functional linkage of water basins and streams: flows v1 arcgis tools

Functional Linkage Functional Linkage of Water Basins and of Water Basins and

Streams: Streams: FLoWS v1 ArcGIS FLoWS v1 ArcGIS

toolstoolsDavid Theobald, John Norman, David Theobald, John Norman,

Erin PetersonErin PetersonNatural Resource Ecology Lab, Dept of Natural Resource Ecology Lab, Dept of

Recreation & Tourism, Colorado State Recreation & Tourism, Colorado State UniversityUniversity

Fort Collins, CO 80523 USA Fort Collins, CO 80523 USA 17 17 May 2006May 2006

Project contextProject context

Challenges of STARMAP (EPA STAR):Challenges of STARMAP (EPA STAR): Addressing science needs Clean Water ActAddressing science needs Clean Water Act Integrate science with states/tribes needsIntegrate science with states/tribes needs

Develop landscape-based indicators to Develop landscape-based indicators to assist in testing tenable hypotheses assist in testing tenable hypotheses generated using understanding of generated using understanding of ecological processesecological processes

PremisePremise

Challenges to develop improved landscape-scale indicators (Fausch et al. 2002; Gergel et al. 2002; Allan 2004) are:- clearer representation of watersheds & hierarchical relationship;- incorporate nonlinearities of condition among different watersheds and along a stream segment

Need to characterize spatial heterogeneity & scaling of watersheds when developing indicators of biological condition

Goal: to develop indicators that more closely represent our understanding of how ecological processes are operating

From From watersheds/catchments as watersheds/catchments as hierarchical, overlapping hierarchical, overlapping

regions…regions…

River continuum concept (Vannote et al. 1980)

““Lumped” or watershed-based Lumped” or watershed-based analysesanalyses % agricultural, % urban (e.g., ATtILA)% agricultural, % urban (e.g., ATtILA)

Average road density (Bolstad and Swank)Average road density (Bolstad and Swank) Dam density (Moyle and Randall 1998)Dam density (Moyle and Randall 1998) Road length w/in riparian zone (Arya 1999)Road length w/in riparian zone (Arya 1999) But ~45% of HUCs are not watershedsBut ~45% of HUCs are not watersheds

EPA. 1997. An ecological assessment of the US Mid-Atlantic Region: A landscape atlas. EPA ATtILA 2002.

… … to network of to network of catchmentscatchments

Network Dynamics Hypothesis - Benda et al. BioScience 2004

Reaches linked to Reaches linked to catchmentscatchments

1 to 1 1 to 1 relationshiprelationship

Properties of Properties of the the watershed watershed can be linked can be linked to network to network for for accumulation accumulation operation operation

Covariates: landscape Covariates: landscape contextcontext1.1. Co-variate(s) at spatial location, site contextCo-variate(s) at spatial location, site context

- E.g., geology, elevation, population density at a point- E.g., geology, elevation, population density at a point

2.2. Co-variate(s) within some distance of a locationCo-variate(s) within some distance of a location- Housing density at multiple scales- Housing density at multiple scales

3.3. Watershed-based variablesWatershed-based variables- Proportion of urbanized area- Proportion of urbanized area

4.4. Spatial relationships between locationsSpatial relationships between locations- Euclidean (as the crow flies) distance between points- Euclidean (as the crow flies) distance between points- Euclidean (as the fish swims) hydrologic network - Euclidean (as the fish swims) hydrologic network

distance between pointsdistance between points

5.5. Functional interaction between locationsFunctional interaction between locations- Directed process (flow direction), anisotropic, multiple - Directed process (flow direction), anisotropic, multiple

scalesscales- How to develop spatial weights matrix?- How to develop spatial weights matrix?- Not symmetric, stationary - Not symmetric, stationary violate traditional violate traditional

geostatistical assumptions!?geostatistical assumptions!?

Local vs. accumulated Local vs. accumulated (e.g., Human Urban Index)(e.g., Human Urban Index)

LocalLocal

AccumulatedAccumulated

AccumulatedAccumulated

USGSNHD,NED

USGSNHD,NED

Pre-processingPre-processingGenerating reach contributing Generating reach contributing

areas (RCAs)areas (RCAs)Automated delineationAutomated delineation

Inputs: Inputs: stream network (from stream network (from

USGS NHD or other)USGS NHD or other) topography (USGS NED, 30 topography (USGS NED, 30

m)m)

Processes:Processes: 1. traditional WATERSHED 1. traditional WATERSHED

command requires FILLed command requires FILLed DEM – “hydro-conditioned”DEM – “hydro-conditioned”

2. Cost-distance using 2. Cost-distance using Topographic Wetness & Topographic Wetness & Position IndicesPosition Indices

“true” catchments

“adjoint”catchments

Segments

Generating RCAs: FILLedGenerating RCAs: FILLed1.) Filled DEM 2.) Flow Direction

Artifacts?Artifacts?

Generating RCAs: cost-Generating RCAs: cost-distancedistance1.) DEM

2.) Topographic Wetness Index 3.) Topographic Position Index

Generating RCAsGenerating RCAs 4) Stream Reaches 5.) RCAs (Yellow)

Evaluation of RCAsEvaluation of RCAs

““Truth”Truth” Hand-delineated from 1:24KHand-delineated from 1:24K

Modeled (1:100K, 30 m DEM):Modeled (1:100K, 30 m DEM): A. traditional (FILL-ing)A. traditional (FILL-ing) B. cost-distanceB. cost-distance

Measure: Jaccard’s similarity Measure: Jaccard’s similarity coefficient: coefficient: b / (a + b + c)b / (a + b + c)

a

b

c

Preliminary resultsPreliminary resultsFILLed DEM 50 m/WATERSHEDMean accuracy: 78%

Cost-distance RCAsMean accuracy: 85%

Within RCA hydro-Within RCA hydro-weightingweighting

Overland flow Overland flow (hydro distance to stream)(hydro distance to stream)

Instream flow Instream flow

(hydro network distance to outlet)(hydro network distance to outlet)

Landscape NetworkLandscape NetworkLandscape network features and associated relationships table

From graph theory perspective, reaches are nodes, confluences are edges

Network connectivity errors

SelectionsSelections User-defined fieldUser-defined field Polylines or RCAsPolylines or RCAs Cumulative Cumulative

(distance from (distance from selected feature)selected feature)

AnalysisAnalysis

Estimated dischargeEstimated discharge

Average annual Average annual precipitation & precipitation & temperature, basin temperature, basin areaarea

Vogel et al. 1999Vogel et al. 1999

Vogel

AnalysisAnalysis

Export to distance Export to distance matricesmatrices

Straight-line Instream distance

Distance matrices (cont.)Distance matrices (cont.)Downstream only Upstream only

Distance matrices (cont.)Distance matrices (cont.)Proportion upstream Proportion downstream

Distance matrices (cont.)Distance matrices (cont.)Downstream portion dist onlyNumber of confluences

Example: Example: Coho Coho

salmon salmon distancesdistances

SummarySummary River Continuum to NetworkRiver Continuum to Network

From overlapping waterbasins to network spatial From overlapping waterbasins to network spatial structurestructure

OpenOpen Simple data structureSimple data structure Python linked to GeoProcessing objectPython linked to GeoProcessing object Non-GIS (thru Access, SQL, etc.)Non-GIS (thru Access, SQL, etc.)

FlexibleFlexible User-defined variables to accumulate, navigate networkUser-defined variables to accumulate, navigate network Different selection sets, combinationsDifferent selection sets, combinations Compute framework once, use with many point Compute framework once, use with many point

configurations (samples)configurations (samples) RobustRobust

Flow-based vs. Strahler stream orderFlow-based vs. Strahler stream order Cost-weighted methodsCost-weighted methods Developed, tested (broken), refinedDeveloped, tested (broken), refined E.g, Mid-Atlantic Highlands; Oregon; Central Shortgrass E.g, Mid-Atlantic Highlands; Oregon; Central Shortgrass

Prairie; Alaska; Prairie; Alaska;

Next stepsNext steps Project/tool website: Project/tool website:

www.nrel.colostate.edu/projects/starmapwww.nrel.colostate.edu/projects/starmap FLoWS, FunConn, RRQRRFLoWS, FunConn, RRQRR

FLoWS database to complement toolsFLoWS database to complement tools Attach additional attributes to FLoWS Attach additional attributes to FLoWS

databasedatabase Land cover (urban, ag, “natural”)Land cover (urban, ag, “natural”) Historical, current, future housing densityHistorical, current, future housing density Hydro & slope weighted road densityHydro & slope weighted road density Human accessibilityHuman accessibility

Within reach/segmentWithin reach/segment Streams as 2D featuresStreams as 2D features

SCALE: Grain

Substrate

Biotic Condition

OverhangingVegetation

Segment

River Network

Network Connectivity

Tributary Size DifferencesNetwork Geometry

Stream Network

ConnectivityFlow Direction Network Configuration

Drainage DensityConfluence Density

Cross Sectional AreaChannel Slope, Bed MaterialsLarge Woody Debris

Biotic Condition, Substrate Type, Overlapping VegetationDetritus, Macrophytes

Microhabitat

Segment Contributing Area

Riparian Vegetation Type & ConditionFloodplain / Valley Floor Width

Localized DisturbancesLand Use/ Land Cover

Landscape

ClimateAtmospheric depositionGeology

TopographySoil Type

Microhabitat

ShadingDetritus Inputs

Riparian Zone

Nested Watersheds

Land UseTopography

Vegetation TypeBasin Shape/Size

COARSE

FINE

Reach

Terrestrial Aquatic

Peterson 2005

Example: 2D stream in Example: 2D stream in VirginiaVirginia

Example: 2D stream in Example: 2D stream in VirginiaVirginia

Example: 2D stream in Example: 2D stream in VirginiaVirginia

Example: 2D stream in Example: 2D stream in VirginiaVirginia

Thanks!Thanks! Comments? Questions?Comments? Questions? Thanks to K. Verdin at USGS EROS Data Thanks to K. Verdin at USGS EROS Data

Center for sharing EDNA datasetsCenter for sharing EDNA datasets Funding/Disclaimer: The work reported Funding/Disclaimer: The work reported

here was developed under the STAR here was developed under the STAR Research Assistance Agreement CR-Research Assistance Agreement CR-829095 awarded by the U.S. 829095 awarded by the U.S. Environmental Protection Agency (EPA) to Environmental Protection Agency (EPA) to Colorado State University. This Colorado State University. This presentation has not been formally presentation has not been formally reviewed by EPA.  The views expressed reviewed by EPA.  The views expressed here are solely those of the presenter and here are solely those of the presenter and STARMAP, the Program (s)he represents. STARMAP, the Program (s)he represents. EPA does not endorse any products or EPA does not endorse any products or commercial services mentioned in this commercial services mentioned in this presentation.presentation.

FLoWS: FLoWS: www.nrel.colostate.edu/projects/stawww.nrel.colostate.edu/projects/starmaprmap

davet@nrel.colostate.edudavet@nrel.colostate.edu

CR - 829095

Water basin - StreamWater basin - Stream

Hydrologic Hydrologic distance:distance:-InstreamInstream-Up vs. Up vs. down?down?

FLoWSFLoWS

Overlapping Overlapping watershedswatersheds

Accumulate Accumulate downstreamdownstream

FLoWS (and FLoWS (and SPARROW)SPARROW)

Stand-alone Stand-alone watershedwatershed

Watershed-based Watershed-based analyses (HUCs)analyses (HUCs)

Tesselation of Tesselation of true, adjoint true, adjoint catchments catchments

??

Watershed HUCs/WBD Reach Contributing Areas (RCAs)

Grain (Resolution)

Pro

cess

/Fu

nct

ional

Zonal

A

ccum

ula

te

U

p/d

ow

n

(net.

)