Multivariate Elasticity of Extreme Streamflow in the United States.
Gumennik, Irina.
2015
-
Abstract:
Increasingly, hydrological research reveals that streamflow may be changing in response
to changes in precipitation and land development. The concept of elasticity is used to
investigate the generalized sensitivity of streamflow to changes in multiple basin
characteristics. Multivariate regional regression models are developed for median annual
high and low flows for 18 hydrologic ... read moreunits in the United States. The resulting
coefficients of the multivariate regression models are shown to provide elasticity
estimates for basin characteristics that capture several climatic, morphological,
hydrological, and development influences on watersheds. Explanatory variables that
consistently appear in the regression models reveal patterns that demonstrate the
importance of employing a multivariate sensitivity approach. Regression estimates of
precipitation elasticity based on single variable models are consistently biased when
compared to more accurate multivariate approaches. Overall, precipitation and
groundwater flow exhibit the strongest influence on streamflow with elasticity values
that are an order of magnitude larger than for other explanatory variables. An increase
in land development is associated with an increase in both high and low flows, but at an
order of magnitude smaller compared to precipitation and
groundwater.
Thesis (M.S.)--Tufts University, 2015.
Submitted to the Dept. of Civil Engineering.
Advisor: Richard Vogel.
Committee: Annalise Blum, and Jeffrey McCollum.
Keyword: Hydrologic sciences.read less - ID:
- fn107965r
- Component ID:
- tufts:21425
- To Cite:
- TARC Citation Guide EndNote