Biodiversity and Ecosystem Informatics - BDEI - Spatio-temporal Models of Biogeophysical Fields for Ecological Forecasting: A Cross-Disciplinary Incubation Activity

EIA-0131937 Henebry，内布拉斯加州大学林肯分校BDEI：用于生态预测的生物地球物理场的时空模型：跨学科孵化活动摘要我们现在处于一个密集的地球观测时代：轨道平台产生大量跨越空间、时间、光谱和辐射分辨率的遥感数据流。在接下来的几年里，“天空之眼”的数量和种类将显著增加。这种名副其实的数据洪流需要新的思维方式，将遥感数据转换为关于生态模式和过程的信息。这些数据流为环境决策支持带来了希望。然而，迫切需要能够有效和可靠地表征图像时间序列中所包含的时空模式的理论和工具。我们认为，这些工具必须基于对陆地表面动态的生态预期，类似于气候学预期。生态预期将总结特定区域生物地球物理场空间格局的典型时间发展。我们有一个从遥感数据流中提取生态预期的稳健的主要方法：将图像时间序列投影到典型的公制空间。为了使生态预测成为一种可操作的可能性，我们需要建立和更新复杂的时空基线的能力，这将使预测和识别、量化和评估不寻常的情况成为可能。美国国家航空航天局最近举办的关于地球科学数据挖掘的讲习班确定异常检测是科学数据挖掘的一个关键特征；然而，生物地球物理数据的时空数据挖掘实例相对较少。我们的方法是时空数据挖掘，它由相关领域的专业知识提供信息。数据库中时空实体和字段的表示必须支持复杂的时空查询：这是目前尚不存在的功能。

EIA-0131937Henebry, GeoffreyUniversity of Nebraska - LincolnBDEI: Spatio-temporal models of Biogeophysical Fields for Ecological Forecasting:A Cross-Disciplinary Incubation Activity SummaryWe are now in an era of intensive earth observation: orbital platforms generate myriad remote sensingdatastreams across a range of spatial, temporal, spectral, and radiometric resolutions. The number andvariety of "eyes in the skies" are scheduled to increase significantly over the next few years. Thisveritable data deluge necessitates new ways of thinking about transforming remote sensing data intoinformation about ecological patterns and processes. These datastreams hold the promise forenvironmental decision support. Yet, there is a critical need for theories and tools that will enable efficientand reliable characterization of spatio-temporal patterns contained in image time series. We think thatsuch tools must be based on ecological expectations of land surface dynamics, analogous toclimatological expectations. Ecological expectations would summarize across specific regions the typicaltemporal development of spatial pattern in biogeophysical fields. We have a robust principal method forextracting ecological expectations from remote sensing datastreams: projecting image time series intopattern metric spaces. To make ecological forecasting an operational possibility, we need the capabilityto establish and to update complex spatio-temporal baselines that will enable prediction of the usual andidentification, quantification, and assessment of the unusual. A recent NASA workshop on Earth Sciencedata mining identified anomaly detection as a key characteristic of scientific data mining; yet, there arerelatively few examples of spatio-temporal data mining of biogeophysical data. Our approach is spatio-temporal datamining that is informed by relevant domain expertise. Representation of the spatio-temporalentities and fields in databases must support sophisticated spatio-temporal queries: a capability that doesnot currently exist.