@InProceedings{Alemu_etal2011,
author="Alemu, Henok
and Velpuri, N.M.
and Senay, Gabriel B.
and Angerer, J.",
editor="in",
title="A multi-index approach to delineate surface water bodies in the pastoral regions of Mali using ASTER imagery [abs.]",
booktitle="Fall Meeting",
year="2011",
edition="5-9 December 2011",
publisher="American Geophysical Union",
address="Washington, D.C.",
volume="Fall Meeting Abstracts",
optkeywords="Africa",
optkeywords="area",
optkeywords="ASTER",
optkeywords="band",
optkeywords="bands",
optkeywords="conference abstracts",
optkeywords="data",
optkeywords="dynamics",
optkeywords="early warning",
optkeywords="East Africa",
optkeywords="hydrologic modeling",
optkeywords="identification",
optkeywords="image",
optkeywords="imagery",
optkeywords="livestock",
optkeywords="Mali",
optkeywords="map",
optkeywords="migration",
optkeywords="modeling",
optkeywords="monitoring",
optkeywords="NDWI",
optkeywords="Normalized Difference Water Index",
optkeywords="research",
optkeywords="resolution",
optkeywords="resources",
optkeywords="satellite",
optkeywords="satellite data",
optkeywords="spatial",
optkeywords="spatial distribution",
optkeywords="spatial resolution",
optkeywords="spectral band",
optkeywords="surface water",
optkeywords="temporal",
optkeywords="trend",
optkeywords="warning system",
optkeywords="water",
optkeywords="water bodies",
optkeywords="water index",
optkeywords="water level",
optkeywords="water resource",
abstract="Information on the location and availability of water resources is a day-to-day challenge for pastoralists in the Sahelian region of Mali. They move seasonally along their migration corridors in search for water and forage. Satellite data can be used to map the spatial and temporal dynamics of these water resources. In this work, ASTER imagery is selected for its high (15 m) spatial resolution and suitable spectral bands for water body identification. Our research indicates that as most of the waterholes of interest in the study area are very shallow and heavily sediment-laden, using only one of those commonly used water identification indices such as the Simple Band Ratio (SBR), or the Normalized Difference Water Index (NDWI) alone does not help in effectively characterizing all the surface water bodies in the region. As a result, we used four different spectral indices to identify surface water features: (i) Simple Band Ratio (SBR), (ii) Normalized Difference Water Index (NDWI), (iii) Modified Normalized Difference Water Index (MNDWI), and (iv) the Mean Absolute Deviation (MAD) to identify and delineate surface water bodies using 91 ASTER images. Initial results indicate that the SBR method identified 17 waterholes while the NDWI 18, the MNDWI 36, and the MAD method identified 28 waterholes. However, by combining the results from the four aforementioned spectral indices following a multi-index approach, 89 waterholes that were previously unidentified by a single approach alone were identified. Furthermore, our analysis indicates that the SBR and the NDWI methods identify relatively clearer waterholes better (29\% of the waterholes), whereas MNDWI and MAD proved to be good indices for identifying sediment-laden waterholes. Identifying the location and spatial distribution of surface water bodies is the first step towards monitoring their seasonal dynamics using a hydrologic modeling system, similar to an existing setup for east Africa (http://watermon.tamu.edu/). Seasonal trends in relative surface water levels are one of the most important inputs in the livestock early warning system (LEWS) along with forage and livestock market prices.",
optnote="exported from refbase (http://eros.usgs.gov/refbase/show.php?record=23329), last updated on Fri, 21 Sep 2012 11:59:54 -0500",
issn="http://www.agu.org/meetings/fm11/waisfm11adv.html",
opturl="http://www.agu.org/meetings/fm11/waisfm11adv.html",
language="FY 2012"
}