urban cluster development and agricultural land loss in...
Post on 08-Oct-2020
7 Views
Preview:
TRANSCRIPT
Urban cluster development and agricultural land loss in China and India: A multi-scale and multi-sensor analysis
Background • By 2030, the urban population of China and India will grow by 700 million, the largest urban transition in history. • Economics, urban planning, and political science research show that patterns of urban growth are not fully captured
within administrative boundaries, and emphasize urban clusters as an important unit for policy and land-use planning. • No prior remote sensing‐based studies have evaluated urban cluster growth or its underlying causes.
A. Hotspots of urban cluster growth
C. Drivers of urban cluster growth and land conversion D. New remote sensing methods for characterizing urbanization
This project is supported by NASA Land-Use/Land-Cover Change Grant NNX11AE88G. Poster design by Nick Allen
Fiscal Transfers Local governments take a stronger role in urban expansion • Area of urban
land conversion greater in counties with lower central government transfers
Higher Education
Drivers of Urban Land Expansion Primary predictors of urban land conversion in China are local, not regional factors • Urban wages, agricultural investments per
capita, and the ratio of agricultural to urban land rents are the dominant predictors of urban land conversion
• Agricultural investments driving farmland conversion suggest a policy failure
3D Urban Structure Novel dataset, SeaWinds, adds new dimension to urban form
• SeaWinds backscatter used to derive coarse-resolution urban form and joined with NTL-derived urban extent to characterize changes in 3D structure over time
• Data show Chinese cities growing more vertically than Indian cities, mirroring real estate and investment patterns
Frolking S, Milliman T, Seto KC, & Friedl MA. 2013. A global fingerprint of macro-scale changes in urban structure from 1999 to 2009. Environ. Res. Lett. 8(2), 024004.
Karen C. Seto (PI) • Yale School of Forestry & Environmental Studies • karen.seto@yale.edu • urban.yale.edu
Zhang, Qian, Wallace J, Deng X, Seto KC. 2014. Central versus local states: Which matters more in affecting China’s urban growth? Land Use Policy 38: 487–496.
Decomposition analysis of determinants of China’s land conversion, 1989-2000
Estimated Parameter
% change in variable
Impact on converted land (%)
Contribution (%)
Agriculture/Urban Land Rent Ratio
-0.705 -0.11 0.078 0.137
Urban Wages 0.308 1.17 0.360 0.632
Agricultural Investment 0.063 0.59 0.037 0.065
Converted Land 0.57 1
Jiang L, Deng X, Seto KC. 2012. Multi-level modeling of urban expansion and cultivated land conversion for urban hotspot counties in China. Landsc. Urban Plan.108(2–4): 131–139
University establishments drive regional growth via infrastructure development, skilled labor training
• Contribution increases with the age of the university • Diminishing returns on co-located universities suggest
investments should focus on underdeveloped regions
NTL temporal profiles can detect urban growth rates and underlying processes
• Rates of urban growth in China and India were higher than the developed world in most two-year periods from 1992 to 2008
• Urban growth accelerated more in 1990s than 2000s in both countries
• Less accurate in less-developed and low-lit regions
Identifying Urbanization Typologies
Zhang, Qian, Seto KC. 2013. Can night-time light data identify typologies of urbanization? A global assessment of successes and failures. Remote Sens. 5(7), 3476–3494.
Zhang, Qingling, Schaaf C, Seto KC. The Vegation Adjusted NTL Urban Index: A new approach to reduce saturation and increase variation in nighttime luminosity. Remote Sens. Environ.129, 32–41
Zhang, Qingling, Seto KC. 2011. Mapping urbanization dynamics at regional and global scales using multi-temporal DMSP/OLS nighttime light data. Remote Sens. Environ. 115(9), 2320–2329.
China’s agricultural land intensifying and spreading • The net loss of cultivated land
in developed provinces suggests intensification
• Additional land brought under cultivation offsets agricultural land loss due to urbanization
Loss of Indian agricultural land concentrated in areas most suited for cultivation • Concentrated in industrialized states • 0.7 million ha lost, 5x the size of Delhi
Pandey B, Seto KC. Accepted. Urbanization and agricultural land loss in India: comparing satellite estimates with census data. J. Environ. Manage.
Fixed Effects Model R2, p-value, (t-statistic)
Random Effects Model R2, p-value, (t-statistic)
Intercept 0.720, p<0.01 (14.94)
Converted Land Area (‘000 ha) −0.010, p<0.1 (−2.19) −0.016, p<0.01 (−3.75)
Cultivated Land Area (‘000 ha) −0.652, p<0.01 (−6.60) −0.883, p<0.01 (−17.81)
Industrial GDP (billion ¥) −0.010, p<0.01 (−3.99) −0.011, p<0.01 (−5.61)
GDP per cap. (‘000 ¥) 0.004, p<0.01 (1.89) 0.006, p<0.01 (4.02)
Ag. investment per cap. (‘000 ¥) 0.706, p<0.05 (2.55) 0.671, p<0.01 (3.30)
Length of Highways (km) 0.0007, p<0.01 (6.91)
Distance of county seat to province capital (km)
−0.0005, p<0.01 (−4.46)
Ratio of low-incline land (<8° grade) 0.369, p<0.01 (12.06)
Avg. elevation (km) −0.077, p<0.01 (−4.80)
Avg. annual precipitation (mm) 0.0002, p<0.01 (7.45)
Avg. annual temperature (°C) 0.027, p<0.01 (8.62)
Observations 5010 5010
R-squared 0.21 0.31
Jiang L, Deng X, Seto KC. The impact of urban expansion on agricultural land use intensity. Land Use Policy 35: 33–39
Total area of agricultural land lost to urban growth, 2001–2010
B. Analysis of urban land conversion
NTL temporal profiles identify distinct urbanization typologies
Reducing DMSP/OLS NTL Saturation
Key findings A. Hotspots of urban cluster growth During 1992–2010, India’s urban cluster growth hotspots were Greater Bangalore, the Mumbai–Hyderabad corridor, and the Delhi–Ludhiana corridor; there is no evidence of Delhi–Mumbai corridor growth. These hotspots are consistent over time and are concentrated around existing large cities, with little spillover into surrounding regions. In China, urban cluster growth hotspots were the Pearl River Delta, the Yangtze River Delta, the Shandong Peninsula, and the Beijing–Tianjin corridor. Over time, these hotspots shrank or expanded and new clusters emerged.
B. Analysis of urban land conversion Both China and India have experienced rapid loss of cultivated land to urban expansion in the past two decades. During 2001–2010, India lost an area of agricultural land equivalent to five times the size of Delhi. Urban expansion in India’s smaller cities results in more agricultural land conversion than urban expansion around bigger cities. In China, newly cultivated farmland has largely offset the loss of agricultural land to urban expansion. Agricultural intensification is negatively correlated with urban expansion in China.
C. Drivers of urban cluster growth and land conversion Foreign direct investment spurred urban cluster growth only in India’s primate cities, Mumbai and Delhi. India’s higher education institutions promote regional economic growth through infrastructure development and labor training. India’s major urban renewal program, the JNNURM, has not caused urban land expansion. In China, urban cluster growth is driven by urban wages and FDI, and central government transfers are declining in significance. Agricultural investments induce urban land conversion in China, which may indicate a policy failure.
D. New remote sensing methods for characterizing urbanization Spatial statistical analysis of time series DMSP/OLS nighttime light (NTL) data can detect urban growth hotspots at regional scales. The Vegetation Adjusted NTL Urban Index (VANUI) combines NTL with MODIS NDVI to provide global, timely information about inter-urban variability in form, infrastructure, and energy use. Using SeaWinds with NTL allows for characterization of 3D urban structure.
Leveraging NDVI to correct urban NTL saturation
• The Vegation Adjusted NTL Urban Index (VANUI) increases inter-urban variability using underlying biophysical and urban characteristics
Changes in structural backscatter (PR) and nighttime lights (NL), 1999–2009
Fragkias M, Seto KC. In revision. Do universities drive urban economic growth and regional development in India? Evidence from remote sensing and higher education data. Urban Economics.
Night-time lights shows cluster growth around major urban areas and transport corridors in India
China’s hotspots are concentrated around major urban regions and show greater inter-period variation
2,000 km 1,000 500 0
1,000 km 500 250 0
1992–1997
1992–1997 1997–2003
1997–2003 2003–2010
2003–2010
NTL time series show large inter-regional variations in urban growth rates
More urbanized counties have smaller fiscal transfers from Beijing as a percentage of revenue
∆PR
∆NL 1999
2009
Delhi (National Capital Region) Bangalore Chennai
Shenzhen Shanghai
Urban cover per cell (%)
0 – 19.9% 20 – 29.9% 30 – 39.9% 40 – 49.9% 50 – 59.9%
70 – 79.9% 60 – 69.6%
90 – 99.9% 80 – 89.9%
100%
Arrows represent non-water, 0.05°cells in an 11×11 grid around each urban center; tail and head are at 1999 and 2009 coordinates of cell PR and NL
Pearl River Delta Shanghai Beijing
New Delhi Chennai Bangalore
VANUI VANUI VANUI NTL NTL NTL 20 km
N 0.0 1.0
Hotspot Coldspot
State boundary District boundary
Hotspot city
2,000 km 1,000 500 0
Net fiscal transfers as a percentage of
total revenue
< 0.25 0.26–0.40 0.26–0.40 0.56–0.70 > 0.70 District Province
1,000 km 500 250 0
Area estimate (in hectares)
0 6–37 44–100 106–187 194–325
331–519 531–819 825–1519 1550–3737 3762–27756
Beijing
0
10
20
30
40
50
60
70
1992 1994 1996 1998 2000 2002 2004 2006 2008
Mea
n N
TL D
N
Year
1. Constant urban activity
2. Early urban growth
3. De-urbanization
4. Constant urban growth
5. Recent urban growth
-40
-30
-20
-10
0
10
20
30
40
50
92-94 94-96 96-98 98-00 00-02 02-04 04-06 06-08
Rate
of u
rban
gro
wth
(%
)
Two-year period
China
India
Japan
US
top related