You are here:
Uncertainty Management in Urban Water Engineering Adaptation to Climate Change - abstract
Citation:
Yang, J. Uncertainty Management in Urban Water Engineering Adaptation to Climate Change - abstract. Presented at EWRI Annual Meeting, Cincinnati, OH, May 20 - 22, 2013.
Impact/Purpose:
Abstract submitted for EWRI annual conference in May 2013.
Description:
Current water resource planning and engineering assume a stationary climate, in which the observed historical water flow rate and water quality variations are often used to define the technical basis. When the non-stationarity is considered, however, climate change projection contains a large uncertainty, particularly in precipitation. This model performance is region-specific and will unlikely improve in the near future to the degree adequate for deterministic engineering and planning of urban water systems. Both the non-stationary assumption and the climate model uncertainty make it difficult to make adaptation decisions. This paper describes the nature of the uncertainty, using two hydroclimatic regions – the Ohio River Basin and Floridian Gulf Coast, to examine probabilistic, rather than deterministic, approach in evaluating hydrological parameters for the next 20-30 years. Specifically, historical precipitation records of long duration (>80 years) are analyzed using wavelet spectrum analysis to show long-term and periodic changes in the frequency and magnitude of low-probability (i.e., 10%) precipitation events. Robust statistics suggests pertinent changes in precipitation return intervals since 1980s, and the type of change is location-specific. The GEV rainfall distribution is not adequately captured in Bulletin-18 or in the SCS method of return storm period. Furthermore, the uncertainty or method error in design storm is propagated in the values of determined future flow rate and water quality parameters, and thus in uncertainties of engineering design parameters. The non-linear uncertainty propagation and associated effects on water systems are illustrated in two examples: HSPF simulation of stream flow and water quality in a small Ohio urban watershed, and secondly, process modeling of water treatment for source water quality changes. Based on these results, a probability-based approach is outlined to define required system capacity reserve (CR) and probabilistic cost at a desired degree of system resilience.
