H34B-01 INVITED 16:00h
Integrated, Multi-Scale Modeling for Process, Parameter and Policy Research at the River Basin Scale
A new strategy for large-scale integrated hydrological modeling is proposed, which reduces the governing partial differential equations (PDE's) to ordinary differential equations (ODE's) using the semi-discrete finite volume method (FVM). The distributed model is formed on an unstructured grid constructed from domain decomposition using Delaunay Triangulation. The finite volume elements are prisms, projected from an unstructured grid (triangular irregular network, TIN) with constraints. This approach results in a local ODE system referred to as the model kernel. The global ODE system is formed by assembling the local system over the chosen hydrologic domain. The constraints are related to the river network and the watershed boundary, elevation contours, geology, vegetation, etc. The system is solved with an efficient ODE solver. The model is designed to capture "dynamics" in multiple processes with a minimum of elements, while maintaining the conservation of mass at all cells, as guaranteed by the finite volume formulation. The complexity or dimension of the model is determined by the choice of support scale of the geospatial data and the purpose of the model. The present paper focuses on how integrated, multi-scale models can utilize qualitative characteristics of climate, vegetation, landuse, and hydrogeologic conditions for developing dynamic water management scenarios in ungauged or poorly gauged regions. A "simulation scenario" is presented showing how natural climate variation might affect seasonal, interannual and decadal recharge; and how climate change interacts with the time scales of water development policy decisions. The scenario is meant to show how the evolution of human water practice and natural climate variability in a river basin, together affect the long-term water balance in a river basin. The scenario also illustrates the importance of coupled models to the dynamics of complex human-hydrologic landscapes.
H34B-02 16:15h
The Use of a Flow Forecasting System to Support Decision Making of Publicly-Owned Dams in Quebec
The Centre d'expertise hydrique du Quebec is responsible for the management of publicly-owned dams located mainly in southern Quebec. These dams are used for multiple objectives like flood control, water supply, tourist and recreational activities, and hydroelectricity. Such opposite objectives implies that dam operations in critical situations must be supported by scientifically sounded information in order to provide rational beyond these operations to the stakeholders. Scientifically sounded information can also increase the acceptability of operations that have some negative impacts by demonstrating that these operations help to avoid even more negative impacts. The presentation will provide a brief description of a flow forecasting system developed to support decision making. Some examples of its use in the management of the dams on lakes Kenogami and Saint-Francois will be presented. Specifically, the examples will examine the winter low flows of 2002, which followed an exceptionally dry period, and there effect on water supply and hydroelectricity. In this case, modelling yielded information on future inflows and eased the communication with stakeholders in a situation where historical observed data were useless given the exceptional hydrological conditions.
H34B-03 16:30h
Integrated Modeling of Water Policy Futures in the Imperial-Mexicali Valleys
Divided by an international border, the Imperial-Mexicali Valleys (IMVs) are linked by shared history, natural resources, culture and economy. This region is experiencing changes driven by policy makers both within and outside the IMVs. The largest external decision, the Colorado River Quantification Settlement Agreement (QSA) of 2003, opens the door to a laboratory for studying the consequences of a massive transfer of agricultural water to municipal users. Two irrigation districts, two urban water agencies and the State of California have agreed to a 75 year of more than 30 million acre-feet of Colorado River water from agricultural to urban use. Although Imperial Valley farmers will be compensated for water conservation and land fallowing, the economic, environmental and social consequences are unclear. Farmers who fallow will likely cause a greater impact on local businesses and government than those choosing on-field water conservation. Reduced agricultural water use causes reduced flow of irrigation runoff, at higher salinity than before, to the Salton Sea that, in turn, impacts the population dynamics of Ichthyan and Avian species at the Salton Sea. Municipal wastewater discharged into the New River by Mexicali, Mexico is also an important source of inflow to the Salton Sea that will be reduce by plans to reclaim the wastewater for various uses, including cooling water for two new power plants in the Mexicali. A restoration program is funded to produce a Sea with much reduced surface area. But this approach may, in turn, lead to increases in windblown dust from the dry lakebed that will contribute to an air basin already designated as a federal nonattainment area for particulate emissions. Additional water will be conserved by lining the All American and Coachella canals. But, eliminating seepage from the All American canal reduces groundwater recharge to aquifers used by Mexican farmers. A complex interplay of water-related issues must be accounted for if planners and residents of the IMVs are to make sound socioeconomic and environmental policy decisions. We use a spatially-explicit, stochastic, simulation compartment model (based on difference equations with a daily time step programmed with STELLAr software) to simulate the hydrologic system that underlies our broader modeling of the socioeconomic and environmental future of the IMVs. Alternative future scenarios are defined and used to explore the hydrologic and environmental implications of variations in future Colorado River flows, and various water-related policy decisions. The results of a suite of simulations, made assuming that the Sea is not impounded in small sub-Seas, suggest that the salinity of the Salton Sea is most likely to continue increasing. If this is the case, and if restoration of the Salton Sea continues to be a high priority, then more aggressive water conservation methods or a much smaller Salton Sea will be required. More aggressive conservation will lead to greater socioeconomic concerns, while a smaller Sea will increase concerns regarding windblown dust from the exposed lakebed.
http://www.borderplus20.sdsu.edu/
H34B-04 16:45h
Integration of Hydrologic, Sediment Yield, Sediment Delivery, Hydrodynamic and Sediment Transport Models in Large Great Lakes Watersheds
Computational tools to evaluate surface and subsurface water flow and sediment transport are commonly used by environmental and engineering practitioners. However, different parts of the hydrologic system (e.g. hillslope overland flow, groundwater, river channel flow) are often treated separately and at disparate spatial and temporal scales. Overland flow models typically have no explicit channel representation and vice-versa, making integrated assessments of water and sediment delivery from catchment to channel difficult. This is problematic when appraising the influence of land use change (urbanization, modification of riparian buffer strips, changes in tillage and forestry practices etc.) on catchment sediment movement and river flood hydrographs. A 'budgetary' approach was taken to defining the sources and sinks of water and sediment within large catchments in the Great Lakes area under a variety of land uses. These budgets were derived from existing datasets including digital elevation models, river flow and sediment load records, and dam sedimentation surveys. Numerical models of watershed hydrology and sediment delivery, 2-D river flow and sediment transport were constructed to develop a general understanding of the hydrologic and geomorphic behavior of these systems, and to predict the effects of changing land use and riparian buffer zone modification. Models were calibrated against river flow and sediment transport records, reservoir sedimentation surveys and harbor dredging records. The challenges and benefits of combining these diverse approaches and their implementation in best management practices are discussed.
H34B-05 17:00h
Integrated Hydrologic Models for Closing the Water Budget: Whitewater River Basin, Kansas
Groundwater and its recharge are unobserved and unmeasured components of the water cycle of a river basin. The objectives of this study were: 1) to evaluate the groundwater component of the water balance for the Whitewater River Basin using a 3-D saturated groundwater model, 2) to compare the groundwater model results with a fully integrated hydrologic model and, 3) to describe the spectral frequency response of the basin to long-term climate forcing. The basin is the Whitewater River, near Wichita, Kansas. The basin has an area of 1,100 square-kilometers, an elevation range of 380 - 470m (amsl), and an average annual precipitation of 858 millimeters. The near-surface geology is comprised of a weathered shale overlying limestone bedrock of Mississippian age. Streamflow and weather records are available from 1960. A steady-state saturated groundwater model (MODFLOW) was implemented assuming a simple two-layer conceptual model. A total of 422 wells with static water levels were available. Using a subset of the wells, a steady-state calibration of MODFLOW was performed by adjusting permeability between the two layers. Steady-state calibration resulted in an R2 of 0.89 for predicted and observed water levels. The remaining wells were used for validation, with an R2 of 0.92. The next step constructed the transient model using a fixed percentage of rainfall as groundwater recharge. For a single observation well the R2 was 0.89 (observed vs. predicted) for the transient calibration and 0.77 for the validation for a year simulation. The final step was to compare MODFLOW to an integrated model to provide a more complete representation of surface hydrologic dynamics. Here MODHMS (developed by HydroGeologic Inc, Herndon, VA) was used since it is MODFLOW-based with 3D variably-saturated groundwater flow, 2D overland flow, and 1D channel flow. MODHMS allows for canopy interception and evapotranspiration so total precipitation and potential evaporation were input to the model for a better estimate of recharge through complete energy and water balance. Singular spectrum analysis (SSA) was used to analyze the temporal response of precipitation, streamflow and groundwater levels from selected points in the model both for MODFLOW and MODHMS results. This paper demonstrates the use of integrated models for determination of groundwater recharge. Time series analysis proved to be a useful tool in identifying climate response within the watershed.
H34B-06 17:15h
The Physical and Chemical Effects of Mid-winter Water Removal from Alaskan North Slope Lakes
For many years, the oil industry and support services have withdrawn water from freshwater lakes to build ice roads and pads for increased access to remote sites with decreased maintenance costs. This technique is quite important to the oil industry in that it allows oil field development or maintenance while avoiding the environmental disturbance associated with construction of gravel roads and pads. Construction on ice-roads and pads begins in December or January when the tundra mat is adequately frozen to support construction traffic and continues through April (depending upon weather). Recently numerous questions have been raised regarding the potential environmental consequences of water withdrawal. Possible effects of pumping include impacts to the water balance, direct impacts to aquatic organisms (including fish and invertebrates), and impacts to the pond water chemistry (with subsequent effects on aquatic creatures). There may also be associated cumulative impacts as ponds are repeatedly pumped year after year. Questions have also been raised on pumping ponds and the consequent effects on neighboring (and potentially connected) unfrozen zones within frozen rivers that serve as over-winter fish habitat. This study includes continuous monitoring of water characteristics in selected lakes in the Kuparuk and Alpine Oilfields near Prudhoe Bay, Alaska. Lakes that may be affected by pumping activities were identified and monitored to permit evaluation of the factors that impact biological populations and chemical concentrations. The measurements collected during the first winter (2002-2003) did not provide ample evidence to make any conclusive statements regarding the potential effects of water removal from these tundra lakes. A relatively small volume of water was pumped from Kuparuk lakes K209 and K214 in December, January and February, but not enough to definitely state that greater water removal will not cause some measurable impact. Year two saw substantially more water removed from several lakes, one of which (K209) was pumped to its physical limits. Remote instrumentation are capable of detecting changes in volume (when normalized for snow loads) but have not displayed an unequivocal effect of pumping (i.e. measurable difference between pumped and non-pumped lakes). We have demonstrated the ability of monitoring lakes remotely utilizing electronic sensors and radio telemetry. We established and refined techniques to collect water samples that were representative of the lakes and generally have low variation among samples within the Kuparuk and Alpine fields. We have observed variation in basic water quality parameters such as alkalinity and conductivity at the Alpine lakes as compared to Kuparuk lakes. We documented the baseline chemistry concentrations and physical parameters in these tundra lakes. We have observed the variation in ice thickness within and among lakes and have related this variation to surface snow thickness and have validated models for estimating ice thickness. This background information will be valuable for use in characterizing changes that may be observed in the upcoming winter sampling series.
http://www.uaf.edu/water/projects/NorthSlope/lake_recharge/
H34B-07 17:30h
Integration of Complex Models Into a System Dynamics Based Basin Scale Planning Model for the Upper Rio Grande
As the finite, and often over-allocated water resources of the western United States are challenged by a myriad of growing demands, computer based simulations can be a powerful tool for evaluation of potential water use scenarios for hydrologic decision making and water policy analysis. To maximize their usefulness for policy analysis, such simulations should accurately represent the physical system as well as its interconnectedness to the socio-economic systems relevant to water planning without losing user accessibility or run speed. One solution to these constraints is system dynamics (SD) modeling at a relatively coarse spatial and temporal resolution. The challenge of this approach is in maintaining sufficient physical accuracy despite coarse resolution and SD's simple modeling framework. In this paper, the development of a reach-based monthly time-step system dynamics model of the upper Rio Grande River (from the headwaters in Colorado to Elephant Butte Reservoir in New Mexico) is discussed. Within this SD model, temporally and spatially coarse physical and operational relationships are abstracted from a variety of existing models with higher resolutions, including an operations model (Upper Rio Grande Water Operations Model (URGWOM)), a land surface rainfall-runoff model, an evapotranspiration model, and two groundwater models. Abstraction and calibration methods and implications of information loss associated with this scaling are considered.
H34B-08 17:45h
Water Quality and Quantity Modeling for Hydrologic and Policy Decision Making
This paper presents the results of a research project that elucidate the excesses of nitrogen and phosphorous using a spatial-temporal modeling approach. The project uses the approach of integrating biophysical and socio-economic knowledge to offer sound solution to multiple stakeholders within a watershed context. The aim is to promote rural development and solve environmental conflicts by focusing on the internalization of externalities derived from watershed management, triggering the transference of funding from urban to rural populations, making the city invest in environmental goods or services offered by rural environments. The integrated modeling is focused towards identifying causal relationships between land use and management on the one hand, and water quantity/quality and sedimentation downstream on the other. Estimation of the amount of contaminated sediments transported in the study area and its impact is also studied here. The soil runoff information within the study area is obtained considering the characteristics of erosion using a MUSLE model as a sub-model of SWAT model. Using regression analysis, mathematical relationships between rainfall and surface runoff and between land use or management practices and the measured nitrate and phosphate load are established. The methodology first integrates most of the key spatial information available for the site to facilitate envisioning different land use scenarios and their impacts upon water resources. Subsequently, selected alternatives scenarios regarding the identified externalities are analyzed using optimization models. Opportunities for and constraints to promoting co-operation among users are exposed with the aid of economic games in which more sustainable land use or management alternatives are suggested. Strategic alliances and collective action are promoted in order to implement those alternatives that are environmentally sound and economically feasible. Such options are supported by co-funding schemes designed with the private and public stakeholders having a role in the study area. The significance of this research is clearly depicted by the results of the different models applying here for the assessment of water quality parameters and for modeling upper catchments terrain surface change in the study area. Application of the methodology is presented for the Fuquene Lake Basin in Cundinamarca, Colombia. Additional research needs and limitations of the methodology are highlighted.