9s soil and landform for continental Australia analysis of biodiversity pattern: aggregated from 3s data

A selection of 9-arcsecond resolution substrate surfaces (soil and landform) for the Australian continent, aggregated from 3-arcsecond source data. These substrate surfaces have been selected because they have been found to be relevant to biodiversity modelling using generalised dissimilarity modelling. These data are intended to be used along with a similarly compiled and spatially standardised 9-arcsecond gridded climatic layers. See links for related collections. \nLineage: The 3-arcsecond gridded data were aggregated to a 9-second grid for use in biodiversity modelling. Details of the soil aggregation process are given in the supporting document (SLGA_9s_aggregation_metadata.docx), provided with this collection. Briefly, the 9-arcsecond aggregated soil properties are spatially aggregated versions of the 3-arcsecond Soil and Landscape Grid of Australia (SLGA) soil attributes, aligned with the 9-arcsecond DEM of Australia. Each soil attribute is represented as three grids for each depth layer, giving the 5th, 50th and 95th percentile values. The aggregation method accounts for the probability distribution of values in the SLGA data and the different alignments of the 3-arcsecond SLGA grid and the 9-arcsecond DEM, using cumulative distribution functions weighted by cell area contributions to the output grid cells. The landform variables were spatially aggregated using bilinear sampling. Datasets were selected that minimised spatial discontinuities, as well as their ecological relevance.

本数据集为澳大利亚大陆9弧秒分辨率的基底地表（土壤与地貌）精选集合，其数据源自3弧秒原始资料经聚合处理所得。该数据集经筛选入选，因其适用于基于广义相异建模（generalised dissimilarity modelling）的生物多样性建模研究。本数据集可与经类似方式整合、空间标准化后的9弧秒网格化气候图层配套使用。相关数据集集合可参阅对应链接。
数据溯源：3弧秒网格化原始数据经聚合处理后生成9弧秒网格，以适配生物多样性建模需求。土壤聚合流程的详细说明详见本数据集配套提供的辅助文档（SLGA_9s_aggregation_metadata.docx）。简言之，经聚合得到的9弧秒土壤属性数据，是3弧秒分辨率的澳大利亚土壤与地貌网格（Soil and Landscape Grid of Australia, SLGA）土壤属性的空间聚合版本，且与澳大利亚9弧秒数字高程模型（Digital Elevation Model, DEM）空间配准一致。每个土壤属性按深度分层，各层均包含3个网格化数据集，分别对应5%、50%（中位数）与95%分位数的属性值。该聚合方法采用以输出网格单元的像元面积占比为权重的累积分布函数，兼顾了SLGA数据的属性值概率分布，以及3弧秒SLGA网格与9弧秒DEM之间的空间配准偏差。地貌变量则通过双线性采样法完成空间聚合。数据集的筛选同时兼顾了空间连续性最优与生态相关性两项要求。