Upload
others
View
0
Download
0
Embed Size (px)
Citation preview
Geomatics Tools for Geomatics Tools for Mapping Headwater Mapping Headwater
StreamsStreamsAdam HoggAdam Hogg
Inventory Monitoring & Assessment, Ministry of Natural Inventory Monitoring & Assessment, Ministry of Natural ResourcesResources
Headwaters WorkshopHeadwaters WorkshopMay 20, 2009May 20, 2009
Outline
The Challenge
What is a headwater stream?
Mapping history
Map scale
Suggested Methods & Data
Know your data
Data available in Southern Ontario
Two ways to map headwater streams
Accuracy assessment
Case Example – North Carolina
Outline
Current related efforts in Ontario
Ongoing projects involving geomatics
Example of how geomatics can help
Summary
The ChallengeThe Challenge
What is a Headwater Stream?
The furthest point from the watershed outlet from which flowing water is:
measurable year round……… OR
measurable at sometime in the year
First and second-order streams
streams that have no upstream tributaries or only first-order tributaries (Gomi et al. 2002, Meyer and Wallace 2001)
Wetlands?
Represent the transition zone
Where water is channelized
Constructed drainage?
Mapping History
Over the last few decades many agencies and researchers have attempted
Medium scale topographic mapping (~1:10000 – 1:24000)
Sources like OBM, USGS
Past maps:
greatly underestimate # and length
misrepresent flow duration
Unsuitable for identifying streams
Variable scale a problem
Seeing through the coniferous trees
Sources: Morisawa 1957; Hansen 2001; Firman and Jacobs 2002; Heine et al. 2004; Paybins 2002; Colson 2006
1:10 000
Map Scale
A need for large scale (detailed)
Water Resource Information Program (WRIP) & Conservation Ontario large scale hydrology mapping standards
Note: 1:2 000 stream mapping courtesy of Niagara Peninsula CA
111111
1122
2233
1111
11
Stream
1:2 000
Methods & DataMethods & Data
Know Your Data
All spatial data is a model with inherent error
Data search: Province (LIO), CA, engineering firms, your network
Data dictionary, a table including information on:
source, accuracy, scale/resolution, intended use
stream mapping, orthophoto’s, DEM’s.
Data Available in Southern Ontario
Streams & DEM’s
OrthophotosSoilsSurficial Geology
Interpretive
Spring photo’s ideal
Existing stream data
Look for hydrologic indicators
Understand the topography
Stream Prediction Model
DEM upslope contributing area
D-Infinity (Tarboton, 1997)
Topographic Index (Beven & Kirkby, 1978)
flow accumulation
local slope
soil conductivity?
Calibrate model
Interpretively with orthophoto
Other comparative and verified stream data
*In the field
Strengths & Weaknesses
Interpretive & the “human element”:+ Intelligent selection and accurate (time permitted)– A snapshot in time, expensive and time consuming,
random error
Prediction & the “computer element”+ Efficient and more cost effective, can be altered
easily, easy to standardize for large “non- jurisdictional” areas
– Less intelligence and accuracy, difficult to employ intelligent selection
Accuracy Assessment
Purpose of accuracy assessment
Crucial to understanding quality and use
Rules of thumb:
1/3 the total mapping effort
Independently evaluated
Statistically sound (>50 validation samples)
Validation source should be at least 3 times more precise
Report overall accuracy, omission, commission and absolute and/or relative spatial error
Sources:
“Assessing the accuracy of remotely sensed data: principles and practices” (Congalton and Green, 1998)
USGS – NPS Vegetation Mapping Program (USGS Center for Biological Informatics, 1994)
Case Example: North Carolina
Mapping Headwater Streams: Intermittent and Perennial Headwater Stream Model Development and Spatial Application (Russell, 2008)
Use high precision DEM for first and second order stream prediction
Survey grade GPS mapping for calibration and accuracy assessment
Case Example: North Carolina
Successes:
Spatially accurate, presence absence, length of permanent streams
Challenges:
Commission error, stream length of non-perennial streams, labeling of permanence
Other interesting findings:
Sedimentary areas similar to Southern Ontario 70% of headwater streams were intermittent or ephemeral
Bedrock dominated areas similar to Ontario Shield majority are ephemeral or perennial
Current Related Efforts In Current Related Efforts In Ontario Using GeomaticsOntario Using Geomatics
Existing Projects
Large Scale Mapping Standards
Conservation Ontario and OMNR
Ontario Headwater Restoration Initiative (OHRI)
OMNR led
Identify and map all roadside potential headwater stream and wetland restoration opportunities
Restore a few headwater systems
Oak Ridges Moraine, Lake Simcoe
Wetland loss
Ducks Unlimited Canada led
Estimate wetland loss using current land cover (SOLRIS) and soils
Losses > 10ha
Southern Ontario
CA’s?
Toronto, Niagara ,Grand River, Credit Valley
How Geomatics Could Help
OHRI field data comparison with stream prediction
Standardized mapping
Screening tool
Summary
It is possible to map headwater streams
Mapping and characterization will be challenging and potentially expensive depending project objectives
Know all data available and use a multi-layered approach
Rigorous accuracy assessment is crucial
A few related projects
Questions?
Source: GOES Project, NASA - http://goes.gsfc.nasa.gov
MitigationMitigation
ProtectionProtection
RestorationRestoration
Note: source SOLRIS wetlands
A Possible Decision Support Tool
Note: Source topographic index
MitigationMitigation
ProtectionProtection
RestorationRestoration
A Possible Decision Support Tool
Note: source SOLRIS frequently tilled agriculture
MitigationMitigation
ProtectionProtection
RestorationRestoration
A Possible Decision Support Tool
Note: Source untilled SOLRIS agriculture
MitigationMitigation
ProtectionProtection
RestorationRestoration
A Possible Decision Support Tool
Cyan Ellipse: Development MitigationCyan Ellipse: Development Mitigation
Prevent developmentPrevent development
Magenta Ellipse: ProtectionMagenta Ellipse: Protection Blue Ellipse: RestorationBlue Ellipse: Restoration
A Possible Decision Support Tool
Case Example: North Carolina
Results:
Overall accuracy 83%
Commission 17%
Omission 3%
Length
Over predicted stream length by 24%
Most error in intermittent and ephemeral
Permanency
Low accuracy ~30%
Spatial
70% were within 2.5 meters of actual centerline