H53A-0957
Spatial modeling of coupled hydrologic-biogeochemical processes for the Southern Sierra Critical Zone Observatory
One of the primary roles of modeling in critical zone research studies is to provide a framework for integrating field measurements and theory and for generalizing results across space and time. In the Southern Sierra Critical Zone Observatory (SCZO), significant spatial heterogeneity associated with mountainous terrain combined with high inter-annual and seasonal variation in climate, necessitates the use of spatial-temporal models for generating landscape scale understanding and predictions. Science questions related to coupled hydrologic and biogeochemical fluxes within the critical zone require a framework that can account for multiple and interacting processes. One of the core tools for the SCZO will be RHESSYs (Regional hydro-ecologic simulation system). RHESSys is an existing GIS-based model of hydrology and biogeochemical cycling. For the SCZO, we use RHESSys as an open-source, objected oriented model that can be extended to incorporate findings from field-based monitoring and analysis. We use the model as a framework for data assimilation, spatial-temporal interpolation, prediction, and scenario and hypothesis generation. Here we demonstrate the use of RHESSys as a hypothesis generation tool. We show how initial RHESSys predictions can be used to estimate when and where connectivity within the critical zone will lead to significant spatial or temporal gradients in vegetation carbon and moisture fluxes. We use the model to explore the potential implications of heterogeneity in critical zone controls on hydrologic processes at two scales: micro and macro. At the micro scale, we examine the role of preferential flowpaths. At the macro scale we consider the importance of upland-riparian zone connectivity. We show how the model can be used to design efficient field experiments by, a-priori providing quantitative estimate of uncertainty and highlighting when and where measurements might most effectively reduce that uncertainty.
H53A-0958
A role for high frequency hydrochemical sampling in long term ecosystem studies
Monitoring of surface waters for major chemical constituents is needed to assess long-term trends and responses to ecological disturbance. However, the typical fixed-interval (weekly, monthly, or quarterly) sampling schemes of most long-term ecosystem studies may not capture the full range of stream chemical variation and do not always provide enough information to discern the landscape processes that control surface water chemistry and solute loadings. To expand upon traditional hydrochemical monitoring, we collected high frequency event-based surface water samples at an upland, forested basin of the Sleepers River Research Watershed (Vermont, USA), one of five intensively studied sites in the Water, Energy, and Biogeochemical Budgets (WEBB) program of the US Geological Survey. We present several examples that highlight the importance of linking long-term weekly data with intensive, high frequency sampling. We used end-member mixing analysis and isotopic approaches to trace sources of stream nutrients (e.g. nitrate, dissolved organic carbon) and quantified how atmospheric pollutants (e.g. nitrogen, sulfate, and mercury) affect stream chemistry. High frequency sampling generates large numbers of samples and is both labor and resource intensive but yields insights into ecosystem functions that are not readily discerned from less-frequent sampling. As the ecological community contemplates the scope and foci of environmental observatories as benchmarks for deciphering the effects of natural and anthropogenic change, incorporating high frequency hydrochemical sampling will further our understanding of ecosystem functions across a range of ecosystem types and disturbance effects.
H53A-0959
Ground-Water Sources, Flow-Paths, and Residence Times in the Middle Verde River Watershed, Northern Arizona
Geochemical tracers serve as valuable tools for characterizing basin hydrogeology. By combining stable and radioactive isotopic analyses with solute concentrations and discharge data, one can constrain hydrologic flow- paths and water sources in an area of complex hydrogeology. These techniques are applicable to the Middle Verde River watershed, a region located in the transition zone between the Southern Colorado Plateau and the Basin and Range structural provinces. As the population within the Verde River Valley is projected to double by 2050, efforts to improve the conceptual understanding of basin hydrogeology and to quantify recharge rates within the watershed are critical for water resources management. The primary objective of the investigation is to determine the hydrologic connection between aquifers underlying the Colorado Plateau and adjacent aquifers in the Verde River watershed through analysis of oxygen and hydrogen stable isotopes, tritium, carbon-14, and major solute concentrations. The secondary objective is to gain an understanding of how these water sources and flow-paths contribute to and sustain Verde River base-flow. Two surface-water datasets collected from the Middle Verde River and its tributaries (Oak Creek, Wet Beaver Creek, and West Clear Creek) in November 2006 and June 2007 serve as snapshots of winter and summer base-flow conditions, respectively. Ground-water samples complement these datasets by serving as end members for base-flow source mixing models. Preliminary analyses based on solute relationships (i.e. chloride- sulfate and bromide-chloride) show evidence of separate solute sources for the Verde River and its tributaries. The distinct Verde River trends, including overall increases in solute concentrations along two reaches (kilometers 15 to 30 and 57 to 66, as measured upstream from USGS gauge 09506000), suggest dissolution of evaporite deposits within the Tertiary lakebed-derived Verde Formation. Notably, ground-water from wells drilled in the Verde Formation near the Verde River is geochemically more similar to surface-water from the river's tributaries than to the river itself. This suggests that the Verde Formation is supplied by waters from the surrounding formations and this water then undergoes evolution based on the geologic variability of the Verde Formation. Seasonal sources of base-flow are constrained by oxygen and hydrogen stable isotope values: mixing models using regional precipitation averages as end members show greater contribution of winter precipitation than summer precipitation.
H53A-0960
Examining the Effects of Geomorphology on Hydrologic Transit Times Using Liquid Water Isotopes
In recent years there has been resurgence in improving physically based and spatially distributed hydrological response models. However there continues to be many obstacles in accurately representing the basic processes governing rainfall runoff responses. Much of these inaccuracies can be attributed to such problems as a lack of understanding in runoff processes, unknown heterogeneity both at the surface and subsurface, variations in driving forces and the effects of geomorphology on the transformation of rainfall to stream flow. We hope to improve on such ambiguity is by examining the relationships between geomorphology and hydrology through the investigation of transit time distributions, which can be used as a fundamental descriptor of catchments" characteristics such as storage and flow pathways. By examining stable isotopic variability in precipitation, soil moisture and stream flow to determine transit times, we hope to better understand the effects of topographic land structures on the hydrologic response system. The first step in this process has been to fully instrument a series of hill slopes with similar structural and pedologic characteristics, located in the Marshall Gulch region of the Santa Catalina Mountains. Equipment including suction and non- suction lysimeters, tipping bucket rain gauges and automatic flow samplers with data loggers positioned to take stream flow and precipitation samples have been used to collect samples throughout the region. A description of preliminary results will be presented.
H53A-0961
Modelling Stream Aquifer Interactions During Floods and Baseflow Upper San Pedro River, Southeastern Arizona
Streams and groundwaters may interact in distinctly different ways during flood versus base flow periods. Flooding periods may recharge riparian aquifer systems with this water destined to sustain baseflows in a river for long periods after the flood wave has passed. The riparian ecosystem along the Upper San Pedro River (southeastern Arizona) is one of only a few free flowing rivers in the Southwest. Recent research using isotopic and chemical data shows that (1) near-stream, or ‘riparian,' groundwater recharged during high streamflow periods is a major contributor to streamflow for the rest of the year, and (2) the amount of riparian groundwater derived from this flood recharge can vary widely (10-90%) along the river. Riparian groundwater in gaining reaches is almost entirely basin groundwater, whereas losing reaches are dominated by prior streamflow. The above results give rise to the questions of (1) how much flood recharge occurs at the river-scale, and (2) subsequently, what is the relative importance of flood recharge and basin groundwater in maintaining the hydrologic state of the entire riparian system. To address these questions, a coupled hydrologic-solute model was constructed for 45 Km of the riparian system. The model domain is divided into segments, with each segment representing a distinctly gaining or losing reach. Surface-subsurface water exchange is regulated by hydraulic properties of the system based on observed groundwater level response to flood waves. Daily discharge data at three points and chemical/isotopic river and groundwater data at various locations along the river were used to calibrate the model from 1995 to the present. Model results suggest good agreement between our model and the overall hydrologic and chemical/isotopic behavior patterns of the riparian system. Less than 51% of total summer flood recharge occurs in the upper two- thirds of the river, where gaining conditions dominate. However, recharge in this upper portion of the river on average accounts for more than 34% of the fall, winter and spring baseflow. Flood recharge along the lower losing reaches maintains the shallow water tables in the riparian aquifer that are essential to much of the vegetation composing the riparian forest.
H53A-0962
A Simulation Framework for Evaluating Sampling Strategies and Determining Load Accuracies in Suspended Sediment Loads
Excessive river sedimentation can cause extensive economic and ecological damage. Expensive dredging operations are needed to keep navigation channels clear and to maintain the capacity of water supply reservoirs. Deposition of fine sediments in rivers can eliminate pool habitats, decrease embryo survival rates of certain fish, and affect macroinvertebrate density and diversity. Sedimentation is often associated with anthropogenic watershed activities (e.g. urbanization and agricultural practices). Effort has been spent on developing Best Management Practices (BMPs) to reduce the sediment loads caused by specific watershed activities. Sediment monitoring networks have also been implemented to measure loads within streams and help determine the efficacy of BMPs over time. Yet, fundamental questions remain regarding how accurately loads can be estimated. Research suggests watershed hydrologic and geomorphic characteristics, sampling method and frequency, along with the method used to develop sediment-discharge rating curves can substantially affect the accuracy and precision at which sediment load estimates are made. The confidence at which one can estimate sediment loads, based on a specific sampling protocol, is one of several important pieces of information that hydrologic observatories need to understand in order to help monitor load trends. A computer program is being developed that allows one to estimate sediment loads using several sediment- discharge rating curves and bias correction factors. Using USGS mean daily sediment data for the Illinois River at Valley City, the program is employed to perform Monte Carlo simulations to predict confidence limits for loads estimated using different sampling protocols (e.g. weekly, monthly and hydrologic event based sampling). Results of the different sampling approaches are compared. A discussion regarding how these results, combined with future simulations representative of different sediment monitoring locations, can help guide future monitoring efforts is provided.
H53A-0963
Optimization of geometry and modeling parameters of artificial neural networks using genetic algorithms
In recent years, artificial neural networks (ANNs) appear to be viable alternative to models that use phenomenological hypotheses (i.e. knowledge based models) for cases (1) the available data are not detailed and sufficient for using a process based model and (2) the detailed complex physics of the system is partially understood. ANNs have been widely used in many fields such as chemical and environmental engineering, hydrology, and water resources applications for optimum prediction of system parameters and variables. However, in most cases, parameters and system variables were forecasted employing suboptimal ANNs. The geometry and modeling parameters of an artificial neural network (ANN) and the training dataset have significant effects on its predictive performance efficiency. The combination of ANN modeling parameter and geometry arranged in the modeling domain (i.e. lower and upper bounds of each modeling parameter and geometry) is large enough (i.e. greater than 100000) that it is difficult to examine all cases using trial and error approach for the selection of an optimum set. Thus, one could easily end up with finding a set of suboptimal values. This study presents the use of genetic algorithms (GAs) to search for the optimal geometry and values of modeling parameters of a multilayer feedforward backpropagation neural network (BPNN) and a radial basis function network (RBFN). The predictive performance efficiency of the GA and ANN combination is examined using two datasets derived from the same population for training. It is illustrated that (1) the GA optimized ANN outperforms to the ANN using a trial and error approach, and (2) ANN predictive performance and geometry depend on the number of samples and the characteristics of samples included in the training dataset.
H53A-0964
Quantitative Analysis of Cations in Solutions Using ion Exchange Filter Paper by X-ray Fluorescence Spectrometry
The strong acid cation exchange filter, CP-1, were applied to X-ray fluorescence (XRF) for trace elements in aqueous solution. In this study, the objective elements were Cu, Zn, Pb, and Cd. The measurement of the X-ray intensity determined the optimum methods of XRF analyses using the CP-1 filter. The optimum area that sample solutions penetrated was inside the determining area. The proper material to underlay CP-1 filter was glass bead. The calibration curves that determined by this study calculated the limit values of determination. These values were 0.72 micro gram for Cu and Zn. The method of separation and concentration using the CP-1 filter enable the quantitative analyses of trace heavy metals easily.