不同空间粒度下地表水体分布格局多样性研究

Distribution Patterns of Surface Water Body Diversity at Different Grain Sizes

选取河南省北部、中部和南部典型研究样区,将空间粒度方法引入并应用于地表水体多样性的研究中。在总结前人研究水体多样性测度方法的基础上,提出新的方法——空间分布长度指数(MSHDLI),并在1 km×1 km网格尺度下研究MSHDLI、水网密度(RD)和空间分布面积指数(MSHDAI)的粒度效应、相关性和相关系数的粒度响应、尺度效应关系及回归模型。结果表明:随着粒度的增加,MSHDLI、RD和MSHDAI的粒度响应曲线均为下降型;MSHDLI与RD和MSHDAI有显著的相关关系,平均相关系数r分别为0.997和0.878(P<0.01),且MSHDLI与RD、MSHDAI相关系数的粒度效应属于显著正相关关系;MSHDLI与RD的拟合度函数均是对数函数,MSHDLI与MSHDAI的拟合度函数均是多项式函数;由三者的多元线性回归模型可知MSHDLI与RD、MSHDAI线性相关且对它们的解释程度达到95%以上。MSHDLI可以将RD描述水资源广度和MSHDAI描述水资源空间分布离散性的优势结合起来。

Three case study areas from north, central and south Henan Province in China were selected for applying spatial grain size method to the research of distribution patterns of surface water body diversity. On the basis of summarizing and adapting the previous research methods of surface water body diversity from related references, a new method of space distribution length index （MSHDLI） was put forward. Then all the selected information was further processed to present the effects of different grain sizes, correlativity and effects of grain size between the correlation coefficient of MSHDLI, water network density （RD） and MSHDAI （modified Shannon diversity area index） , and scaling relations between the three used indices and parameter, and multiple regression analysis model at a resolution of 1 km ~ 1 km. Results showed that with the increase of spatial grain size, the granularity of the response curves of MSHDLI, RD and MSHDAI belonged to drop type; MSHDLI maintained significant correlativity with RD and MSHDAI with average correlation coefficient of 0. 997 ~11 0. 878 （P 〈 0.01 ） , respectively. Effect of grain size of correlation coefficient was significantly positive correlation between MSHDLI, RD and MSHDAI. The fitting function of MSHDLI and RD was logarithmic function, MSHDLI and MSHDAI was polynomial function. Multivariate linear regression model of MSHDLI, RD and MSHDAI indices was also conducted to show that MSHDLI was the dependent variable and RD, MSHDAI were the independent variables. The fitting model showed that MSHDLI and RD, MSHDAI had linear correlationship, and the level of MSHDLI to explain the others combination of the indexes of RD and surface water resource distributions was above 95%. To a certain extent, MSHDLI MSHDAI to describe the breadth and the spatial can be a good discreteness in