Statistical Methods for Assessing Climatic and Anthropogenic Impacts on Streamflow, Storage Reservoir Yield, and Effectiveness of Water Conservation Programs.
Tsai, Yu-Shiou.
2012
-
Abstract: Statistical methods are applied for evaluating the impacts of
climatic and anthropogenic influences on streamflow, for determining the ability of storage
reservoirs for managing hydrologic variability and for assessing the effectiveness of water
conservation programs. The impacts of changes in climate, land use and water use on both
mean flows and droughts across the Eastern U.S. are... read moreestimated in terms of dimensionless
elasticity using three multivariate regression approaches. These three influences have
never been considered simultaneously in any previous research particularly with reference
to floods and droughts, and yet we find that such impact assessments for which only one
influence is considered are most likely to give biased estimates. Across the eastern U.S.
both mean and drought flows are found to be quite sensitive to changes in climate and land
use, with both mean and drought flows being most sensitive to precipitation except for the
New England Watershed (region 1), where the drought flows are most sensitive to
temperature. Across the eastern U.S., impacts of land use changes on both annual mean and
drought flows are quite significant. Particularly for region 1, where urbanization is
intense, the overall land use changes have greater impact than climate (precipitation) on
annual mean flow. The impact of water use on annual mean flow in region 1 is found
significant while for other regions impact of water use on both mean and drought flows are
insignificant. The impacts of precipitation (P) and potential evapotranspiration (PET) on
floods, droughts and mean flows across the continental U.S. are evaluated simultaneously
using two multivariate regression approaches. Remarkably no previous research has evaluated
climate impacts on mean, flood and drought flows systematically across a national scale
study area. Stochastic based regression approaches are found to provide better fits than
their deterministic counterparts. Flood flows are found to be more sensitive (1) to both P
and PET for basins with a drier climate, (2) to P for basins with less snow, smaller
drainage areas, smaller lakes/wetlands area, smaller soil permeability and steeper
terrains, and (3) to PET for basins with more snow and flatter terrains. Drought flows are
found to be more sensitive (1) to P for basins with a wetter climate, larger lakes/wetlands
areas for attenuating runoffs and lower baseflow, and (2) to PET for basins with a drier
climate and smaller soil permeability. This study further documents the importance of
spatial and temporal scales on estimation of the climate elasticity of streamflow. An
optimization model is combined with generalized storage-reliability-yield relationships to
determine the optimal capacities and locations of additional storage reservoirs needed for
the Zambezi River Basin. Since economic growth within the Zambezi River Basin thrives on
provision of high reservoir yield reliability which also results in a reduction of
hydrologic variability downstream to a reservoir, we advocate use of yield reliability for
quantifying the hydrologic variability associated with reservoir planning and operations.
In addition, the results of the optimization analysis suggest the locations within the
region where additional storage capacities are needed. Finally, the effectiveness of four
water conservation programs is evaluated using controlled experiments in four towns within
the Ipswich River Watershed in Massachusetts. This study introduces nonparametric
statistical methods, which have seldom been applied in similar previous research, for
evaluating the effectiveness of these conservation programs in reducing water use.
Installation of weather-sensitive irrigation controller switches (WSICS) in municipal ball
fields, free residential water use audits with water fixture retrofit kits, rebates for
low-water-demand toilets and washing machines, and soil amendments in ball fields are found
to reduce water use significantly. Installation of WSICS in residences is found to be more
effective in reducing water use of high water users. Rainwater harvesting is found to
provide substantial rainwater use, but these volumes are small relative to total domestic
water use and relative to the natural fluctuations in domestic water use.
Thesis (Ph.D.)--Tufts University, 2012.
Submitted to the Dept. of Civil Engineering.
Advisor: Richard Vogel.
Committee: Sara Cohen, Stephen Levine, and Jeffrey Zabel.
Keywords: Hydrologic sciences, and Water resources management.read less - ID:
- xk81jx992
- Component ID:
- tufts:21031
- To Cite:
- DCA Citation Guide EndNote
