Characteristics of the spatial pattern and aspatial correlation of seawater quality in Jinhae Bay of Korea

摘要

The spatial autocorrelation measure Moran's Index was used in this study to understand spatial distribution pattern of seawater quality that has been seasonally monitored for three years in Jinhae Bay of Korea. When calculating Moran's Index, neighborhood range for each target point is generally set as a fixed distance. However, this method is not appropriate for current study as the coastline is complicated and spatial distribution of monitoring stations is non-uniform. Accordingly, a program was developed to automatically calculate Moran's Index by setting the neighborhood ranges using a Gabriel network of monitoring stations. Relationships among seven indicators of ocean water quality were analyzed for autumn, as it experiences the highest frequency of clustering pattern of the four seasons (57%). Results of aspatial correlation analysis and spatial distribution pattern analysis were summarized into the following four cases: 1) both aspatial correlation and spatial autocorrelation were high, 2) aspatial correlation was high but spatial autocorrelation was low, 3) aspatial correlation was low but spatial autocorrelation was low, or 4) both types of correlation were low. The findings suggest that correlation analysis as a traditional statistical analysis should be conducted in conjunction with spatial autocorrelation analysis in order to understand the spatial distribution characteristics of ocean water quality.