Contributions of climate, elevated atmospheric CO2 concentration and land surface changes to variation in water use efficiency in Northwest China.

• WUE showed a significant increasing trend with a break point in 2002 in NWC. • A robust coupled attribution analysis with elasticity method was employed. • GPP contributed 88%, while ET a contributed 12%, to variation in WUE. • Precipitation was the primary factor driving the 66% increase in WUE. • A structure diagram was developed to present the driving mechanism of WUE variation. Besides contributing to a broader understanding of ecosystem-hydrology interactions, ecosystem water use efficiency (WUE) is a key factor operating in an ecosystem's balance between carbon assimilation and water transpiration. However, the effects of climate, elevated atmospheric CO 2 concentration and land surface changes on variation in regional WUE are sparsely documented. To elucidate the driving mechanisms of spatiotemporal variation in WUE in northwest China (NWC) between 1982 and 2015, the contribution of a changing environment — represented by the WUE's sensitivity to Gross Primary Productivity (GPP) and evapotranspiration (ET a) — was examined under different landscape types. Annual WUE series for NWC showed a significant increasing trend and a break point in 2002 with an increasing rate (7.40 mg C kg−1 H 2 O yr−1) prior to the break, and a slightly increasing rate (1.1 mg C kg−1 H 2 O yr−1) thereafter in NWC. A robust (Bias = 2.6%) attribution analysis revealed that GPP was the dominant factor contributing 88% variation in WUE, while ET a contributed 12%. Precipitation was the principal indirect driving factor that affected WUE variation in NWC, while the secondary factor was different among different landscapes. Regionally, precipitation drove 66% of the increase in WUE, wind speed contributed −15% of variation in WUE, and temperature contributed 12% of increase in WUE, while the remaining four factors contributed <3%. Compared to energy-related factors, water-related factors were the more sensitive drivers to WUE variation. The results can guide agricultural irrigation, integrated water resources management and ecological restoration under the changes in climate coming to Northwest China. [ABSTRACT FROM AUTHOR]