HR: 10:20h
AN: H31G-01 INVITED     [PDF]
TI: Identifying Characteristic Responses of Hydrological and Geochemical Behavior at Hillslope and Catchment Scales
AU: * Newman, B D
EM: bnewman@lanl.gov
AF: Earth and Environmental Science Division, Los Alamos National Laboratory, MS J495, Los Alamos, NM 87545 United States
AU: Duffy, C J
EM: cxd11@psu.edu
AF: Dept. of Civil and nvironmental Engineering, Penn St. Univ, 212 Sackett Bldg., University Park, PA 16802 United States
AU: Hickmott, D D
EM: dhickmott@lanl.gov
AF: Earth and Environmental Science Division, Los Alamos National Laboratory, MS D462, Los Alamos, NM 87545 United States
AB: Time series of hillslope or catchment water content, subsurface flow, and solute chemistry typically show complex and highly variable behavior. Event-based approaches are often used to characterize input-output responses (e.g., tracking an isotopically unique precipitation event, or by fitting a seasonality curve to the data). However, such approaches typically quantify just one characteristic response when we know that hillslopes and catchments have a distribution of characteristic time scales. Thus, we seek an alternative approach, one where we can identify distributions of characteristic responses and quantify individual variance contributions. Besides improving our conceptual understanding of how hillslopes and catchments work, the goal is to develop low-dimensional "empirical" and/or physically based models that more fully represent the distribution of characteristic responses. To this end, singular spectrum time series analysis is used to identify characteristic times of hydrologic and geochemical variables at hillslope and catchment scales. For our semiarid study sites we find that soil water content, spring flow (representing catchment scale behavior), and geochemical responses can be characterized by a small suite of periodic or quasi-periodic characteristic "modes' that account for a major fraction of the total system variance. The remaining modes (i.e., those that only account for a small fraction of the variance) typically display "red-noise" type behavior suggesting a simple stochastic model. Our results show that singular spectrum analysis offers new insights on hydrologic and geochemical response and should be a fruitful way of developing improved hillslope- and catchment-scale models.
DE: 1831 Groundwater quality
DE: 1832 Groundwater transport
DE: 1860 Runoff and streamflow
DE: 1866 Soil moisture
SC: Hydrology [H]
MN: 2003 Fall Meeting