Insciences J. 2011, 1(4), 194-210;doi:10.5640/insc.0104194

Full Research Paper, Section: Sensors

Diachronic Mapping of LUCC in the Northwest of China over the last half Century: Conversion of CORONA Panchromatic reflectance into LANDSAT NDVI Imagery

Frank Veroustraete^{1 *} , Hui Fang², Chen Xi², An Ming Bao², Qinghan Dong¹, Liu Tie^2,3, Patrick Willems³

¹ Flemish Institute for Technological Research (VITO),Unit of Spatial and Environmental Aspects (RMA).Boeretang 200, BE2400 Mol, Belgium.
² Xinjiang Institute for Ecology and GIS (XIEG). 40-3 South Beijing Road, Urumqi, Xinjiang 830011, China.
³ Catholic University of Leuven (K. U. Leuven), Department of Civil Engineering. Kasteelpark Arenberg 40, BE3001 Heverlee, Belgium.

^* Author to whom correspondence should be addressed.

Published: November 22, 2011

Complete Article

PDF Download

Abstract

It is only by using the first collection of intelligence satellite imagery acquired by CORONA, that time series of NDVI imagery can go back for a time period of nearly 50 years, based on diachronic mapping. This is an exceptionally long time period if not the longest, for which remote sensing imagery is available to estimate land use change for specific regions on the globe. The P.R. of China and the former USSR can be documented to nearly 50 years back in time, using CORONA satellite imagery. Typically – for many regions of the globe – is that a 50 year time span represents the principal time period in which many anthropogenic land use changes have taken place. In this paper, an account is given of the land use change which took place over the last 44 years in the Chinese province of Xinjiang, located in the North-west of the country. Declassified CORONA imagery was used, as well as LANDSAT MSS and ETM+ imagery. The Xinjiang province has one of the largest inland water basins on the globe, e.g. that of the Tarim river. Moreover, the province has grown into a major cotton and fruit producer in China. Based on diachronic mapping with CORONA and LANDSAT imagery, we can illustrate that over the last 44 years the Tarim river basin has undergone drastic land use changes, e.g., from nonfragmented landscapes of natural vegetation to fragmented mosaics of agrcicultural plots and desert. The region of interest in this study is the downstream area of the Tarim basin in the province of Xinjiang in the P.R. of China. A transfer function was established between canopy panchromatic reflectance (p_c,pan) and canopy NDVI (NDVI_c). This allows to reach the objective to increase the back casting time period of land use change observations based on satellite observations. To this end, a coupled PROSPECT-SAIL radiative transfer modelling approach was applied. Our approach elicited that the relationship between NDVI and p_pan at leaf level, converges to a single relationship depending on leaf structure. The canopy level transfer function used to convert p_c,pan into NDVI_c was established using different chlorophyll contents, LAI values and a bright soil in the PROSPECT/SAIL radiative transfer models. A main result is that for a time lapse of 44 years, a drastic change in land use took place in the downstream reaches of the Tarim river basin. In many places natural vegetation has converted into high productivity irrigated agricultural plots. On the other hand natural vegetation also degraded into desert and semi-desert areas of very low to zero level productivity. A relatively small area (13%) of the Tarim downstream area has remained unchanged during the course of 44 years. Typically, the radiation regime and especially the hydrology of the the Xinjiang province has changed significantly over the last 44 years due to this drastic land use change. During this period the Tarim river retracted upstream, several hundreds of km, leaving behind desert land inapt for agricultural production and hence inapt to sustain human life.

CORONA sensor, Diachronic mapping, land use change, NDVI, Tarim river basin, Xinjiang province