Abstract Diverse on-farm agronomic data, gathered via precision agriculture (PA), require depuration and joint analysis of multiple georeferenced field characteristics such as crop yields, and soil and topographic aspects. Multivariate analysis of spatial variability has been recommended to understand variation of data within fields and classify sites into zones of broad similarity to support crop management. However, most of the statistical techniques available for spatial data require advanced skills. Cleaning and statistical analysis of big spatial data can only be effectively implemented if there are easy-to-use computer programs integrating the analytical steps. This paper explains the development and implementation of FastMapping, an interactive web application to automatically clean PA raw data, generate spatial variability field maps, and delineate multivariate management zones. The application uses an interface developed in R language that automatically depurates datafiles; in addition, through automatic spatial interpolation of each variable, it allows us to merge data layers on the same spatial grid in such a way that each site within the agricultural field has values of all the measured variables. On this grid, FastMapping identifies homogeneous zones, in a multivariate and spatial way, and provides data reports including validation of the delineated zones. FastMapping outputs for zone delineation were compared with results from Management Zone Analyst (MZA) software in several fields. The new software yielded a similar zoning to that of MZA and added graphs and statistical results that are lacking in most PA software tools. The flexibility of FastMapping to import, clean and visualize PA data in a single computer environment is highly encouraging. FastMapping can be used by the PA community to support site-specific agricultural management.

中文翻译：

---

FastMapping：用于在精准农业中创建田间地图和识别管理区域的软件

摘要 通过精准农业 (PA) 收集的各种农场农艺数据需要净化和联合分析多种地理参考田间特征，如作物产量、土壤和地形方面。已建议对空间变异进行多变量分析，以了解田间数据的变化，并将场地分类为具有广泛相似性的区域，以支持作物管理。但是，大多数可用于空间数据的统计技术都需要高级技能。大空间数据的清洗和统计分析只有在有易于使用的计算机程序集成分析步骤的情况下才能有效实施。本文解释了 FastMapping 的开发和实现，这是一个交互式 Web 应用程序，用于自动清理 PA 原始数据、生成空间可变性场图、并划定多元管理区。该应用程序使用 R 语言开发的接口，可自动净化数据文件；此外，通过每个变量的自动空间插值，它允许我们合并同一空间网格上的数据层，这样农田内的每个站点都有所有测量变量的值。在这个网格上，FastMapping 以多元和空间的方式识别同质区域，并提供数据报告，包括对划定区域的验证。区域划分的 FastMapping 输出与多个领域的管理区域分析 (MZA) 软件的结果进行了比较。新软件产生了与 MZA 类似的分区，并添加了大多数 PA 软件工具所缺乏的图表和统计结果。FastMapping 导入的灵活性，在单个计算机环境中清理和可视化 PA 数据非常令人鼓舞。PA 社区可以使用 FastMapping 来支持特定地点的农业管理。