| 摘要: |
| 为了对比研究不同空间插值方法在沉积物粒度空间分布表述上的精度及适用性，在2004年11月荣成宁津小型海湾 黑泥湾综合调查的基础上，探讨了表层沉积物粒度参数的空间插值，比较了IDW(反距离加权法)、Kriging(克里金插值法)、Spline(样条插值法)与NN(自然邻域法)4种GIS空间分析方法的特征、差异及实效性，并对影响插值结果准确性的因素进行分析。结果显示，在黑泥湾表层沉积物粒度结果的空间插值计算中，从插值准确性和空间表达能力两方面考虑，IDW，Spline，Kriging和NN中以IDW法较为适宜，但要考虑到“牛眼效应”的出现会与局部实际情况存在差异；Kriging法和NN法的插值结果准确性较高，但其空间表达能力稍逊；Spline法在近岸带表层沉积物粒度特征插值中的应用性相对较差。海洋调查要素空间分布的内在规律性是控制插值结果的主要因素，表层沉积物分布以长期稳定的潮流、地形特征为主导因素，呈现由岸向海条带状分布的特征。数据均匀分布区域的插值结果要优于边界区和突变区；另外，在选取的3个观察尺度上，不同插值方法的误差均与野外取样网格间距呈显著线性正相关。 |
| 关键词: 黑泥湾 表层沉积物 地理信息系统 插值方法 |
| DOI： |
| 分类号: |
| 基金项目:中国科学院海洋地质与环境重点实验室开放基金项目(MGE2009KG04)；我国近海海洋综合调查与评价专项项目(SD-908-01-03-01)国家自然科学基金项目(40676037) |
|
| GIS-based analysis of spatial interpolations of sediments grain size in Heini Bay |
|
YAN Li-wen,HUANG Hai-jun,LIU Yan-xia
|
| Abstract: |
| Interpolation applications on grain size parameters distribution of surface sediments in Heini Bay with GIS-based methods were discussed. Four interpolation methods (IDW, Kriging, Spline, NN) were compared and the factors influencing the interpolation accuracy were discussed from several aspects. The results indicate that considering interpolation accuracy and spatial expression IDW is relatively best among four interpolation methods in process of grain size parameter interpolation in Heini Bay. But “bulls eye” effects should be considered in IDW. Kriging and NN methods have a high accuracy and relatively bad spatial description. The applications of Spline methods isn't adaptive to grain size interpolation in nearshore sea. Compared with dissimilarity of different interpolation methods inner regularity, spatial distributions and sampling density contribute more in interpolation precision. The grain sizes have a thinning tendency seaward on control of steady currents and topography elements. The study also showed that interpolation errors are typically smaller in proportional spacing regions than that in border and abrupt changing regions. The interpolation errors of different methods are all positive correlative to the sampling density in situ. |
| Key words: Heini Bay surface sediments geographic information system (GIS) interpolation methods |
