Spatial-temporal patterns of water use efficiency and climate controls in China's Loess Plateau during 2000-2010

Accurate assessments of spatial-temporal variations in water use efficiency (WUE) are important for evaluation of carbon and water balances. In this study, the spatial and temporal patterns of WUE and associated climate controls in China's Loess Plateau are investigated over 2000-2010 by utilizing remote sensing data and multiple statistical methods; which provides a greater understanding about how WUE changed after the Grain to Green Program (GTGP) launched. Carbon sequestration (i.e., net primary productivity, NPP) is estimated with the CASA model and water consumption (i.e., evapotranspiration, ET) is obtained from the MODIS product (i.e., MOD16). Our results identify an increasing trend in the regional mean NPP that amounted to 7.593 g C/m(2) center dot yr with an average value of 310.035 g C/m(2) center dot yr. Changes in ET are segmented into three stages, the growth (2000-2003), decline (2004-2006) and stable (2007-2010) stages. Regional WUE is measured at 0.915 g C/mm center dot m(2) and shows an upward trend at a rate of 0.027 g C/mm center dot m(2) center dot yr. Spatially, significant regional heterogeneity is found in both NPP and WUE with gradients decreasing from the southeast to the northwest, but sharp rises detected in northern Shaanxi. At the biome level, the annual average WUE of the four groups decrease in the order of grasslands > woodlands > shrublands > croplands. Moreover, all biomes in the grassland ecosystems exhibit a growth in WUE as does the arid desert zone in the northwestern region, suggesting that vegetation in moderately water-deficient areas may have a higher tolerance to drought. Among different meteorological factors, precipitation and drought severity index (DSI) in the Loess Plateau show a latitudinal zonality and influences the WUE, which indicated that the moisture rather than temperature would be the major control factor of the regional WUE. Finally, significant variation in vegetation WUE sensitivity in response to meteorological factors is noted. Temperature is found to be the dominant driving factor of shrublands WUE, whereas precipitation primarily influenced the WUE of grasslands, croplands, and woodlands. (C) 2016 Elsevier B.V. All rights reserved.