Integrated Approaches in Hydrological Process Studies II Posters
Presiding: K McGuire, Georgia Institute of Technology; M Weiler, University of British Columbia
H21C-01 0830h
Graph Theory Data Objects Applied to Stream Flow Network Representation in an Integrated Hydrological Model
The Management Simulation Engine (MSE) component of the Regional Simulation Model (RSM) incorporates a multi-level hierarchical control architecture which emphasizes the decoupling of hydrological state information from the management information processing applied to the states. A crucial aspect of effectively storing and accessing state information for water resource management purposes is the maintenance of an efficient storage mechanism which associates hydrological state information with the proper managerial abstractions. In the RSM this is done by storing hydrological and managerial information relevant to a water control unit (WCU) in a data storage object defined in the MSE Network. The MSE Network is an abstraction of the stream flow network and control structures suited to the needs of water resource routing and decisions. It is based on a standard graph theory representation of a flow network comprised of arcs and nodes. The MSE Network data objects serve as state and process information repositories for management processes. They maintain appropriately filtered state information, parameter storage relevant to WCU or hydraulic structure managerial constraints and variables, and serve as an integrated data source for any MSE algorithm. It also provides a mathematical representation of a constrained, interconnected flow network which facilitates efficient graph theory solutions of network connectivity and flow algorithms. This paper describes the MSE Network implementation in the RSM, an integrated hydrological computation engine aimed at meeting the needs for comprehensive integration of management features in coupled hydrological models [1]. [1] Belaineh, G., Peralta, R. C., Hughes, T. C., Simulation/ Optimization Modeling for Water Resources Management, ASCE Journal Water Resources Planning Management, 125(3), p 154-61, 1999
H21C-02 0830h
Explore a Fine Resolution MM5 for Reservoir Temperature Forecasting
TVA's three nuclear plants and some fossil plants rely on reservoir water for cooling. When weather is hot during summer, variable meteorological conditions such as wind and cloud cover often affect reservoir water and make its temperature prediction particularly challenging. A critical point is occasionally reached when a decision must be made on whether to derate a generating unit to comply with thermal discharge restrictions. The decision to derate can depend on differences of a few tenths of a degree in the water temperature prediction. Predicted reservoir temperatures for cooling of a TVA nuclear plant, for example, is produced by hydrodynamic models with input meteorological condition for the entire lake approximated by a National Weather Service Forecast Office (NWSFO) forecast issued for the closest airport to the water discharge location. One of the potential areas for improving reservoir temperature prediction is the non-homogeneous meteorological forcing over the body of water. A high resolution PSU/NCAR MM5 model was explored for improving reservoir temperature forecasting. Three summer periods representing extremely hot weather conditions were selected for MM5 simulations over the cooling water source of one TVA nuclear plant. The MM5 was configured with five concentric modeling domains. Grid resolution ranged from 27 km for the coarsest to 0.3 km for the finest modeling domain. Model results from the finest resolution domain were evaluated and analyzed with observations from a meteorological tower of the plant and reservoir temperature profiles from permanent buoys near the plant. Preliminary results indicate that the fine resolution MM5 model is capable of simulating large air temperature variations resulting from extreme weather events. The model can also simulate non-homogeneous weather conditions and produce different wind patterns across the reservoir that can affect water temperature through wind induced turbulent mixing. Detailed modeling results and analyses will be presented at the joint assembly.
H21C-03 0830h
Improved Flood Prediction in an Urban Watershed Using a Physically-Based Modeling Approach
A modeling strategy has been developed to improve the real-time forecasting of medium to large magnitude floods in an urban watershed, Brays Bayou (260 km2), in Houston, TX. Severe flooding of downstream areas of the watershed around the Texas Medical Center (TMC) has been an increasing problem over the past few decades, particularly during Tropical Storm (TS) Allison in June 2001. Since late 2003, the physically-based distributed hydraulic model, VfloTM, using a relatively coarse (122m) finite element grid, has been run alongside HEC-1 as an ensemble for real-time flood forecasting at strategic locations on the main channel. The models are driven by local NEXRAD L2 radar data. Both VfloTM and HEC-1 are incorporated into the Flood Alert System 2 (FAS2), which was developed for the TMC through Rice University. Both FAS2 and its predecessor FAS have done well in predicting the magnitude and timing of flooding. However, during TS Allison, stormwater in Harris Gully (20.7 km2) overflowed, resulting in millions of dollars in damages. In addition, the City of Houston is currently installing new stormwater conduits which will change the pattern of flow from Harris Gully to the larger Brays Bayou. The SWMM model has been applied to this subcatchment; however its long run-time precludes its use for real-time forecasting. It is intended to develop a more detailed VfloTM model of the smaller Harris Gully subcatchment to explore several questions, including: 1) in order to improve real-time forecasts further for the TMC, highly accurate quantitative precipitation estimate data at the spatial resolution relevant for urban areas (e.g. a city block) and a distributed hydrologic model of the same spatial resolution are necessary, 2) even in heavily urbanized areas, dependence on flood properties on a basin scale derives from the space-time scaling properties of rainfall, 3) urbanization, particularly in small (less than 100 km2) basins, alters traditional scaling theories of flood response. The Engineering Research Center for Collaborative Adaptive Sensing of the Atmosphere (CASA) aims to provide a new archetype of precipitation product data for the TMC region of downtown Houston. Data available through CASA will enable the research questions to be addressed. In addition, improvement in real-time flood forecasting attributable to the development of new weather radar precipitation estimation (QPE) techniques will be assessed.
H21C-04 0830h
Geology and Geophysical Surveys to Infer the Structure of the Upper San Pedro River Basin, Sonora, Mexico for Use in a Ground-Water-Flow Model
Data on geology and geophysics in the San Pedro River Basin in Sonora, Mexico were combined to develop a three-dimensional conceptual model of the alluvial-fill aquifer in the basin that is being used to construct a regional ground-water-flow model. In Mexico, the headwater region of the river encompasses approximately 1,800 square kilometers of an ungaged catchment system. This feeds a 58 kilometer-long series of intermittent and perennial stream reaches in the United States that extend from just north of the international border to the town of St. David, Arizona. The river forms part of a north-south riparian corridor that provides habitat for more than 100 resident and 250 migratory bird species. Ground water in the basin is used extensively on both sides of the border and information on basin structure and composition will help to address questions regarding ground- and surface-water sustainability and planning. Interpretations of bedrock and alluvial-fill geometry indicate that a significant portion of the catchment area in Mexico is underlain by bedrock composed of highly indurated (compacted) Cretaceous sedimentary, volcanic, volcano-sedimentary, and granitic intrusive rocks. Aeromagnetic surveys were used to estimate depth to bedrock underlying alluvial sediments. Satellite photographs, older geologic maps, and recent field observations were used to delineate the boundaries between bedrock and alluvium. About 655 square kilometers, or 36 percent, of the Mexican portion of the river basin is underlain by alluvial fill. In the southern part of the study area, detailed information on thickness and composition of subsurface layers to depths of 500 meters was derived from drill logs. An extensive network of vertical electrical soundings covering much of the central part of the basin allowed for estimates of the location and thickness of clay layers that are confining units within the aquifer system. Across much of the area, the thickness of the silt and confining units was difficult to determine because of problems in distinguishing between these layers and underlying, electrically-conductive Cretaceous siltstone and mudstone. In general, two hydraulically connected sub-basins were identified: one in the southern part of the study area and one in the northern part.
H21C-05 0830h
An Integrated Approach for Assessment of Groundwater Potentialities Arid Areas
We applied an integrated approach to assess the ground water potentiality in the Eastern Desert (ED) of Egypt. Inferences from a wide range of data sets (Remote sensing: [Landsat TM, SIR-C, SRTM, TRMM]; geochemical [O, H stable isotope composition, solute chemistry], geological [structures, lithology], hydrological [lithology, water table, etc.,] were used to address the nature and sources of groundwater in the Eastern Desert (ED) and to determine the groundwater potentialities across area. Data integration, analysis, and visualization was conducted in a GIS environment. Three potential ground water sources were identified. The first is the shallow alluvial aquifers that fossil water of the Nubian Aquifer. The Nubian Aquifer deep water access the deep and sub-vertical fault systems defining the River Nile graben, and the Gulf of Suez graben. Criteria used in the definition of these reservoirs are: (a) isotopic compositions similar to those of paleo-water in the Eastern Desert, (b) NW-trending fault systems intersecting or bounding the alluvial sediments in question, and (c) presence of thick alluvial deposits proximal to the deep seated faults. The second is meteoric precipitation over the Red Sea hills that are channeled as surface runoff in the valleys and as subsurface groundwater flow in the underlying alluvial aquifers; a portion of this water gets trapped within underlying highly fractured basement reservoirs. Criteria used in the definition of these reservoirs are: (a) well-developed drainage network across a large area within the crystalline basement (~300 km2) (b) within the targeted watersheds, valleys are narrow and are floored by minimal alluvial deposits to minimize losses within the valley network to evaporation, (c) intersecting fracture patterns and/or shear zones that promote porosity, (d) presence of rock types that readily dissolve (e.g., carbonates) within the fractured system. The third is sporadic precipitation over extensive networks encompassing the crystalline basement and adjacent sedimentary that ultimately recharge the alluvial aquifers draining the main valleys within the network. Criteria used in the definition of these reservoirs are: (a) presence of extensive drainage network across the Eastern Desert, (b) isotopic compositions similar to those of present meteoric precipitation, and (c) thick alluvial deposits (e.g., proximity of River Nile Graben).
H21C-06 0830h
A Methodology for Solute Transport in Unsteady, Nonuniform Streamflow with Subsurface Interaction
An advection-dispersion-reaction model can generally be used to describe one-dimensional stream solute transport if the flow is steady and if the channel is smooth and uniform. When applied to unsteady, nonuniform streamflows, a model based on the Fickian analogy needs to be modified to account for the temporal and spatial variation of the cross-sectional area of the stream channel. In this paper, we explore this topic with a simple approximation method as well as an elaborate one, both of which are incorporated into a conjunctive stream-aquifer transport model and are applied to a hypothetical stream-aquifer setting. The simple method, while easier to implement, displays a persistent pattern of error in simulation results. The elaborate method, while accurate in computation, results in a more complicated model and requires extensive procedures to overcome the efficiency problem when simulating complex stream-aquifer interactions. However, by coupling the latter with the adaptive stepsize control for the Runge-Kutta method in a conjunctive stream-aquifer model, it not only greatly improves model efficiency but also results in more realistic modeling than previously reported.
H21C-07 0830h
The Analysis of Seasonally Varying Flow in a Crystalline Rock Watershed and Calibration of an Integrated Groundwater and Surface Water Model
Minimal scientific knowledge is available to describe the interaction of surface water and groundwater flow systems within watersheds in the crystalline rock setting of the Canadian Shield. The surface water and groundwater flow in a small watershed are being investigated using the fully coupled groundwater / surface water model HydroSphere. The watershed has an area of 6.92 square kilometers and contains multiple lakes, the largest with a surface area of 0.94 square kilometers and a maximum depth of 9.0 meters. The surface water drainage system includes wetlands, small streams, and beaver dams while the topographic relief has a range of 50 meters. A bathymetric survey has been completed and lake circulation patterns have been determined for the small lake within the watershed using RMA2. The overburden is thin with granite bedrock outcrops occurring at numerous locations throughout the basin. The conceptual model for the subsurface system includes the surface sediments while three categories are used to characterize the bedrock with these being moderately fractured rock, sparsely fractured rock, and fracture zones. These categories are consistent with those used in the characterization of the crystalline rock at the Atomic Energy of Canada Limited (AECL) Underground Research Laboratory near Lac du Bonnet, Manitoba. ArcView GIS was used to facilitate data synthesis, analysis, and visualization. Data layers include orthophotos and a digital Ontario Base Map. Hydrologic data for the basin includes 2 rain gauges and a continuous water level recorder for the watershed's largest lake. Additional climate data were obtained from nearby meteorological stations. The HydroSphere model is based on the FRAC3DVS three-dimensional groundwater model and the MODHMS surface water simulator. HydroSphere was used to investigate and calibrate the parameters for groundwater and surface water flow in the basin for a spring to fall time period. Model output is compared to the observed continuous hydrologic record. The work provides a better understanding of model and physical parameters for integrated surface water / groundwater systems of watersheds in Canadian Shield settings. This research will contribute to the ongoing development of HydroSphere such that it can be applied to the analysis of a wide range of watersheds.
H21C-08 0830h
Continuous Daily Simulation of Chloride Flux for a 12 km2 Forested Catchment in the Central Appalachians
Quantifying the processes by which water and conservative tracers are transformed from rainfall into streamflow is a key prerequisite to understanding biogeochemical cycling, pollutant movement, and aquatic ecosystem functioning in headwater catchments. An ideal process-based model would represent the key features of the observed watershed-scale response in the simplest terms possible, while remaining faithful to hydrologist's general understanding of runoff generation mechanisms. We present a conceptual model for daily water and chloride flux for Paine Run, a 12.4 km2 forested watershed in the Virginia Blue Ridge. The flow submodel is based on a decomposition of hydrologic response into "fast" and "slow" runoff components, and is parameterized independently of the solute data. Chloride flux is represented in terms of a cascading set of two travel-time distributions; one representing sporadic movement and evapoconcentration in the shallow subsurface, and the other representing continuous movement along deeper "baseflow" pathways. The model is calibrated against a seven-year series of weekly stream samples, as well as episodic data for selected events. We evaluate the model in terms of its simulation performance, its calibrated parameter values, and its credibility, addressing topics such as old water fraction in stormflow, overall residence time distribution, and frequency-domain response characteristics.