National threshold runoff estimation utilizing GIS in support of operational flash flood warning systems

Citation
Tm. Carpenter et al., National threshold runoff estimation utilizing GIS in support of operational flash flood warning systems, J HYDROL, 224(1-2), 1999, pp. 21-44
Citations number
31
Categorie Soggetti
Environment/Ecology,"Civil Engineering
Journal title
JOURNAL OF HYDROLOGY
ISSN journal
00221694 → ACNP
Volume
224
Issue
1-2
Year of publication
1999
Pages
21 - 44
Database
ISI
SICI code
0022-1694(19991012)224:1-2<21:NTREUG>2.0.ZU;2-J
Abstract
Threshold runoff is the amount of excess rainfall accumulated during a give n time period over a basin that is just enough to cause flooding at the out let of the draining stream. Threshold runoff estimates are indicators of ma ximal sustainable surface runoff for a given catchment, and are thus an ess ential component of flash flood warning systems. Used in conjunction with s oil moisture accounting models and areal rainfall data, they form the basis of the US National Weather Service (NWS) flash flood watch/warning program . As part of their modernization and enhancement effort, the NWS determined that improved flash flood guidance and thus improved threshold runoff esti mation is needed across the United States, with spatial resolution commensu rate to that afforded by the WSR-88D (NEXRAD) radars. In this work, Geograp hic Information Systems (GIS) and digital terrain elevation databases have been used to develop a national system for determining threshold runoff. Es timates of threshold runoff an presented for several locations in the Unite d States, including large portions of the states of Iowa, Oklahoma, and Cal ifornia, and using several options in computing threshold runoff. Analysis of the results indicates the importance of channel geometry in flash flood applications. Larger threshold runoff estimates were computed in Oklahoma ( average value of 34 mm) than in Iowa (14 mm) or California (9.5 mm). Compar isons of the threshold runoff estimates produced by the GIS procedure with those based on manually computed unit hydrographs for the selected catchmen ts are shown as a preliminary measure of the accuracy of the procedure. Dif ferences of up to about 15 mm for hourly rainfall durations were obtained f or basins larger than 50 km(2). (C) 1999 Elsevier Science B.V. All rights r eserved.