Contiguous suitable area data of the Fitzroy catchment WA, Darwin catchments and Mitchell catchment Qld generated by the Northern Australia Water Resource Assessment

These contiguous suitable area data are based on the land suitability data from the Northern Australia Water Resource Assessment (NAWRA). To address on-ground operational farming constraints imposed by parcels of suitable land being too small or oddly shaped according to natural variability of land, or physical limits on suitable farming land parcel sizes, contiguous suitable area data was generated. This contiguous suitable area data is based on crop-specific minimum areas and minimum length/width of contiguous suitable land and is produced as standalone data products for all crop groups. The rules are provided for download. The data was generated to remove the component of landscape complexity that natural distributions of soil and land variability and specific crop requirements produce. This data provides improved land evaluation information to identify opportunities and promote detailed investigation for a range of sustainable development options. The land suitability evaluation methods used to produce the underlying data are a modification of the Food and Agriculture Organisation (FAO) land evaluation approach. The land suitability analysis is described in full in the CSIRO NAWRA published report ‘Land Suitability of the Fitzroy, Darwin and Mitchell Catchments’. A technical report from the CSIRO Northern Australia Water Resource Assessment to the Government of Australia. The naming convention for these data is; ‘crop group’ underscore ‘major crop’ underscore ‘season code’ underscore ‘irrigation type code’ underscore ‘catchment code’ underscore ‘data type’ eg ‘CG7_CottonGrains_D_Fw_F_ContigArea’ is Cotton and grain crops grown in the dry season with furrow irrigation in the Fitzroy catchment contiguous suitable area data. The codes for season are; W – wet season; D – dry season; P – perennial. The codes for irrigation type are; S – overhead spray irrigation; T – trickle irrigation; Fd – flood irrigation; Fw – furrow irrigation; R – rainfed. It is important to emphasize that this is a regional-scale assessment: further data collection and detailed soil physical, chemical and nutrient analyses would be required to plan development at a scheme, enterprise or property scale. Several limitations that may have a bearing on land suitability were out of scope and not assessed as part of this activity (refer to the report), these limitations include biophysical and socio-cultural. For example these land suitability raster datasets do not include consideration of the licensing of water, flood risk, risk of irrigation induced secondary salinity, or land tenure and other legislative controls. Some of these may be addressed elsewhere in NAWRA eg flooding was investigated by the Earth observation remote sensing group in the surface water activity. The Northern Australia Water Resource Assessment provides a comprehensive overview and integrated evaluation of the feasibility of aquaculture and agriculture development in the Fitzroy, Darwin and Mitchell catchments as well as the ecological, social and cultural (indigenous water values, rights and aspirations) impacts of development.\nLineage: These contiguous suitable area raster datasets have been generated from a range of inputs and processing steps. Following is an overview. For more information see the CSIRO NAWRA published report 'Land Suitability of the Fitzroy, Darwin and Mitchell Catchments’. A technical report from the CSIRO Northern Australia Water Resource Assessment to the Government of Australia. 1. Collated existing data (relating to: soils, climate, topography etc, remotely sensed, of various formats: reports, spatial vector, spatial raster etc). 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. Create Digital Soil Mapping (DSM) attribute raster datasets. DSM data is a geo-referenced dataset, generated from field observations and laboratory data, 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, remote sensing images and some geophysical measurements. 7. Land management options were chosen and suitability rules created for DSM attributes. 8. Suitability rules were run to produce limitation subclass datasets using a modification on the FAO methods. 9. Final suitability data created for all land management options. 10. Companion predicted reliability data was produced from the 500 individual Random Forest attribute models created. 11. QA Quality assessment of these land suitability data was conducted by two methods. Method 1: Statistical (quantitative) assessment of the "reliability" of the spatial output data presented as a raster of the Confusion Index. Method 2: Collecting independent external validation site data combined with on-ground expert (qualitative) examination of outputs during validation field trips. A two-week validation field trip was conducted using a new validation site set which was produced by a random sampling design based on conditioned Latin Hypercube sampling. The modelled land suitability value was assessed against the actual on-ground value. These results are published in the report referenced above. 12. A two-step process was developed to simplify the data and was applied across the suitability data of the catchments. First the five suitability classes were aggregated to two: ‘suitable’ for suitability classes 1, 2 and 3, or ‘not suitable’ for class 4 and 5. Second, to further simplify the data, and to reflect the on-ground spatial constraints of farming practices, isolated one or two pixels of ‘not suitable’ contained in larger ‘suitable’ areas were reclassified as ‘suitable’. 13. For each crop group, a minimum area and width were defined based on knowledge of farming practices. Depending on the possible land use, minimum areas were deemed as 2.5 ha, 5 ha, 10 ha or 25 ha and minimum widths of 80 m or 120 m (rules are provided for download). 14. For each crop rule the minimum width was imposed by removing those parts of the suitable area that are narrower (in any direction) than the required minimum width. The remaining groups of connected cells were then tested to see if they meet the required minimum area and removed if they did not. 15. For the 25ha rule a consultant provided a further quantification to the processing, see supplied file Contiguous_suitable_areas.docx\n

这些连续适宜区数据基于澳大利亚北部水资源评估（Northern Australia Water Resource Assessment, NAWRA）的土地适宜性数据。为解决因土地自然变异性导致适宜地块过小或形状不规则，或适宜耕作地块规模存在物理限制而产生的实地农业操作约束，生成了连续适宜区数据。该连续适宜区数据基于作物特异性的连续适宜土地最小面积及最小长宽比，针对所有作物组生成独立数据产品，相关规则可供下载。生成此数据旨在消除土壤与土地变异性的自然分布及特定作物需求所产生的景观复杂性成分。该数据提供了更完善的土地评估信息，以识别机遇并推动对一系列可持续发展方案的详细调查。生成基础数据所采用的土地适宜性评估方法是对联合国粮食及农业组织（Food and Agriculture Organisation, FAO）土地评估方法的改良版本。土地适宜性分析详情载于澳大利亚联邦科学与工业研究组织（CSIRO）NAWRA项目发布的报告《菲茨罗伊、达尔文及米切尔流域的土地适宜性》，该报告是CSIRO澳大利亚北部水资源评估项目提交给澳大利亚政府的技术报告。数据命名规则如下：“作物组”下划线“主要作物”下划线“季节代码”下划线“灌溉类型代码”下划线“流域代码”下划线“数据类型”，例如“CG7_CottonGrains_D_Fw_F_ContigArea”代表菲茨罗伊流域旱季采用沟灌种植棉花与谷物作物的连续适宜区数据。季节代码：W——雨季；D——旱季；P——多年生。灌溉类型代码：S——高架喷灌（overhead spray irrigation）；T——滴灌；Fd——漫灌；Fw——沟灌；R——雨养。需强调的是，本评估为区域尺度：若要在方案、企业或地块尺度规划发展，需进一步收集数据并开展详细的土壤物理、化学及养分分析。部分可能影响土地适宜性的限制因素未纳入本活动范围（详见报告），包括生物物理及社会文化因素。例如，这些土地适宜性栅格数据集未考虑水资源许可、洪水风险、灌溉诱发次生盐渍化风险、土地权属（tenure）及其他立法管控。其中部分因素在NAWRA项目其他部分有所涉及，如洪水问题由地球观测遥感组在地表水活动中进行了研究。
数据谱系：这些连续适宜区栅格数据集由一系列输入数据及处理步骤生成，概述如下。详情参见CSIRO NAWRA项目发布的报告《菲茨罗伊、达尔文及米切尔流域的土地适宜性》，该报告为CSIRO澳大利亚北部水资源评估项目提交给澳大利亚政府的技术报告。
1. 整理现有数据（涉及土壤、气候、地形等；来源包括遥感数据；格式多样：报告、空间矢量、空间栅格等）。
2. 通过在协变量数据空间应用条件拉丁超立方统计抽样法，选择额外的土壤与土地属性站点数据位置。
3. 开展实地调查，收集新属性数据、土壤分析样本，并建立对地貌及景观过程的认知。
4. 进行数据库分析，依据属性建模所需的特定选择标准提取数据。
5. 采用R统计编程环境进行属性计算。利用ranger R包实现的随机森林（Random Forest）方法，基于选定输入数据及协变量数据构建预测学习模型。
6. 生成数字土壤制图（Digital Soil Mapping, DSM）属性栅格数据集。DSM数据是一种地理参考数据集，通过定量关系将实地观测与实验室数据、环境协变量数据相结合生成。它应用土壤计量学（pedometrics）——即利用数学与统计模型，将土壤观测信息与相关环境变量、遥感影像及部分地球物理测量信息相融合。
7. 选择土地管理方案，并针对DSM属性制定适宜性规则。
8. 应用改良的FAO方法，运行适宜性规则生成限制子类数据集。
9. 为所有土地管理方案生成最终适宜性数据。
10. 基于创建的500个独立随机森林属性模型，生成配套的预测可靠性数据。
11. 采用两种方法对土地适宜性数据进行质量评估（QA）。方法1：对空间输出数据的“可靠性”进行统计（定量）评估，结果以混淆指数栅格呈现。方法2：收集独立外部验证站点数据，并结合实地专家在验证调查期间对输出结果的（定性）检查。基于条件拉丁超立方抽样的随机抽样设计生成新验证站点集，开展为期两周的验证调查，将模拟土地适宜性值与实地实际值进行对比。相关结果载于上述参考报告。
12. 开发两步法简化数据，并应用于流域适宜性数据。第一步：将五类适宜性等级聚合为两类——1、2、3级归为“适宜”，4、5级归为“不适宜”。第二步：为进一步简化数据并反映实地农业操作的空间约束，将较大“适宜”区域中孤立的1-2个“不适宜”像素重新分类为“适宜”。
13. 基于农业操作知识，为每个作物组定义最小面积及宽度。根据可能的土地利用类型，最小面积设定为2.5公顷、5公顷、10公顷或25公顷，最小宽度设定为80米或120米（规则可供下载）。
14. 针对每个作物规则，通过移除适宜区域中（任意方向）窄于所需最小宽度的部分来施加最小宽度约束。随后测试剩余连通单元组是否满足所需最小面积，不满足则移除。
15. 针对25公顷规则，咨询顾问对处理过程提供了进一步量化说明，详见文件Contiguous_suitable_areas.docx