Norfolk Island soil clay content DSM attributes

Surface soil clay % and maximum clay % to 2m soil depth are two of nine soil attributes analysed as one component of a digital soil mapping exercise undertaken as part of the Norfolk Island Water Resource Assessment (NIWRA). It was selected due to its importance to groundwater recharge, surface water storage and managed aquifer recharge. Clay percent is a parameter influencing surface water infiltration, subsoil permeability and water holding capacity and has also been applied in the gully dam suitability rules in relation to construction and engineering properties for gully dams. These raster data represent modelled datasets of clay percent and are derived from field measured site data, limited laboratory analysis and environmental covariates. Data values are expressed as a percent. The measured site data attribute soil texture is used to represent clay content as defined by the National Committee on Soil and Terrain 2009 (NCST). Companion datasets presenting reliability of these data is also provided and can be found described in the lineage section of this metadata record. Processing was carried out in the ranger package inside R and attributes were modelled using a Random Forest approach. Further clay percent information can be found in the NIWRA technical report (Petheram et al., 2020). The DSM process is described in Appendix E of the NIWRA technical report.\nLineage: The surface soil clay % and maximum clay % to 2m soil depth datasets have been generated from a range of inputs and processing steps. The following is an overview of the methods detailed in Petheram et al. 2020. 1. Collated existing data (relating to: soils, climate, topography, natural resources, remotely sensed, of various formats: reports, spatial vector, spatial raster). 2. Selection of additional soil and land attribute site data locations by a conditioned Latin hypercube statistical sampling method applied across the covariate data space. 3. Fieldwork was carried out to collect new attribute data, soil samples for analysis and build an understanding of geomorphology and landscape processes. 4. Database analysis was performed to extract the data to specific selection criteria required for the attribute to be modelled. 5. The R statistical programming environment was used for the attribute computing. Models were built from selected input data and covariate data using predictive learning from a Random Forest approach implemented in the ranger R package. 6. Created surface soil clay % and maximum clay% to 2m soil depth Digital Soil Mapping (DSM) attribute raster dataset. DSM data is a geo-referenced dataset, generated from field observations coupled with environmental covariate data through quantitative relationships. It applies pedometrics - the use of mathematical and statistical models that combine information from soil observations with information contained in correlated environmental variables and remote sensing images. 7. Companion predicted reliability data was produced from the 500 individual Random Forest attribute models created. 8. QA Quality assessment of these DSM attribute data was conducted by three methods. Method 1: Statistical (quantitative) method of the model and input data. Testing the quality of the DSM models was carried out using data withheld from model computations and expressed as OOB and R squared results, giving an estimate of the reliability of the model predictions. Method 2: Statistical (quantitative) assessment of the spatial attribute output data presented as a raster of the attributes “reliability”. This used the 500 individual trees of the attributes RF models to generate 500 datasets of the attribute to estimate model reliability for each attribute. For continuous attributes the method for estimating reliability is the Coefficient of Variation. This data is supplied. Method 3: On-ground expert (qualitative) examination of outputs.

表层土壤黏土占比（%）与2米土层深度内最大黏土占比（%），为诺福克岛水资源评估（Norfolk Island Water Resource Assessment, NIWRA）所开展的数字土壤制图工作中分析的9项土壤属性中的两项。该属性被纳入分析，因其对地下水补给、地表水存储以及人工含水层补给均具有重要意义。黏土占比是影响地表水入渗、下层土壤渗透性以及持水能力的参数，同时也被应用于与谷坊建设及工程特性相关的谷坊适宜性评价规则中。本次提供的栅格数据为黏土占比的模拟数据集，其数据源包括野外实测样点数据、有限的实验室分析结果以及环境协变量，数据值以百分比形式表示。本次研究依据2009年国家土壤与地形委员会（National Committee on Soil and Terrain 2009, NCST）的定义，采用实测样点的土壤质地属性来表征黏土含量。本数据集还附带了数据可靠性相关的配套数据集，其详细说明可参见本元数据记录的谱系（Lineage）章节。

数据处理在R语言的ranger包中完成，土壤属性的建模采用随机森林（Random Forest）方法。更多黏土占比相关信息可查阅NIWRA技术报告（Petheram等，2020），数字土壤制图（Digital Soil Mapping, DSM）的具体流程详见该技术报告的附录E。

谱系说明：表层土壤黏土占比及2米土层深度内最大黏土占比数据集由多类输入数据经多步处理生成，以下为Petheram等（2020）中详述的方法概述：
1. 整合现有数据（涵盖土壤、气候、地形、自然资源、遥感等领域，格式包括报告、空间矢量、空间栅格等）。
2. 基于协变量数据空间，通过条件拉丁超立方统计抽样方法选取额外的土壤与土地属性样点位置。
3. 开展野外调查工作，采集新的属性数据与用于分析的土壤样本，并加深对地貌与景观过程的认知。
4. 开展数据库分析，按照属性建模所需的特定筛选标准提取数据。
5. 采用R统计编程环境进行属性计算，基于筛选后的输入数据与协变量数据，通过ranger R包实现的随机森林预测学习方法构建模型。
6. 生成表层土壤黏土占比及2米土层深度内最大黏土占比的数字土壤制图（DSM）属性栅格数据集。数字土壤制图数据为地理参考数据集，通过定量关系将野外观测数据与环境协变量数据相结合生成，其依托土壤计量学（pedometrics）——即结合土壤观测信息与相关环境变量、遥感影像所蕴含信息的数学与统计模型方法。
7. 基于构建的500个独立随机森林属性模型，生成配套的预测可靠性数据。
8. 采用三种方法对上述DSM属性数据开展质量评估（Quality Assurance, QA）：
方法1：针对模型与输入数据的统计（定量）评估。通过预留未参与模型训练的数据检验DSM模型质量，以袋外误差（Out-of-Bag, OOB）与决定系数（R²）结果表示，以此估算模型预测结果的可靠性。
方法2：针对空间属性输出数据的统计（定量）评估，生成属性“可靠性”栅格图层。该方法利用随机森林模型的500棵独立决策树生成500个属性数据集，以此估算每个属性的模型可靠性；对于连续型属性，可靠性估算采用变异系数（Coefficient of Variation）方法，该配套数据已随本数据集一同提供。
方法3：基于实地的专家定性检验输出结果。