Abstract Soil information with high spatial and temporal resolution is crucial to assess potential soil degradation and to achieve sustainable productivity and ultimately food security. The spatial resolution of existing soil maps can commonly be too coarse to account for local soil variations and owing to the cost and resource needs required to update information these maps lack temporal information. With improved computational processing capabilities, increased data storage and most recently, the increasing amount of freely available data (e.g. Landsat, Sentinel-2A/B), remote sensing imagery can be integrated into existing soil mapping approaches to increase temporal and spatial resolution of soil information. Satellite multi-temporal data allows for generating cloud-free, radiometrically and phenologically consistent pixel based image composites of regional scale. Such data sets are of particular use for extracting soil information in areas of intermediate climate where soils are rarely exposed. The Soil Composite Mapping Processor (SCMaP) is a new approach designed to make use of per-pixel compositing to overcome the issue of limited soil exposure. The objective of this paper is to demonstrate the automated processors ability to handle large image databases to build multispectral reflectance composite base data layers that can support large scale top soil analyses. The functionality of the SCMaP is demonstrated using Landsat imagery over Germany from 1984 to 2014 applied over 5 year periods. Three primary product levels are generated that will allow for a long term assessment and distribution of soils that include the distribution of exposed soils, a statistical information related to soil use and intensity and the generation of exposed soil reflectance image composites. The resulting composite maps provide useful value-added information on soils with the exposed soil reflectance composites showing high spatial coverage that correlate well with existing soil maps and the underlying geological structural regions.

中文翻译：

---

构建裸露土壤复合处理器 (SCMaP)，用于使用 Landsat 图像绘制土壤的时空特征（1984-2014）

摘要 具有高时空分辨率的土壤信息对于评估潜在的土壤退化、实现可持续生产力和最终实现粮食安全至关重要。现有土壤图的空间分辨率通常太粗糙，无法解释当地土壤变化，并且由于更新信息所需的成本和资源需求，这些地图缺乏时间信息。随着计算处理能力的提高、数据存储的增加以及最近免费可用数据量（例如 Landsat、Sentinel-2A/B）的增加，遥感图像可以集成到现有的土壤测绘方法中，以提高土壤的时空分辨率信息。卫星多时相数据允许生成无云、辐射测量和物候一致的基于像素的区域尺度图像合成。此类数据集特别适用于提取土壤很少暴露的中等气候地区的土壤信息。土壤复合映射处理器 (SCMaP) 是一种新方法，旨在利用逐像素复合来克服土壤暴露有限的问题。本文的目的是展示自动化处理器处理大型图像数据库的能力，以构建可支持大规模表层土壤分析的多光谱反射复合基础数据层。SCMaP 的功能使用 1984 年至 2014 年在德国上空应用 5 年的 Landsat 影像进行了演示。生成三个初级产品级别，允许对土壤进行长期评估和分布，包括暴露土壤的分布、与土壤使用和强度相关的统计信息以及暴露土壤反射率图像合成的生成。由此产生的复合图提供了有用的土壤增值信息，暴露的土壤反射复合图显示出高空间覆盖率，与现有土壤图和下层地质结构区域密切相关。