Developing Potential Scenarios of Changes in Hydroclimatic Variables for Analyzing the Impacts on a Regional Water Supply System
Abstract
Examining the impact of potential changes in hydroclimatic attributes (e.g., changes in precipitation, temperature, and streamflow) is critical to understand the system vulnerability and reliability in water supply systems under potential climate change. This study focuses on Tampa Bay Water, a regional water supply agency in the west coast of Florida, which considers a diverse portfolio of surface water to meet its continuing growing demand. Recently, wavelet based autoregressive model have been proposed to reproduce the time-frequency properties of the observed rainfall. Given the evidence of low frequency events such as ENSO in the southern United States, we propose an alternative wavelet-based hidden Markov model, a flexible model which considers multiple state-space conditions, to capture the time-varying low-frequency characteristics in observed streamflow. Our analysis show that the proposed hidden Markov model (HMM) simulation provides better results than an autoregressive model in terms of reproducing the low frequency variabilities of hydroclimatic variables by better preserving observed mean, variance, skewness, lag-1 auto-correlation and also flow duration curves. We also propose an innovative approach to bootstrap the simulations of wavelet-based HMM for developing synthetic streamflow generation scenarios that exhibit a shift in the marginal distribution of flows under potential climate change. Thus, the proposed synthetic generation scheme can be useful for examining hydroclimatic attributes under stationarity and under changing streamflow conditions suggested by climate model inter-comparison projects (CMIP5) scenarios.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2021
- Bibcode:
- 2021AGUFM.H25U1257C