We propose to conduct a pilot project to develop and test the physical basis of a statistical scaling theory for peak flows for individual rainfall-runoff events in natural river basin. The theory rests on a peak-flow-scaling hypothesis that can be stated as: Statistical distributions of peak-flows for complete H-S streams in rainfall-runoff events exhibit scale invariance as a fundamental organizing principle, which can be predicted from knowledge of space-time rainfall; infiltration and runoff generation; network topology; geometry; hydraulic geometry; channel losses due to transpiration from riparian vegetation and/or infiltration; and channel gains due to recharge. Scale invariance or scaling is an emergent property of a complex nonlinear system, which is not built into the equations representing physical processes that produce peak flows. Our objectives for a pilot project combining both theoretical analyses and a field program will be carried out on the 150 sq. km. Walnut Gulch basin, AZ, the 21 sq. km Goodwin Creek basin, MS, and the 1100 sq. km. Whitewater basin, KS. Whitewater is a carefully selected basin for investigating heavy-rainfall induced floods resulting dominantly from overland flow. It is a tributary of the Walnut River, which was a site for Atmospheric Boundary Layer Experiments (ABLE) during 1995-2004. Atmospheric Radiation Measurement (ARM)/DOE Climate Research facility (ACRF) support of our project provides use of ABLE instruments that we need, and help with field deployment and maintenance of instruments in Whitewater. Our specific objectives are: (1) Investigate physical basis of statistical scaling in peak-flow using existing data from the two ARS basins, (2) Test preliminary validity of mean peak flow scaling and physically postulated change in it by installing eleven new stream flow gauges in the Whitewater basin that can accurately measure stream flow hydrographs on the basis of a new technology developed by us, (3) Test a new Hortonian fetch hypothesis for transpiration losses, which states that the statistical distributions of forested riparian fetches over which trees extract water from a channel network vary predictably with the H-S order and yield Horton ratios for fetch area and transpired volume. Installation of a twelfth stream flow gauge to accurately measure flow losses in a channel link due to transpiration, (4) Estimate good quality space-time rainfall intensity data using measurements from existing WSR-88D NEXRAD radars and 14 new rain gage sites, (5) Conduct statistical scaling analyses on rainfall data obtained under objective-4 and develop a predictive understanding of rainfall scaling statistics from physical processes, (6) Deploy and test a new network for real-time data collection, transmission and storage from our sensors. Our cooperative interdisciplinary and multidisciplinary project includes senior faculty, post docs, and graduate students from the University of Colorado, Boulder, the University of Iowa, and the University of New Mexico, and scientists from the U.S. Geological Survey. Intellectual Merit: A physical basis of the statistical scaling theory for peak flows for rainfall-runoff events in natural river basins provides a key step towards solving the long-standing problem of flood prediction from ungauged basins (PUB). It also provides a rigorous scientific framework for detecting and predicting how anthropogenic changes in landscape and climate variability modify biophysical processes generating floods, which can be used for predicting flood statistics.
Broader Impacts: Our pilot project is a major first step towards developing a new unified network-based theory that couples hydrology, boundarylayer- and hydro- meteorology, landscape and stream ecology, biogeochemistry, climate variability and anthropogenic change in river basins.
