H13M-01 INVITED
An Incremental and Interactive Process for Watershed Characterization and Modeling: A Case Study in Southwestern North America
Spatiotemporal variability of catchment-scale hydrologic processes in southwestern North America are poorly understood due to strong seasonality in hydroclimatic forcing and complex variations in land surface conditions. Mechanistic understanding of these semiarid catchments may be achieved through integration of field observations, remote sensing and numerical modeling of distributed hydrologic processes. In this study, we present an approach designed to reveal the spatiotemporal structure and response of a mountainous basin in northern Sonora, Mexico. Our approach is based on an incremental and interactive process (IIP) of observations and simulations to yield insight into: (1) the topographic control on hydrological states and fluxes; (2) plant-water interactions and their seasonal dynamics; and (3) spatial and temporal variations in the rainfall-runoff behavior. We discuss how knowledge obtained during early efforts in the ungauged basin has been incorporated into recent field sampling, characterization and modeling activities. The IIP framework has allowed a detailed evaluation of basin characteristics, instrument records and remote-sensing observations necessary to reduce modeling uncertainty. We illustrate the approach using a high resolution network of precipitation gauges installed in response to preliminary observations and modeling efforts. Advances in process understanding and distributed predictions made through this framework will allow regionalization efforts in other ungauged basins in southwestern North America. http://www.ees.nmt.edu/vivoni/sonora/www
H13M-02
Coupling Strontium Isotope and Trace Metal Geochemistry With a Watershed Flow Path Model in the Lamprey River Watershed, New Hampshire
One significant challenge in watershed characterization is accounting for the groundwater flow system. In temperate climates, the hydrologic cycle is dominated by surface processes such as precipitation, evapotranspiration, and shallow groundwater flow. While, volumetrically, groundwater flow is less significant, it remains critically important for maintaining instream flow and closing the water budget within the watershed. The objective of the current study is to assess the adequacy of a simple groundwater flow conceptual model to describe groundwater flow in a mesoscale (50 km2), fractured bedrock, watershed. We hypothesize that groundwater flowpaths and seasonal flowrates at the scale of kilometers can be adequately described by a highly-resolved (30 meter grid), topographically dominated, homogeneous isotropic system. To test this hypothesis, we utilize the distinct and traceable chemical and isotopic signatures of the host bedrock. The combined use of trace elements and radiogenic isotopes (e.g., 87Sr/86Sr) provides great promise for the delineation of groundwater flow paths at the watershed scale, especially when bedrock hosts have distinctive compositions. The Lamprey River, located in Southern New Hampshire, is an especially promising watershed to carry out these types of analyses because groundwater within the watershed is stored primarily in bedrock aquifers with two geochemically contrasting bedrock hosts lying adjacent to each other. The bedrock aquifers include the White Mountain Magmatic Series of Mt. Pawtuckaway, a local topographic feature, and the Massabesic Gneiss Complex. Bedrock measurements, as well as rock leachate solutions, show distinct signatures. Surface water samples collected from ponds underlain by the differing hosts also reflect the bedrock signals with 87Sr/86Sr ranging from 0.70686 in the White Mountain Magmatic Series to 0.71443 in the Massabesic Gneiss Complex. Surface water samples collected along the Lamprey River during Summer 2006 show a decrease in 87Sr/86Sr in the downstream direction as the water generally flows from the magmatic series to the gneiss complex, indicating that low 87Sr/86Sr groundwater is infiltrating the river. These results also indicate groundwater inputs from extrabasinal flow. Further refinements of the model, coupled with forthcoming trace element analyses, will provide a stronger means to quantify these contributions along the gradient of the river.
H13M-03
Generality of Fractal 1/f Scaling in Catchment Tracer Time Series: Implications for Catchment Travel Time Distributions
The mean travel time - the time that it takes a parcel of rainwater to reach the stream - is a basic parameter used to characterize catchments. More generally, a catchment is characterized by its travel-time distribution, which is described not only by its mean but also its shape. The travel time distribution of water in a catchment is typically inferred from passive tracer time series (typically water isotopes or chloride concentrations) in rainfall and streamflow. The catchment mixes precipitation inputs (and thus passive tracers) falling at different points in time; as a result, tracer fluctuations in streamflow are usually strongly damped relative to precipitation. Mathematically, this mixing of waters of different ages is represented by the convolution of the travel time distribution and the precipitation inputs to generate the stream outputs. Previous analyses of both rainfall and streamflow tracer time series from several catchments in Wales have demonstrated that rainfall chemistry spectra resemble white noise, whereas these same catchments exhibit fractal 1/f scaling in stream tracer chemistry over three orders of magnitude. These observations imply that these catchments have an approximate power-law distribution of travel times, and thus they retain a long memory of past inputs. The observed fractal scaling places strong constraints on possible models of catchment behavior: commonly-used exponential or advection-dispersion travel time distribution models do not exhibit fractal scaling. Here we test the generality of the observed fractal scaling of streamflow chemistry, by analyzing long-term tracer time series from 17 other catchments in North America and Europe. Special care is taken to account for the effects of spectral aliasing. We demonstrate that 1/f fractal scaling of stream chemistry is a common feature of these catchments and discuss the implications of this observation to catchment-scale hydrologic modeling. We then present the best-fit travel time distributions for each site and explain differences among the sites in light of available hydrometric information.
H13M-04
Near-Surface Site Characterization Using a Combination of Active and Passive Seismic Arrays
Seismic surveys with an active source are commonly used to characterize the subsurface. Increasingly, passive seismic surveys utilizing ambient seismic frequencies (microtremors) are being used to support geotechnical and hazards engineering studies. In this study, we use a combination of active and passive seismic methods to characterize a watershed site at Haddam Meadows State Park, Haddam, Connecticut. At Haddam Meadows, we employed a number of seismic arrays using both active and passive approaches to estimate the depth to rock and the seismic velocity structure of the unconsolidated sediments. The active seismic surveys included seismic refraction and multi-channel analysis of surface waves (MASW) using an accelerated weight-drop seismic source. The passive seismic surveys consisted of MASW techniques using both linear and circular geophone arrays, and a survey using a 3-component seismometer. The active seismic data were processed using conventional algorithms; the passive seismic data were processed using both the spatial autocorrelation method (SPAC) and the horizontal to vertical spectral ratio (H/V) method. The interpretations of subsurface structure from the active and passive surveys are generally in good agreement and compare favorably with ground truth information provided by adjacent boreholes. Our results suggest that a combination of active and passive seismic methods can be used to rapidly characterize the subsurface at the watershed scale.
H13M-05
An Adaptive Multi-Scale Watershed Characterization Approach Utilizing Geoinformatics and Self-Organizing Maps
Environmental management and research within heterogeneous watersheds provides challenges for consistent evaluation and understanding of system functions. Assessing, mitigating, and managing diverse systems can be difficult due to varying natural characteristics, large geographic areas, domestic and international political boundaries, and varying degrees of spatial, temporal, and empirical data availability and quality. These characteristics often allow only specific geographic areas and research/monitoring topics to be realized. Through the development of data relationships and patterning, existing geographically specific studies and data can be used to infer responses of other areas which have limited available data, but exhibit similar landscape and watershed characteristics. The discussed approach aims to identify patterns from various data sources at a variety of spatial and temporal scales, including terrain morphometry, hydrology, vegetation, land use, soils, and climate and apply this data to active functions in the system such as hydrograph response. Automatic data collection methods have dramatically increased with advances in technology over the past two decades. Despite these advancements, it still remains difficult and expensive to monitor and understand all aspects of a system. This method looks to utilize available and known information at a various watershed scales and apply these observed values to other basins with less resolute or available data. The use of advanced geospatial analysis and Artificial Neural Network (ANN) processes, particularly Self-Organizing Maps (SOMs), is proposed as a method to discover landscape and watershed function patterns and similarities between areas in a watershed that are not only spatially disjointed, but dissimilar in their available data. An adaptive and evolutionary capability is presented, in which varying types of data can be fused to evaluate different management needs such as water quality, aquatic habitat, groundwater recharge, land use, and what-if scenarios.
H13M-06
Multi-Scale Entropy Analysis of Mississippi River Flow
Multiscale Entropy (MSE) analysis was applied to the long-term (131 years) daily flow rates (Q) of the Mississippi River (MR) to investigate possible change in the complexity of the MR system due to human activities since 1940s. Unlike traditional entropy-based method that calculates entropy at only one single scale, the MSE analysis provided entropies over multiple time scales and thus accounts for multi-scale structures embedded in time series. It is found that the sample entropy (SE) for Q of the MR and its two components, overland flow (OF) and base flow (BF), generally increase as time scale increases. More importantly, it is found that there have been entropy decreases in Q, OF, and BF over large time scales. In other words, the MR may have been losing its complexity since 1940s. A possible explanation for the loss in the complexity of the MR system is that the major changes in land use and land cover and soil conservation practices in the MR basin since 1940s -- these changes alter the natural system which existed before human intervention and thus reduce its complexity.
H13M-07
Drainage Density: A Framework for Predicting Peak and Low Flows in Ungaged Catchments
Drainage density, a measure of the linear concentration of streams on the landscape, has been recognized as an important geomorphic descriptor for decades, but has received comparatively little attention as a predictor of hydrologic behavior. Here we present a conceptual and analytical framework linking the underlying geologic structure of a landscape through drainage density to peak flow and drought response. In this framework, drainage density provides a first-order measure of the ratio of surface to sub-surface flow at the landscape scale, hence storage and response times during floods and droughts. We argue that this landscape-level metric can constrain estimates of peak flows and drought response in ungaged basins. We present data from the U.S. Pacific Northwest demonstrating the strong coupling between geology, climate, drainage density, and hydrologic response. We then show that unit peak flows increase with drainage density and minimum 7-day flows decrease with drainage density, although other factors also contribute to these relationships. The strength of the relationship between drainage density and flow increases as the return period between events increases. Thus, this relationship could be used to predict the magnitude of long return period events if one event of comparable return period is known in the region.