ITR: Development of an enhanced computer-assisted analysis system for earth science investigation of laminated sediments and tree rings

[019075]具有逐年分辨率的自然档案，如树木年轮、不同的沉积物、珊瑚、软体动物和洞穴动物，对于理解过去环境变异性的本质及其未来可能发生的变化变得越来越重要。这些档案最明显的用途是了解自然年际到世纪尺度气候变率的模式和原因，以及人类如何改变这种变率。然而，它们也是越来越广泛的地球科学应用的核心，包括记录古地震活动、校准放射性碳时间尺度、调查水文变化和理解古生态系统动力学。它们提供在许多与社会相关的时间尺度上理解这些动态所需的绝对时间尺度的能力尤为重要。目前，从这些来源提取信息是非常劳动密集型的，导致对时间分辨率和复制的不利限制，从而导致有关感兴趣的环境信号的过度不确定性。这一点甚至适用于使用复杂的探测器进行基本测量时，因为必须非常小心地将一系列测量结果串成很长的时间序列，同时保留所有可能在以后的分析中有用的信息。现有的图像分析软件本身并不能解决这个问题。这些记录需要的劳动密集型分析产生了强烈的动机来开发仪器，这些仪器可以用来更有效地处理样品，并且比现成的图像处理软件具有更强的解释能力。这项信息技术研究小额拨款计划（ITR）将支持开发用于自动计数年度分层气候变化代理（树木年轮和阀门）的软件。这些开发是基于计算机辅助树木年代学工作站的原型以及相关的控制和图像分析软件（TREES v. 1），该软件先前由美国国家科学基金会人类学项目（BCS-9601867）资助给Schowengerdt。该奖项将支持创建TREES v.2和VARVES v.1，前者是专门为SUN平台编写的基于linux的PC版本，后者也是基于linux的软件，旨在自动计数分层湖泊和海洋沉积物。＊＊＊

0219075OverpeckNatural archives with year-by-year resolution (isannual resolution proxieslo), such as tree-rings, varved sediments, corals, mollusks, and speleothems, are becoming increasingly critical for understanding the nature of past environmental variability, and how it may change in the future. The most obvious use of these archives is in understanding the patterns and causes of natural interannual to century-scale climate variability, and how humans may be altering this variability. They are also central, however, to an increasingly broad range of Earth Science applications, including documenting paleoseismicity, calibrating the radiocarbon time scale, investigating hydrologic variability, and understanding paleoecosystem dynamics. Their capability to provide the absolute time scale needed to understand these dynamics on many societally-relevant time scales is particularly important. At present, extracting information from these sources is extraordinarily labor-intensive, resulting in undesirable limitations on temporal resolution and replication that result in excessive uncertainties concerning the environmental signals of interest. This applies even when sophisticated detectors are used to make basic measurements, since great care must be taken in stringing series of measurements together into very long time series whilst retaining all information that might turn out to be useful later in the analysis. Existing image-analysis software alone cannot solve this problem. The labor-intensive analysis that these records require has created a strong incentive to develop instrumentation that can be used to process samples more efficiently, and with more interpretive power than is available from off-the- shelf image processing software.This Information Technology Research Small Grants Program (ITR) award will support the development of software for automated counting of annually laminated climate change proxies (tree rings and varves). The developments are based on a prototype computer-assisted dendrochronology workstation and associated control and image analysis software (TREES v. 1) that was previously funded to co-PI Schowengerdt by NSF's Anthropology program (BCS-9601867). This award will support the creation of TREES v.2, a Linux-based PC version of TREES v.1, which was written exclusively for the SUN platform, and VARVES v.1, also Linux-based software designed to automate counting of laminated lake and marine sediments. ***