Transcript DES601-Module06

Basic Hydrology &
Hydraulics: DES 601
Module 6
Regional Analysis
Regional Regression Analysis
• Regional analysis refers to the construction of
regression equations for a region by analyzing
historical discharges on many streams within that
region.
• These equations are statistical models of discharge
based on certain explanatory variables, typically
AREA
SLOPE
SHAPE INDICES
CLIMATE INDICES
Module 6
Watershed Characteristics
• What characteristics influence runoff?
• Where you are
• How large an area
• Gradient
Module 6
Watershed Characteristics
• What characteristics influence runoff?
• Width, shape
• Elevation: minimum, maximum + slope
• Roughness: Channels, overbanks
• Geology and soils
• Climate
• Vegetation
• Land use, including urbanization and
imperviousness
• Controls: Dams, gates, diversions, channel
rectification
Module 6
Watershed Delineating
• Topographic maps
• Hands-on methods
• Marking directly on
map
• Tracing using light
table
• Computerized methods
• DEMs
• GIS software
• Semi-automated
delineation
• Fully automated
delineation
• Automated
measurements should
always be verified
Module 6
Information Sources
•
•
•
•
•
•
USGS quadrangle maps
Aerial photos
Satellite imagery
NRCS soil surveys
Field surveys
Previous investigations
Module 6
NRCS county soils surveys
Module 6
Soil descriptions
WEB SOIL SURVEY:
http://websoilsurvey.nrcs.
usda.gov/app/HomePage
.htm
http://soildata.tamu.edu/
http://soils.usda.gov/soils/techn
ical/classification/scfile/index.
html
Module 6
Learn More
Videos that show watershed delineation, and
measuring some of the watershed
characteristics.
http://ww.rtfmps.com/hec-hms-movies/Module-2/Video-2/
Module 6
Regional Regression Analysis
• The equations are constructed by first fitting an
appropriate probability distribution to observations
at a gaged location (station flood frequency).
• Then the station flood frequency curves are used as
surrogate observations (at a specified AEP) to relate
discharge to select geomorphic variables
v
v
v
v
2
(QAEP  0  1 A  2 S0 L )  

• The “ betas ” are obtained by trying to make
“ epsilon ” small, the AREA, SLOPE, and other
watershed characteristics are the explainatory
variables.
Module 6
General forms of regional equations
• The resulting equations are then expressed in a
power-law form for actual application
1
2
QAEP,Estimate  0 (AREA) (SLOPE) (MAP)
3
• Equation suites for some regions have been
chronically problematic (produce inexplicable or
inconsistent results)
• Traditional regionalization is probably unwise in
Texas – too much variation in climate and
terrain, too little data.
Module 6
Regression Equations in Texas
• 1977 Massey and Schroeder (1977) USGS Water
Resources Investigations 77-110 Open File Report
• 6 Regions and some undefined areas
• Region 6 (trans-pecos Texas) used AREA and
MEAN ANNUAL PRECIPITATION as explanatory
variables
• Region 3 (west-central Texas) used area as the
only explanatory variable
• Analysis done using the “Bulletin 17B” process,
which has since shown to perform erratically in
much of Texas
• Variables used are AREA, SLOPE, and SHAPE, but
their use is inconsistent within a suite of
equations in some cases.
• Depreciated Method
Module 6
Regression Equations in Texas
• 1996 Asquith and Slade in
• 11 Regions
• 16 sets of equations (some regions are
broken at 32 square miles) 96 equations
total
• Number of stations used for developing
equations varies from 27 to 66
• SEE varies from 28% to 160%, with a mean
of 66%.
• Depreciated Method
Module 6
Regression Equations in Texas
• 2009 Asquith and Roussel
• Documented in HDM
• 7 Equations for different AEP
• Mean annual precipitation to account for
climatic variability
• OmegaM used to account for location
(mapped value)
• Area and Slope
• Current (2011) Suggested Method
Module 6
MAP
Module 6
OmegaEM
Module 6
Regression Equations in Texas
Module 6
Summary
• Regional analysis uses watershed characteristics
obtained from maps, reports, etc. to estimate
discharge at a location.
• Watershed characteristics are obtained from map
and other sources – required a way to measure
areas, lengths, etc.
• Regression equations relate these characteristics in
a region to a discharge (and an AEP value).
• In Texas location is conveyed by the OmegaM
parameter and climatic effect through the mean
annual precipitation.
Module 6