Abstract: Relying on RUE (rainfall use efficiency), CRD (cumulative rain use efficiency differences)indicates the development trend of vegetation in arid and semiarid region productivity. Based on the NDVI and meteorological data, using the method of CRD and combining it with the spatial of vegetation cover organically, this paper analyzed the influence of vegetation cover to RUE, and focused on the characteristic of "Process-Service" of vegetation cover and RUE. The precipitation interpolation, which is based on the topographic features, reduced the chance of precipitation interpolation errors and enhanced the scientific accuracy of the data. We probed the spatial pattern of RUE and its response to ecological changes on the Loess Plateau during 2000-2014. The results show as follows: (1)The vegetation was improved in most areas on the Loess Plateau which accounted for 81% of the total area from 2000 to 2014. The spatial of vegetation degradation in the research area presented the trend of marginalization. Overall spatial distribution of CRD on the Loess Plateau presented "high in east and low in west, high in south and low in north" from 2000 to 2014. (2)The positive and negative correlation between RUE and NPP co-existed, but the main correlation was positive from 2000 to 2014. The area that had significant positive correlation and extremely significant positive correlation between RUE and NPP accounted for 24.7% and 6.5% of the total area, respectively; that had significant negative correlation accounted for no more than 8.5%. (3)In order to analyze the internal-external connection between RUE and vegetation, by using the precipitation on the Loess Plateau, we explored the change of RUE and the correlation between RUE and NDVI. In the area that lacks precipitation, both precipitation and NDVI had significant negative correlation with RUE; in the area with abundant precipitation, both precipitation and NDVI had significant positive correlation with RUE. (4)The correlation coefficient of CRD and the trend of NPP were significant. Under the background of vegetation degradation, vegetation rainfall use efficiency was positively correlated with the degradation degree from 2000 to 2014; under the background of vegetation restoration, vegetation rainfall use efficiency was closely correlated with "Grain for Green" project. It shows that the improvement of regional vegetation led to the increases of RUE, drought trend was further enhanced and the burden of water resources had increased. But, the calculation of RUE was over-reliance on precipitation, so the rapidly increasing or decreasing of precipitation might cause pseudo degradation or pseudo recovery of vegetation. Currently, there is no better method to estimate RUE; in subsequent studies, we will probe the formation mechanism of RUE, and then provide a scientific basis for the change of vegetation.

Key words: vegetation restoration, vegetation degradation, rainfall use efficiency, Loess Plateau