Design and demonstration of a distributed sensor array for predicting water flow and nitrate flux in the Santa Fe Basin

0609968格雷厄姆流域是动态的。因此，对流域关键科学问题的回答需要流域特征的观测，并依赖于开发新的工程能力，以高时间频率和空间密度采样数据;评估采样的最佳频率和密度;并分析这些数据，以提高对过程的机械理解。需要改进采样和分析能力的流域的基本时间相关特性的例子包括：存储中的水、营养物、沉积物和能量的质量大小;存储中的停留时间;存储之间的通量;以及存储之间的流路。这项工作提出了新的工程方法，以解决科学问题的空间和时间变化的监管机构的硝酸盐负荷在流域;地质环境，土地利用和地形如何影响加载和交付动态;流动混合物和硝酸盐负荷之间的关系;和使用替代措施的可行性估计流域尺度硝酸盐动态的空间和时间。工程方法是双管齐下的。（1）市场上可买到的硝酸盐、电导率、温度和压力传感器将被安装在一个阵列中，该阵列将以低于日常的频率进行采样，并通过手机技术与中央计算机进行真实的实时通信。(2)现有的数据将被用来开发一个probabilisticalgorithm的基础上，信息论，以确定最小的空间和时间间隔的采样必要，以减少不确定性，以规定的水平，这个算法将与新收集的data.The研究网站是圣达菲河流域中北部的佛罗里达。圣菲河是苏瓦尼河的主要支流，苏瓦尼河被认为是CUAHSI清洁水和环境研究系统网络的潜在地点。圣菲河具有与萨旺尼河流域相似的特征，但规模较小，使其成为开发推进流域科学和工程所需的新采样和分析能力的理想试验台。这两条河流都穿过科迪悬崖，这标志着地质封闭和非封闭部分的弗洛里丹含水层之间的边界，以及从地表水到地下水为主的系统的变化。整个地区正在经历从最小的人为影响到土地使用实践的多种压力因素的过渡，包括人口增长和消费性使用，拟议的水转移，以及加速发展，使得预测流域规模的科学至关重要。传感器网络和概率算法的开发代表了如何开发水和环境研究网络的小规模良好的初步测试。通信端口将被配置为测试任何新开发的或现成的传感器，并欢迎外部用户在这些站点测试仪器。作为这项工作的一部分，将开发的概率算法将是足够通用的，他们将是有用的规划传感器部署在任何流域。这一通用特性将使他们有用的分析变量的可预测性的不确定性，除了硝酸盐。人力资源将通过支持两个博士水平的研究生（和博士后研究员）的影响。为这项工作收集和整理的数据将存档在ArcHydrogeodatabase中，供非当地研究人员使用。这些数据也将被演示用于AP环境科学类在高中水平，环境科学和水文课程在大学水平，并作为一个互动展览在佛罗里达博物馆的自然历史盖恩斯维尔。

0609968Graham Watersheds are dynamic. Consequently, answers to critical sciencequestions about watersheds require observations of watershed characteristics and depend ondeveloping new engineering capabilities to sample data at high temporal frequency and spatialdensity; to assess the optimal frequency and density of sampling; and to analyze those data for theimprovement of mechanistic understanding of processes. Examples of fundamental timedependentcharacteristics of watersheds that require improved sampling and analyticalcapabilities include the magnitude of mass of water, nutrients, sediments, and energy in stores;the residence time in the stores; fluxes between the stores; and flowpaths among the stores. Thisproposed work develops new engineering approaches to address science questions about thespatial and temporal variability of regulators of nitrate load in watersheds; how geologic setting,land use and terrain influence loading and delivery dynamics; the relationship between flowmixtures and nitrate loads; and the feasibility of using surrogate measures for estimating basinscale nitrate dynamics in space and time. The engineering approaches are two-pronged. (1)Commercially available nitrate, conductivity, temperature, and pressure sensors will be installedin an array that will sample at sub-daily frequencies and communicate via cell phone technologyin real time to a central computer. (2) Existing data will be used to develop a probabilisticalgorithm based on information theory to identify minimum spatial and temporal spacing ofsampling necessary to reduce uncertainty to a prescribed level; this algorithm will be validatedwith the newly collected data.The study site is the Santa Fe River watershed in north-central Florida. The Santa FeRiver is a major tributary to the Suwannee River which has been cited as a potential location forCUAHSI-CLEANER WATer and Environmental Research Systems Network. The Santa FeRiver has similar characteristics to the Suwannee River Basin, but at a smaller scale, making it anideal test-bed for developing new sampling and analytical capabilities required to advancewatershed science and engineering. Both rivers cross the Cody escarpment, which marks theboundary between geologically confined and unconfined portions of the Floridan aquifer, and thechange from a surface water to ground water dominated system. The entire region is undergoinga transition from minimal anthropogenic impact to multiple stressors from land-use practicesincluding population increase and consumptive use, proposed water diversions, and acceleratingdevelopment making predictive basin-scale science essential.Broader Impacts. Development of the sensor network and probabilistic algorithm represent agood initial test at a small scale of how to develop water and environmental research networks.The communication ports will be configured for testing of any newly developed or off-the-shelfsensors, and outside users will be welcome to test instrumentation at these sites. The probabilisticalgorithms to be developed as part of the work will be sufficiently generic that they will be usefulfor planning sensor deployments in any watershed. This generic characteristic will make themuseful for analyses of the uncertainty of predictability of variables in addition to nitrate. Humanresources will be impacted through support of two PhD-level graduate students (and a postdoctoral researcher). Data collected and collated for the work will be archived in an ArcHydrogeodatabase to be made available to non-local researchers. These data will also be demonstratedfor use in AP Environmental Science classes at the high school level, Environmental Science andHydrologic courses at the college level, and as a an interactive exhibit at the Florida Museum ofNatural History in Gainesville.