基塘水面空间形态度量方法的比较与应用

中国拥有丰富的基塘景观。基塘水面作为基塘的主要部分，其空间形态关乎基塘经济、生态效益的发挥。为此，该研究旨在探寻能用于准确分析基塘水面空间形态的度量方法。该研究首先通过6个不同类型基塘水面样例，比较多个度量方法的有效性，据此选出合适方法；然后以珠三角的佛山市杏坛镇为研究区，将该方法用于分析区内基塘水面的空间形态特征及差异。研究结果表明：加权聚合邻近度（Weighted Aggregation and Closeness，WAC）指数对不同形态的度量结果相差10%以上，比其他15个常用度量方法更能有效地区分出规则塘和不规则塘的空间形态差异。研究区中，以整治后基塘为代表的规则塘的WAC均值比以未整治基塘为代表的不规则塘高40.18%，说明空间形态更加紧凑。规则塘空间形态在不同区域的差异较小；不规则塘的差异大于15%，在边缘位置的更为零碎。该研究为进一步探明不同基塘水面的形态特征与基塘经济、生态效益的关系奠定基础，同时也为分析其他不连接地理要素的空间形态提供应用参考。

Comparison and application of spatial pattern measurement methods for water surface of dike-pond

Dike-pond is a landscape system where water and land areas are integrated through ecological and agricultural processes. This type of system is commonly found in the low-lying river delta of China''s eastern coast. The water surface of the dike-pond plays an important role in the local economic development and ecological services. However, the use of existing landscape metrics to depict the spatial pattern of the water surface of the dike-pond has not been fully studied. This research aims to fill the gap by comparing a range of such metrics through a case study based on the Xingtan Township of Foshan City, China. We divide the dike-ponds in the study areas into two types: regular and irregular, which represent after and before dike-pond consolidation project respectively. We then test 16 landscape metrics using six cases of compact or scatter pattern. The metric that can distinguish various cases is applied to measure the water pattern of the dike-pond in the study area. Our comparison results suggest that the metric called Weighted Aggregation and Closeness (WAC) is the only one that can accurately capture all the water patterns exhibited in our study area. Among other 15 metrics, Mean Patch Fractal Dimension (MPFD), Area Weighted Mean Patch Fractal Dimension (AWMPFD), Mean Patch Size (MPS), Number of Patches (NumP), and Total Landscape Area (TLA), poorly perceive the pattern difference of the same pond types. And the metrics including Area-Perimeter ratio (AP), Perimeter-Area ratio (PA), Total Edge (TE), Mean Shape Index (MSI), Area Weighted Mean Shape Index (AWMSI), Mean Patch Edge (MPE), and Median Patch Size (MedPS) give measurement results that are contrary to the water pattern. Our application results show that the average WAC value of the regular pond is 40.18% higher than that of the irregular ponds. All the regular ponds have WAC values above 0.20, of which 66.76% are greater than 0.25. Only 4% of the irregular ponds have WAC values above 0.30, but 77.22% of them have WAC values lower than 0.25. These indicate that the pattern of the regular pond is more compact than that of the irregular pond. This is because the regular ponds are formed by large-scale consolidation projects, so they are gridded and closed to each other; the irregular ponds are randomly formed by the individual farmers and scattered around the edge area of the town and the regular ponds. The compact water pattern is beneficial to the dike-pond economy by providing a large proportion for aquaculture and exhibiting agglomeration effects, while the scatter water pattern is beneficial to the dike-pond ecology by promoting the land-water interaction. We note that the pattern difference of the regular pond is not obvious in the study area since the consolidation projects are standardized and uniform. There are more spatial differences of the water pattern of the irregular pond. The WAC values of such type in the central of the study area is 15% higher than that in the northern and southern, due to the central ponds have not been consolidated. These findings are useful to reveal the relationships of spatial pattern with the economic and ecological benefits of the dike-pond. Furthermore, WAC is not only effective for measuring the water pattern of the dike-pond, but also for other patterns with disconnected features.