HR: 0830h
AN: B21E-0761 [PDF]
TI: Using remote sensing to estimate global impounded water in small artificial reservoirs
AU: * Blaisdell, B
EM: betsyblaisdell@hotmail.com
AF: Earth Sci., University of New Hampshire, Durham, NH 03824
AU: * Blaisdell, B
EM: betsyblaisdell@hotmail.com
AF: EOS, University of New Hapmshire, Durham, NH 03824
AU: Sahagian, D
EM: dork.sahagian@unh.edu
AF: Earth Sci., University of New Hampshire, Durham, NH 03824
AU: Sahagian, D
EM: dork.sahagian@unh.edu
AF: EOS, University of New Hapmshire, Durham, NH 03824
AU: Xiao, X
EM: xiangming.xiao@unh.edu
AF: EOS, University of New Hapmshire, Durham, NH 03824
AU: Braswell, R
EM: rob.braswell@unh.edu
AF: EOS, University of New Hapmshire, Durham, NH 03824
AB:
The impoundment of water has served to increase global water resources through time in support primarily of growing
agricultural needs. If future capabilities or policies retard the increase in water impoundment, then there will be an impact
on agricultural productivity that may not be accounted for in current projections. Thus, it would be desirable to determine
the total amount and rate of water impounded to date in order to better estimate the effects on usable water resources and
global sea level. Dams have been built for many years, but only the large, registered reservoirs are counted in international
tallies of impounded water resources. To date there is not even a rough estimate of the amount stored in the literally
millions of small reservoirs such as farm ponds and rice paddies that have resulted from small scale agricultural and other
land use. The volumes of these cumulatively may exceed that of the few large reservoirs. Further, there has been no estimate
at all of the amount of ground water impounded by any dams, large or small. While it is impractical to inventory millions of
small agricultural impoundments, it is possible to use remote sensing data to estimate the quantity of this impounded water
so that more accurate projections of water resource availability as well as 21st century sea level rise can be made.
One approach to this problem is based on hydrologic land use regimes globally. Hydrologic land use characterization would
involve determining the amount of water impounded for agriculture locally using Landsat data along with ground-based
information, and extending to the global scale using MODIS data. Comparison with climate data will make it possible to
estimate the area of irrigated land using mid-infrared and optical bands. Based on this information one can calculate the
global areal extent of agricultural land irrigated from impounded water resources. A second approach involves unmixing MODIS
data for sub-pixel analysis of small water bodies directly. The MODIS analysis can be "trained" by high-resolution analysis
of Landsat TM data, leading to global impounded reservoir area from MODIS without the need for hydrologic land use
characterization. In both approaches, water volume can be determined by impounded water area inferred from remote sensing,
and typical water depths obtained from local studies. For all reservoirs (big and small) impounded ground water can be
estimated using topographic databases, combined with information regarding typical porosities and dam height. Integrating
this information with impounded surface water, total increases in water resources can be determined.
DE: 1640 Remote sensing
DE: 1655 Water cycles (1836)
DE: 1803 Anthropogenic effects
SC: Biogeosciences [B]
MN: 2003 Fall Meeting