Surface transient storage in dead zones: Residence times from stream morphology, velocity and CFD modeling

河流中的瞬变储存是指水和溶质在涡旋和其他“死区”(地表瞬变储存，简称STS)和溪流的次表层(海流暂态储存，简称HTS)中的暂时滞留。暂态储存区是河流中新陈代谢的主要场所，对河流中热量和污染物的去向和传输也非常重要。然而，科学家们还没有简单、可靠的方法来定量分离STS和HTS。此外，在STS中，我们不能非常准确地知道停留时间分布(RTD)及其平均值如何与诸如水流速度、死区大小和旋涡数量等物理特性有关，尽管这种关系是必然存在的并且是直观的。初步现场试验表明，在考虑物理尺寸和流速差异后，STS死区的平均停留时间比人工渠道死区和丁坝工程死区的平均停留时间有更大的差异。大涡模拟(LES)模型的初步工作表明，使用这种先进的计算流体力学(CFD)方法来模拟和理解自然河流中的死区停留时间是可行的。我们将对两次排放(总共10次)下的5个死区进行详细的工作，以表征它们的物理尺寸、粗糙度、速度和停留时间。其中三个将用LES进行模拟。反过来，LES将被用于训练精度较低但成本较低的雷诺平均纳维斯托克斯(RANS)模型的相同死区。将建立所有10个死区的RANS模型，并测量RTD和平均停留时间对每个物理特性的敏感性。我们将使用LES和RANS模型来建立RTD和死区物理特性之间的定量关系。这种关系将在20个死区上进行测试，这些死区的物理特征和RTD已被测量，并且没有CFD模型。RTD关系将基于现场可测量的参数，不需要CFD模型。为了了解RTD关系的限制，我们将在现场针对一些非理想的STS特征进行测试，例如具有大木材或巨石的STS特征，以及速度更快、更大的溪流中的STS特征。除了在不同的STS站点上进行的3个大涡模拟外，我们还将对大涡模拟的结果进行尺度分析，以预测更高速度和更大的雷诺数的结果(以及RTD关系的限制)。最后，我们将开发一个基于物理的STS分类方案。这将使我们能够用易于现场使用的定性和定量的术语来说明RTD关系的界限，这也将有助于水文学家和河流生态学家就STS进行交流。STS的定量RTD关系将使水文学家和河流生态学家能够更好地预测河流和河流中营养物质、污染物和热量的运动。反过来，这将有助于管理墨西哥湾的缺氧、溪流中的营养物质负荷、过高的溪流温度以及污染物泄漏的清理。RTD关系的发展将培养几名水文学和CFD建模方面的研究生和本科生，并将推动CFD建模领域的发展。我们将举办一个关于水文及相关科学的CFD建模的短期课程和会议。

Transient storage in streams is the temporary retention of water and solutes in eddies and other "dead zones" (surface transient storage, or STS) and in the stream's subsurface (hyporheic transient storage, or HTS). Transient storage zones are where most of metabolism takes place in streams, and they are also very important to the fate and transport of heat and pollutants in streams. However, scientists have no easy, reliable method for quantitatively separating STS from HTS. Furthermore, in STS we do not know very precisely how the residence time distribution (RTD) and its mean relate to physical characteristics such as stream velocity, size of the dead zone, and number of eddies, even though such a relationship must exist and is intuitive. Preliminary field work indicates that the mean residence time in STS dead zones has greater variance (after accounting for differences in physical dimensions and velocity) than in dead zones in artificial channels or engineered dead zones in groynes. Preliminary work with a large-eddy simulation (LES) model suggests that it will be feasible to use this advanced computational fluid dynamics (CFD) method to simulate and understand dead zone residence times in natural streams.We will do detailed work on 5 dead zones at two discharges (for a total of 10) to characterize their physical dimensions, roughness, velocities, and residence times. Three of these will be simulated with LES. The LES, in turn, will be used to train less precise but cheaper Reynolds Averaged Navier-Stokes (RANS) models of the same dead zones. RANS models of all 10 dead zones will be developed, and the sensitivity of the RTD and mean residence time to each physical characteristic will be measured. We will use the LES and RANS models to develop a quantitative relationship between the RTD and the physical characteristics of the dead zone. This relationship will be tested on 20 dead zones where the physical characteristics and RTDs have been measured and that have no CFD model. The RTD relationship will be based on field-measureable parameters, and will not require a CFD model. To understand the limits of the RTD relationship, we will test it in the field against a number of non-ideal STS features, such as those with large wood or boulders and those in higher-velocity, larger streams. In addition to the 3 LES simulations on different STS sites, we will do a scaling analysis on the LES results to much higher Reynolds numbers to predict results (and the limits of the RTD relationship) for higher-velocity and larger streams. Lastly, we will develop a physics-based classification scheme for STS. This will allow us to state the limits of the RTD relationship in terms that are qualitatively and quantitatively easy to use in the field, and it will also help hydrologists and stream ecologists to communicate about STS.A quantitative RTD relationship for STS will allow hydrologists and stream ecologists to do a better job at predicting the movement of nutrients, pollutants, and heat in streams and rivers. This, in turn, will help with management of problems such as hypoxia in the Gulf of Mexico, nutrient loading in streams, excess stream temperature, and the cleanup of pollutant spills. The development of the RTD relationship will train several graduate and undergraduate students in both hydrology and CFD modeling, and will advance the field of CFD modeling. We will run a short course and conference on CFD modeling in hydrologic and allied sciences.