艾比湖区域水质空间分布特征及其与土地利用/覆被类型的关系

为深入了解艾比湖周边水质空间格局的变化,利用2014年5月及2014年10月实测水样采样点研究了艾比湖区域4个水质指标的空间变化趋势,包括p H、电导率、矿化度及全盐。结果表明:(1)探讨2014年内干湿季水质的空间分布趋势,从总体上看,p H表现出明显的交错模式;电导率、矿化度及全盐在干湿季变化差异较大,高值均聚集在艾比湖周围,而在耕种土地、天然梭梭林附近值则较低;(2)分析各水质的空间自相关性,p H的HH主要分布于艾比湖北部,LL则主要集中于艾比湖南部。HL的分布,说明有较高p H值的点被有较低p H值的点所包围,表明了水质在一定程度上存在空间异质性。电导率、矿化度及全盐在沿着艾比湖周边到甘家湖梭梭林自然保护区随着地势的增高呈现出由HH-LH-LL的转变。(3)建立水质参数与各土地利用/覆盖类型之间的关系,研究表明研究区内林草地、盐渍地及耕地对水质的影响相对较为显著。(4)为了研究水质变化因素,选取耕地、林草地、盐渍地和未利用地与水质参数分别建立优选多元线性回归模型,所得到的相关系数R分别为0.58、0.72、0.74、0.71,结果表明优选拟合模型与数据的拟合程度较好。总之,开展艾比湖区域水质空间格局的变化趋势研究,对于干旱区水质的时空分布具有重要的理论和实际意义。

Correlation between the spatial water quality and land use/cover in the Ebinur Lake area

Water is an important resource for the development of the economy and has a significant influence on maintaining regional ecological balance. Spatial patterns of water quality methods using pH, electrical conductivity (EC), mineralization, and total salt content were examined at 62 sites in and around the Ebinur Lake area. The results of the water quality analysis showed a seasonal variation in pH, EC, mineralization and salt contents between the wet and dry seasons of 2014: (1) Overall, pH showed obvious crisscross pattern; EC, mineralization and total salt showed the large difference between the wet and dry season, high values were recorded around Ebinur Lake, and lower values in farm lands and Haloxylon forests; (2) analysis of spatial autocorrelation of the water quality showed the high-high aggregation(HH) of pH was mainly concentrated in the northern part of the Ebinur Lake and low-low aggregation (LL) was mainly in the southern part. Along the periphery of Ebinur Lake to the Haloxylon forest, EC, mineralization, and salt content decreased with HH- low-high aggregation (LH)-LL transformation per height of the terrain; (3) the relationship between water quality parameters and land use/cover types showed that the influence of forest-grassland, saline soil, and farm land on water quality is relatively significant; (4) to study the changes of water quality, we used farm land, forest-grassland, salinized land, and abandoned (dormant) land to establish an optimized multiple linear regression model with water quality parameters. The correlation coefficients of R were 0.58, 0.72, 0.74, and 0.71, respectively. These indicated the optimal fitting model is a better method for analysis. In summary, this study showed the water quality variations among different land use/cover in the desert region (Ebinur Lake area) can be used for comparison with other desert environments.