Title: FFPI Dataset for AFG
1Flash Flood Potential Index
Greg Smith AHPS/ Flash Flood Services FFPI
Program Lead Senior Hydrologist Colorado Basin
River Forecast Center
FFPI Dataset for AFG
2Offices offering review, feedback,
recommendations, for the FFPI concept
Define direction/continuation of this
project Define an operational concept
3Fairbanks CWA FFPI
FFPI Datasets for Review Analyze and provide
feedback - compare known flash flood
areas Compare/log FFPI indices corresponding with
this seasons events Feedback, guidance, value
of this type of information
Reasoning Method Examples
4FFPI Concept
simple numerical index tied to runoff response
potential
Attempts to account for those land-surface
features affecting hydrologic response
5Flash Flood Factors
Meteorological Factors
Vertical Moisture Flux Mean Wind Speed Cell
Movement Atmospheric Cap Storm Cell
Propagation/Training Quasi-Stationary
Systems Lifting Mechanisms (Orographic
Enhancement) Moisture Source (warm-cool rain
process)
Rainfall Rates Duration Storm Movement
Non-Meteorological Factors
Basin Geometry
Slope Forest / Vegetation cover Soil
Type Land-Use Practices (Urbanization) Wildfire
Activity Basin Geometry (Hydraulic) Antecedent
Soil Moisture (Hydrologic)
Physiographic
Runoff Response Flash Flood Potential
6The Motivation (Influenced by Western Region
issues needs)
- FFMP emphasized need for improved finer
resolution FF information - FFG is coarse, non-existent, or rule-of-thumb
based - Basins with different physical features respond
the same - Local office request for additional FF
information and better guidance - Modernized FFG method issues
- Scale, model coverage, soil moisture dependency,
bankfull definitions
7Flash Flood Guidance Local and Rule-of-Thumb
procedures lack spatial resolution
Implication All basins have the same hydrologic
response
Flash Flood Precipitation Thresholds
Observed Precipitation
FFMP Flash Flood Basins
Pre-defined Climate Zones
Areas range from 15-130 KM2
Largest exceeds 13,000 KM2
Scale Issue Precipitation Thresholds and Radar
Observed Precipitation
8Coarse Resolution FFG
CNRFC FFG
CBRFC FFG
9Developing a Flash Flood Potential Index
How do you account for multiple characteristics
in a small river basin?
?
Vegetation type or density
Soil type
Wildfire
Land Use
Slopes
10FFPI Method
Obtain raster (gridded) datasets representing the
features of interest.
Forest Density
Elevation/Slope
11Methods
FFPI Method Re-classifying datasets - elevation
GIS world (raster)
Real world
Flash Flood Potential Index (FFPI) Values
Relative
Reclassification
User defined processes
Decreasing
Hydrologic Response
Increasing
Map world (DEM)
Math world (eqns)
12Data Re-classification
Re-Classify Data Equal Interval Classification
for 10 categories
Attributes for Forest Density Texture Layer
Attributes for Soil Texture Layer
Decreasing
Hydrologic Response
Increasing
Bedrock/Impervious
13Methods
FFPI Method Re-classifying datasets Land Use
NLCD Categories (Ranking) Commercial/Industrial/T
ransportation High Intensity Residential Bare
Rock Low Intensity Residential Quarries-Strip
Mines-Gravel Pits Shrubland Row
Crops Orchards/Vineyards Grassland Pasture-Hay Dec
iduous Forest Evergreen Forest Mixed Forest Woody
Wetlands Perennial ice - snow Open Water
10
Increasing
Hydrologic Response
Decreasing
1
NLCD - National Land Cover Data Classification
Code 1992
14Methods
FFPI Method
Using Geographic Information System (GIS)
Gather digital data for different layers Soil
type Forest density/cover Slope Land
use/urbanization
Resample and Geo-register Data Consistent
resolution Same projection/datum
Overlapping grid cells representing same location
on the surface of the Earth
15FFPI Method Creating a single FFPI layer
Hydrologic response
3
7
Slope
Greater response and potential
7
8
3
Forest Density
Lesser response and potential
3
6
4
Land Use
4
9
6
Soil
9
Flash Flood Potential Index
16FFPI Method
Scale Issues and characteristics Not all
layers can be viewed as equal to one another Not
all layers respond in a linear fashion
1
10
4
3
2
8
7
6
5
9
Layer 1 Layer 2 Layer 3 FFPI
7
8
9
10
1
2
3
4
5
6
1
2
3
4
5
6
7
8
9
10
Greater response and potential
1
10
Lesser response and potential
17AFG Slope FFPI
Data Source Derived from USGS 2 arc-second DEM
18AFG Land FFPI
Data Source NLCD 2001 Land Cover
19AFG Forest FFPI
Data Source AVHRR Derived Tree Cover Data Global
Land Cover Institute - UMD
20AFG Gridded FFPI
Combination of slope, soil type, land-use,
forest density FFPI layers.
21AFG Gridded FFPI Summed to a Basin Layer
Alaska HUC Basins
22Accumulated hydrologic response information is
displayed as one value
Land Use
Soil type
Vegetation type and density
1
FFPI
10
Increasing Potential
Decreasing Potential
Slopes
23Comparing FFPI Basins and Reality
Lower FFPI values (lighter) Near Moquith
Mountain gradually sloped, sandy basins
Higher FFPI values (darker) Basins in Capitol
Reef National Park
24Flash Flood Potential Index Concept
Simple index Accounts for those land-surface
features affecting hydrologic response Rank
drainage basins Higher index values ? greater
hydrologic response ? greater flash flood
threat Index flexibility Identify flash flood
prone areas Supplement existing methods
(classify basins) Use with areal or rules of
thumb ? better qualify basins response Incorporat
e into alternative methods for generating flash
flood guidance
25FFPI in the Operational Warning Environment
NWS - Warning Decision Training Branch - Advanced
Warning Operations Course
26(No Transcript)
27Operational Use WFO Salt Lake City
FFPI Display These basins on the North Fork
received the heaviest rainfall rates with total
rainfall amounts exceeding FFG by over ½ inch.
Radar reflectivity (above) and storm total precip
(below). FFG was exceeded but a FF Warning was
not issued.
Basins of the North Fork drainage that feed the
Escalante River. Basins are on the eastern slope
of the range but it is heavily forested.
28CNRFC FFG Comparison
FFPI based gridded FFG
NWSRFS Derived
29FFPI Potential Operational Applications
Account For Wildfire
30Overlapping Fires near Tucson, AZ 2002-2003
UTAH
NEVADA
ARIZONA
Bullock Fire Summer 2002 30,563 acres (128 km2)
MODIS - NASA Image
Aspen Fire June 2003 84,750 acres (343 km2)
31Wildfire effects
High Burn Severity All vegetation blackened,
deep soil heating killing roots/seeds, baking
of the soil surface.
Moderate Burn Severity Patchwork of green and
burnt areas. Intermediate between high and
low severity levels.
Low Burn Severity Most vegetation untouched by
fire. No significant effect on soil properties or
water repellency.
32How to apply fire burn severity information ?
High Burn Completely remove forest
density Maximize soil hydrologic response (max
FFPI) Moderate Burn Reduce forest density
50 Increase soil FFPI 50 Low or non burn
No change to hydrologic response
33Aspen Fire Burn Severity Perimeter and Pre-fire
FFPI Grids
Pre-Fire FFPI Grids
Aspen Fire Perimeter
Severe Burn
Moderate Burn
NWSRFS Segment Boundaries
34Fire Modification to FFPI Grids
After Aspen Fire
Before Aspen Fire
35Completed
FGZ Flagstaff GJT Grand Junction PSR
Phoenix RSA CNRFC SLC Salt Lake TWC
Tucson VEF Las Vegas ABQ Albuquerque RNK
Roanoke/Blacksburg SGF Springfield, MO AFG
Fairbanks SJU San Juan PR GJT Grand Junction CO
Status of FFPI Development
AHPS supported test sites
36FFPI 2008 Plan
Summer 2008 (Evaluation) Deliver FFPI to test
sites for review, comment, evaluation,
recommendation Log summer events compare with
FFPI indices Autumn 2008 (conops
recommendations) Provide a possible CONOPS
summary to AHPS FF Services Team Recommendations
from field offices decide the fate/direction of
this project Ongoing Work Enhance FFPI by
including a dynamic soil moisture
layer Establish an automated method to obtain
and modify FFPI based on wildfire data
Recommendations To Date Generate
statistical FFG relating flash flood event data
to FFPI indices Do FFPI indices show a relation
to precipitation return frequencies ? Allow
users to freely weigh and manipulate datasets
(FFA System?) Get FFPI information into FFMP
(format issue)