Green technology innovation is the key to promote the high-quality development of China′s economy, this paper studies whether the technology innovation of China in recent years is green technology innovation, and the related problems such as whether technology innovation can reduce environmental pollution.In theory, under the strict environmental regulation, technological innovation helps to reduce environmental pollution, its mechanism includes three channels as energy saving effect, industrial upgrading effect, population agglomeration effect.Taking 285 Chinese cities from 2003 to 2016as samples, this paper does empirical research based on new urban innovation index.The results show that:in recent years China′s technology innovation is environmentally friendly with a threshold value, lower than which could not reduce environmental pollution.Mechanism analysis based on the mediation effect shows that technology innovation of some regions such as the cities in northeastern and northwestern China is mainly in the service of local industrial development, unable to effectively promote industrial structure optimization, thus failing to reduce environmental pollution.As technology innovation level across the threshold, three effects play a role at the same time, especially in big cities population agglomeration effect reducing the pollution caused by urban sprawl.

To explore the effects of sampling extent and spacing on spatial variability of soil moisture in a gravel-sand mulched jujube orchard, spatial analyses including classical statistics and geo-statistics were conducted with soil water content (SWC) of 0-50 cm in the 32 m×32 m ifeld by changing the sampling extent and spacing. Irrespective of sampling extents (i.e. 32 m×32 m, 28 m×28 m, 24 m×24 m, 20 m×20 m and 16 m×16 m), the SWC decreased with the increase of soil depth while the variation coefifcient increased with the increasing soil depth. The spatial variability of SWC was weak and moderate for all scales; The variation coefifcient (Cv), nugget (C0), and variation range (A) of SWC all increased with the increase of sampling extent. When three sampling spacings (i.e. 4, 8 and 12 m) were concerned, the nugget (C0) increased with the increase of sampling spacing; The variation range (A) decreased with the increase of sampling spacing; theCv was not affected by the sampling spacing. A strong spatial autocorrelation characteristic of SWC was found for all scales. The distribution patterns of SWC in the same soil layer at different sampling spacings were similar and tended to be "lfat" with the increase of the sampling spacing. The reasonable sampling spacing was found to be 8 m in our case.