S. Frolking et al., Agricultural land-use in China: a comparison of area estimates from ground-based census and satellite-borne remote sensing, GLOBAL EC B, 8(5), 1999, pp. 407-416
We compare estimates of total cropland area, paddy rice area, and irrigated
cropland area in China from land cover maps derived from optical remote se
nsing in 1992-93 (1-km resolution NOAA AVHRR) and county-level agricultural
census data for 1990. At national, regional, provincial, and county scales
, the total cropland area estimated by remote sensing is 50-100% higher tha
n reported in the agricultural census. For topographically flat North and C
entral China, there is a high correlation between county-level cropland are
a estimates by the two approaches. For other regions, the correlation betwe
en remote sensing and agricultural census cropland area is much weaker. Thr
oughout China, there is only moderate to weak correlation between remote se
nsing-based and census-bases estimates of paddy rice area and total irrigat
ed cropland area. It is likely that the census data underestimates and the
remote sensing overestimates the actual cropland area. These uncertainties
in agricultural land cover estimates will contribute to uncertainty in any
large-scale biogeochemical analyses.