Improving Streamflow Reconstructions Using Oceanic-Atmospheric Climate Variability

Abstract

Recurring hydrologic drought in Southwestern U.S. has raised questions about the reliability of future water supplies. Streamflow reconstruction using tree-ring chronologies (TRC) is used traditionally to expand the length of past streamflow records, which are helpful in long-term planning of water resource especially by giving information about duration and severity of past regional drought events. This study focuses on improving reconstruction methodology using TRC, by incorporating Sea Surface Temperature (SST) and climate indices of Pacific Decadal Oscillation and Southern Oscillation Index, as predictors, in a stepwise linear regression model. The approach is tested on six unimpaired streamflow gages in the Rio Grande Basin, located in the western U.S. Reconstructions are performed from 1856-2002. Results showed SST as the more influential predictor compared to using TRC only and climate indices. Future work will involve applying the modeling approach in other Western U.S. river basins. Introduction Streamflow is a vital source of water supply. Climate variability is expected to affect streamflow (Dawadi and Ahmad, 2012, Forsee and Ahmad, 2011), and thus may lead to conditions reflected in the form of droughts (Stephen et al., 2010; Puri et al., 2011a) or floods (Mosquera-Machado and Ahmad, 2007; Puri et al., 2011b; Ahmad et al., 2010; Ahmad and Simonovic, 2000; 2001). Climate variability coupled with increasing population calls for sustainable water management for municipal, agricultural, environmental, and public health needs (Wu et al., 2013; Vedwan et al., 2008; Kalra and Ahmad, 2012; Dawadi and Ahmad, 2013; Qaiser et al., 2011; Shrestha et al, 2011; Venkatesan et al., 2011a; Ghumman et al., 2013; Melesse et al., 2011; Impoinvil et al., 2007a; 2007b). Long-term streamflow records can help in understanding the range of flow variability and developing water management strategies (Foster, 2008; Mirchi et al. 2013). The available instrumental data is usually limited and thus might be insufficient to describe the long-term hydrologic variability of a region (Carrier et al., 2013). Streamflow reconstruction, by increasing the length of the data, is an important tool to understand past hydrology (Woodhouse and Lukas, 2006). 846 World Environmental and Water Resources Congress 2014: Water without Borders © ASCE 2014