Effects of urbanization on the hydrologic regimes and geomorphic stability of small streams in southern California
Hawley, Robert Jeffrey
MetadataShow full item record
In southern California streams, altered hydrologic and sediment regimes associated with urbanization (hydromodification) have induced significant morphologic responses such as incision, widening, and planform shifts from single-thread to braided with far-reaching effects to adjacent land and throughout drainage networks. The overarching objective of this dissertation is to improve process-based understanding of these changes such that the risk of future degradation may be mitigated through improved management. Three chapters follow from this fundamental flow of logic: changes in land cover beget changes in flow regimes, leading to increased erosive energy and sediment-transport potential, which, dependent on the relative resistance of the setting, can culminate into substantial changes in channel form. The purpose of Chapter 1 was to understand the first step in this sequence: how urbanization affects the flow regime. Duration Density Functions (DDFs) were developed as histogram-style cumulative duration curves that represent the full range of geomorphically-significant flows as simple power functions. Using long-term data from 52 U. S. Geological Survey (USGS) gauges, empirical models were fit to both peak flows and DDF parameters (i.e., magnitude and shape) as multivariate functions of statistically-significant spatial variables including total impervious area. With little flow control at the subdivision scale to date, total impervious area became an effective hydrologic iv surrogate for urbanization, demonstrating an exponential effect on peak flows, particularly the 1-, 1.5-, and 2-yr events, and increased durations of all sediment-transporting flows. For example, watersheds with ~10% imperviousness typically exhibit a ~5-fold increase in Q1.5 and 2 to 3 times as many days of sediment-transporting flows relative to an undeveloped setting. The models developed in Chapter 1 directly informed the hydrologic components of the subsequent chapters, where impervious area was not found to be a significant predictor of geomorphic response when considered independent of setting or sediment transport. The focus of Chapter 2 was to understand the relative susceptibilities of regional channel types to hydromodification in the context of a 'Screening Tool' that is being developed to help managers assess risk across geomorphic settings. Specifically, Chapter 2 is focused on 1) the general framework of a pre-final version of the susceptibility screening tool, and 2) the development of risk-based analyses of geomorphic thresholds, a central component of key decision nodes in the screening tool. Geomorphic thresholds are real and of great concern in stream management, such that any susceptibility-assessment scheme should account for the proximity to such threshold-based responses. Logistic-regression analyses of braiding, incision, and bank stability directly and probabilistically assess proximity to geomorphic thresholds, and offer a framework for assessing risk that goes beyond expert judgment. Calibrated with local data that were collected in an extensive field campaign, the logistic models were highly significant (i.e., p < 0.005 to p < 0.0001) and correctly classified unstable states in ~90% of the cases using simple but powerful predictor variables that can be measured at the screening/reconnaissance level. A screening tool that incorporates objective probabilistic-based components is novel relative to previous and more subjective classification v schemes, such that regionally-diverse agencies and staff can quantitatively assess channel susceptibility with less variable results. With the objective of developing a process-based understanding of observed channel changes, Chapter 3 presents models that predict relative magnitudes, directions, and risks of channel responses as functions of cumulative sediment-transport capacity ratios (Lr) that contrast 25-yr DDF simulations of urbanized versus undeveloped conditions. Lr was a highly significant term in quantifying channel 'enlargement', whereas logistic regression of Lr in combination with d50 suggested that fine-grained systems (i.e., especially d50 ≤ 16 mm) have little capacity to absorb any increases in sediment-transport potential. A regional Channel Evolution Model (CEM) that includes departures from the original CEM of Schumm et al. (1984) is also presented along with a modified dimensionless stability diagram (sensu Watson et al. (1988)) that provides a conceptual framework for assessing relative departure from equilibrium/reference form for both lateral and vertical channel responses. The overarching conclusion of this dissertation is that urbanization markedly affects the flow regimes of streams in southern California and that the corresponding imbalances in sediment-transport capacity result in substantial geomorphic instabilities across most stream settings. Consequently, mitigation strategies should be tailored to specific stream types and incorporate process-based objectives such as maintaining sediment continuity via duration standards rather than traditional regulations focused exclusively on flow magnitude.