6214
Why Pearl River Delta, Guangdong Province?
• 1988 - 1996 real GDP growth: 350-550%
• Major agricultural region and national leader in
production of: lychees, bananas, pond fish, sugar cane
• Special Economic Zones
• Geographic proximity to Hong Kong
• Cultural ties to overseas Chinese investors
China
Study Area: Pearl River Delta
30 December 1995 TM 432
3
1
2
10 December 1988 TM 432
3 March 1996 TM 432
Land-Use Change Map
water
natural vegetation
agriculture
urban
natural to urban
agriculture to urban
5 km
10 December 1988 TM 432
5 km
3 March 1996 TM 432
water
natural vegetation
agriculture
urban
Land-Use Change Map
water to ag
natural to urban
ag to urban
10 December 1988 TM 432
3 March 1996 TM 432
Land-Use Change Map
water
natural vegetation
agriculture
urban
agriculture to water
natural to urban
agriculture to urban
5 km
Official Estimates vs.
Satellite-Derived Estimates
of Agricultural Land
Seto, K.C., R.K. Kaufmann, and C.E.
Woodcock. 2000.
Agricultural land conversion in
southern China. Nature 406: 121.
Land Use and Land Use Change:
1988 -1996
QuickTime™ and a
Cinepak decompressor
are needed to see this picture.
Made in conjunction with NASA Goddard Space Flight Center
1988 - 1996 Real GDP Growth
350 - 550%
200 - 300%
100 - 190%
25 km
1988 - 1996 Percent Land-Use Change of Counties
High: 17 - 23%
Medium: 14 - 15%
Low: 0 - 9%
25 km
Modeling Socioeconomic Drivers of LUC
Yit = ai + bixit + mit
i = 1,…, N
t = 1,…, T
• Dependent variables:
- agricultureurban
- natural vegetation/waterurban
• Examples of independent variables:
- GDP
- Demography (m/f/rural/urban)
- Gross output value in industry & agriculture
- Wages by sector
Drivers of Land-Use Change
Agriculture to Urban = -0.11
[-5.0]
+ 1.50 *Relative land productivity
[1.97]
- 3.97 *Ag labor productivity
[-4.58]
+ 1.24 *Investment in capital construction [2.74]
+ 0.03 *Average wage
[5.98]
Natural to Urban = 0.028
+ 6.53*Relative land productivity
- 1.39 *Relative labor productivity
+ 2.85*Investment in capital construction
[3.13]
[3.55]
[-3.37]
[3.27]
Seto, K.C. and R.K. Kaufmann, In press, Modeling the drivers of urban land-use change in the
Pearl River Delta, China: Integrating remote sensing with socioeconomic data. Land
Economics
Evaluation of Results
* Panel cointegration--variables share the
stochastic trend
* Hendry forecast test--Regression results stable over
space and time
* Moran’s I--No spatial autocorrelation
* Granger causality--Some evidence that RHS
variables “Granger cause” land use change
no evidence for opposite effect
Results
• Successful mapping of land-use change with high
accuracy (93.5%)
• Amount of developed land has increase by 319%
between 1988 and 1996
• Developed new method to evaluate change in series
of images using time series techniques
• Identified and quantified major drivers of urbanization