Development of a unified framework for environmental monitoring based on a global sampling grid system (GSG)

环境监测已成为许多与生态和环境挑战有关的国际和国家进程的关键因素。防止毁林和森林退化的斗争（REDD+）只是一个例子，表明监测森林资源的状况和变化是至关重要的，因为当国家或项目为避免/减少毁林和森林退化所做的努力应得到补偿时，作为基于绩效的支付的基础。基于模型的遥感分析产生了许多关于森林覆盖变化或碳储量的完整地图产品，这些产品可作为确定全球森林砍伐水平的基准，特别是在热带地区。通过谷歌Earth、Microsoft virtual Earth （bing）或NASA World Wind等虚拟地球仪，大量免费提供的高分辨率和地理参考图像的可用性不断增加，但在全球环境监测背景下的科学应用仍相对未得到充分利用。它们可以作为基于设计的抽样研究的基础，通过对遥感图像中相对较小的观测单元进行视觉解释。一个统一的估算和数据存储框架将有助于利用这些全球数据源，并整合来自不同区域和空间范围的研究。该项目旨在提出一个可扩展的仿射全球抽样设计，并将为其应用制定一个概念性框架，重点是森林资源。特别强调利用可免费获得的遥感图像。计划项目的一个科学关键要素是发展方法基础，以优化观测单位，以目视解释遥感图像。观测设计与分类/解释图像单元的可靠性之间存在相互作用。项目的这个维度是通过研究基于设计的推理和容易出错的观察来解决的。

Environmental monitoring has been emerging as key factor in many international and national processes related to ecological and environmental challenges. The combat against deforestation and forest degradation (REDD+) is but one example showing that monitoring of the state and change of forest resources is essential as a basis for performance based payments when countries or projects shall be compensated for their efforts in avoiding/reducing deforestation and forest degradation. Many wall-to-wall map products of forest cover change or carbon stocks have been generated by model-based remote sensing analysis that shall serve as benchmark to define global deforestation levels, especially for tropical regions. The increasing availability of large archives of freely available high resolution and geo-referenced imagery through virtual globes like Google Earth, Microsoft Virtual Earth (bing) or NASA World Wind and others is still relatively underexploited by scientific applications in this global environmental monitoring context. They could serve as basis for design-based sampling studies by visual interpretation of relatively small observation units in remotely sensed imagery. A unified framework for estimation and data storage would help to utilize these global data sources and to integrate research studies from different regions and spatial extent. The project aims at proposing a scalable and affine global sampling design and will develop a conceptual framework for its application with a focus on forest resources. Special emphasis is laid on utilizing freely available remote sensing imagery. A scientific key element of the planned project is the development of methodological basis for the optimization of observation units for visual interpretation of remote sensing imagery. There is an interplay between the observation design and the reliability of classified/interpreted image units. This dimension of the project is addressed by researching into aspects of design-based inference with error-prone observations.