Generalization for The National Map with emphasis on the National Hydrography Dataset Lawrence Stanislawski, Science Applications International Corporation

Generalization for Generalization for The National Map The National Map with emphasis onwith emphasis on

the National Hydrography Datasetthe National Hydrography Dataset

Lawrence Stanislawski, Science Applications International Corporation (SAIC)Lawrence Stanislawski, Science Applications International Corporation (SAIC)Michael Starbuck, U.S. Geological SurveyMichael Starbuck, U.S. Geological SurveyMichael Finn, U.S. Geological SurveyMichael Finn, U.S. Geological SurveyE. Lynn Usery, U.S. Geological SurveyE. Lynn Usery, U.S. Geological SurveyPat Turley, U.S. Geological SurveyPat Turley, U.S. Geological SurveyJosh Wilkerson, U.S. Geological SurveyJosh Wilkerson, U.S. Geological Survey

2005 ESRI International User Conference, July 25-29. San Diego, CA.

Overview of The National MapOverview of The National Map

Project objectivesProject objectives

Generalization StrategyGeneralization Strategy

National Hydrography Dataset – Flowline networkNational Hydrography Dataset – Flowline network

Geometric characteristics of hydrographic networkGeometric characteristics of hydrographic network

Test DataTest Data

ResultsResults


The National MapThe National Map

•Consistent framework for geographic knowledge needed by the Consistent framework for geographic knowledge needed by the Nation Nation •Provides public access to high-quality, geospatial data andProvides public access to high-quality, geospatial data andinformation from multiple partners to help support decision-information from multiple partners to help support decision-making by resource managers and the public. making by resource managers and the public. •Product of a consortium of Federal, State, and local partnersProduct of a consortium of Federal, State, and local partnerswho provide geospatial data to enhance America's ability towho provide geospatial data to enhance America's ability toaccess, integrate, and apply geospatial data at global, national,access, integrate, and apply geospatial data at global, national,and local scales. and local scales. •Includes eight primary data themes:Includes eight primary data themes:

Orthoimagery,Orthoimagery,Elevation,Elevation,Hydrography,Hydrography,Transportation,Transportation,Structures,Structures,Boundaries, Boundaries, Geographic names, Geographic names, Land use/land cover, Land use/land cover,

•Most themes are populated, and content will continue to beMost themes are populated, and content will continue to beadded and updatedadded and updated


The National MapThe National Map

•Created from many data sources and Web Map Services Created from many data sources and Web Map Services (WMS) (WMS)

owned by many organizations.owned by many organizations.•Data sources are electronically stored at various scales and Data sources are electronically stored at various scales and resolutions. Consequently, when data are extracted from several resolutions. Consequently, when data are extracted from several WMS and horizontally integrated in geospatial data applications, WMS and horizontally integrated in geospatial data applications, the differences between data sources become apparent. the differences between data sources become apparent. •The ability to render one or more scales of The ability to render one or more scales of The National MapThe National Map data into an appropriate, or functionally equivalent, data into an appropriate, or functionally equivalent,

representation at a user-specified scale could substantially representation at a user-specified scale could substantially enhance analysis capabilities of these data.enhance analysis capabilities of these data.

Research Topic: development of automated generalization approach that renders a functionally equivalent dataset at a user-specified scale, focusing on the National Hydrography Dataset (NHD) layer of The National Map.


The National MapThe National MapNational Hydrography Dataset (NHD)National Hydrography Dataset (NHD)


National Hydrography Dataset

• Vector data layer of The National Map representing surface waters of the United States.

• Includes a set of surface water reaches

Reach: significant segment of surface water having similar hydrologic characteristics, such as a stretch of river between two confluences, a lake, or a pond.

A unique address, called a reach code, is assigned to each reach, which enables linking of ancillary data to specific features and locations on the NHD.

Reach code from Lower Mississippi subbasin

08010100000413

region-subregion-accounting unit-subbasin-reach number

08010100000696

08010100000413

The National MapThe National MapNational Hydrography Dataset (NHD)National Hydrography Dataset (NHD)


National Hydrography Dataset

• Divided and distributed at watershed basin and subbasin boundaries.

• Stored in an ArcGIS geographic database (geodatabase) model.

• Three levels of detail (resolutions)

Medium (1:100,000-scale source)

only complete layer

High (1:24,000-scale source)

three-quarters complete

Local (1:12,000 or larger source)Subregions along northern shore of Gulf of Mexico

ObjectivesObjectives


Develop a generalization strategy that can be implemented on subsets of the NHD.

Strategy should produce a dataset in the NHD model format that maintains:• feature definitions, • reach delineations, • feature relationships, and • flow connections between remaining generalized features.

Extracted dataset should function with NHD applications • less detail • faster processing speed

Extracted level of detail:• user-specified, or • defined by display scale

Generalization ApproachGeneralization Approach


Base data: user specified, or highest resolution that covers desired area.

Feature pruning – removal of feature that are too small for desired output scale.• select a subset of network features• select a subset of area features• remove point features associated with pruned line or area features

Feature simplification• removal of vertices• aggregation, amalgamation, merging, etc.

Focus so far has been on network pruning

National Hydrography Dataset (NHD)National Hydrography Dataset (NHD)Flowline NetworkFlowline Network


Drainage network: flow-oriented digital lines that represent the connectivity and flow between surface water features.

Network features types:• Stream/river• Canal/ditch• Pipeline• Connector• Artificial path

Connector represents a path where surface flow is known to exist but was not included in the source material.

Artificial path represents a flow path through an areal water feature that is connected to the drainage network. Areal features that may include artificial paths are:

area of complex channel, canal/ditch, estuary, ice mass, lake/pond, playa, reservoir, stream/river, swamp/marsh, and wash

NHD FeaturesNHD Features

Areal Stream/River

Areal Lake/Pond

Areal Lake/Pond

Linear Streams

Linear Canal/Ditch

Linear Connector

Artificial Paths

National Hydrography Dataset (NHD)National Hydrography Dataset (NHD)Flowline NetworkFlowline Network


MethodsMethodsDrainage network characteristicsDrainage network characteristics


Quantities used to compare geometric characteristics of a stream drainage network:

Catchment area – surface runoff in this area flows into the associated network segment.Upstream arbolate sum – sum of the lengths of all upstream network features that flow

into a point of a network.Drainage proportion – proportion of the arbolate sum at a point on the network to the

total length of the drainage network.Upstream drainage area – sum of all upstream catchment areas up to a point in the

network.Upstream drainage density – ratio of the arbolate sum to the upstream drainage area at

a point in the network.Basin drainage density – the ratio of length of all network features to the sum of all

catchment areas in the network. The sum of all catchment areas in the network may also be referred to as the basin area.

Number of confluence-to-confluence segments – the number of segments between network feature confluences (intersections) that exist in the network.

Stream order – method of numbering network segments within a drainage system, in which the smallest unbranched mapped tributary is called first order. Segments where two or more first order tributaries intersect becomes a second order segment and so on (Horton 1945).

Bifurcation ratio – the ratio of the number of segments of any order to the number of segments of the next highest order.

Basin bifurcation ratio – the average of the bifurcation ratios of each stream order in the drainage network.

MethodsMethods


Network drainage characteristics summarized by NHD subbasin

Arc Macro Language Programs

Python geoprocessing scripts

Rapid method that uses Thiessen polygons to the estimate catchment area for each network drainage feature was developed for these computations

MethodsMethods


Flowline Network Pruning

• Extract the most prominent network features.

• Eliminate network features having an upstream

drainage area that is less than a

threshold. Features that drain larger areas are more likely to be

maintained than features that drain smaller areas. Minimizes the effect of disparate data collection. Produces a uniform distribution of network features.

• Iteratively increase upstream drainage

area threshold until desired output drainage

density is achieved.Requires scale-dependency for ease of use.

MethodsMethodsFlowline Network Pruning


Variation in density of mapped drainage features along 7.5-minute quadrangle boundaries for medium resolution NHD. Background colors delineate subbasins.

MethodsMethodsFlowline Network PruningFlowline Network Pruning


Modeled relationship between • Elevation-derived drainage density• Minimum drainage area for stream formation

Elevation data extracted from the National Elevation Dataset (NED, USGS 2005) and resampled to a 30 meter cell size.

Approximate stream channels were derived from the elevation data through commonly used methods.

Varied minimum drainage area for stream formation on the flow accumulation grid between 0.135 km2 (150 cells) to 90 km2

(100,000 cells)

Completed for two study areas.

MethodsMethods


Gasconade-Osage subregion (1029) falls in the Interior Plains and Interior Highlands physiographic divisions.

MethodsMethods


Colorado Headwaters and Gunnison subregions (1401 and 1402) fall in Intermontane Plateaus and Rocky Mountain System physiographic divisions.

MethodsMethods


Determined that an approximate linear relationship should exist between the square root of map scale and drainage density derived from mapped network features.

Approximate drainage densities determined for the two study areas at four scales of hydrographic vector data:

• 1:24,000, NHD

• 1:100,000, NHD

• 1:500,000, river Reach File version 1 (RF1)

• 1:2,000,000, DLG data

ResultsResultsDrainage Network CharacteristicsDrainage Network Characteristics


Subbasin drainage density (km/km2) for medium resolution NHD



Sample Dataset from Gasconade-Osage subregion:•Pomme De Terre (PDT) subbasin (NHD 10290107).•Medium level of drainage density difference between the high and medium resolution NHD subbasins for the PDT (difference about 0.6421 km/km2).

Relative magnitude of drainage density differences between the high resolution layer and the medium resolution layer of the region. Darker red indicates a higher difference in density.


Pomme De Terre SubbasinPomme De Terre Subbasin


High Resolution Medium Resolution

Value Minimum Maximum Value Minimum Maximum

Reach Statistics

Number of Network Reaches 3367 1079

Average Length of Network Reaches (km) 1.02 1.91

Catchment Areas (km2) 0.001 7.903 0.031 12.846

Arbolate Sum (km) 0.018 3424 0.077 2064

Proportion of Total Drainage Flowing Through Main Outflow Reach (%) 99 100

Upstream Drainage Areas (km2) 0.035 2187.3 0.254 2190.3

Upstream Drainage Densities (km/km2) 0.0938 7.213 0.2164 2.4

Upstream Drainage Density for Main Outflow Reach (km/km2) 1.565 0.9421




High Resolution Medium Resolution


Basin Statistics

Number of Confluence-to-Confluence Segments 3986 1014

Length of Network Features (km) 3428.5 2063.6

Area of Subbasin (km2) 2190.3 2190.3

Basin Drainage Density (km/km2) 1.566 0.9421

Stream Order on Confluence-to-Confluence Segments 1 15 1 6

Bifurcation Ratios 1 4.18 1 8.67

Basin Bifurcation Ratio 2.25 4.28




Divergent to convergent network segments increment stream order when one or both segments receive additional flow.




Graduated symbols of arbolate sum values by network reach for Pomme De Terre subbasin. High resolution (green), Medium resolution (purple)

Results: Drainage Network CharacteristicsResults: Drainage Network CharacteristicsPomme De Terre SubbasinPomme De Terre Subbasin


High Resolution Subbasin 10290107Reach summary

-50

0

50

100

150

200

250

300

350

400

0 2 4 6 8 10 12 14

Catchment Drainage Density (km/sq km)

Freq

uen

cy

Medium Resolution Subbasin 10290107Reach summary

-50

0

50

100

150

200

250

300

350

0 2 4 6 8 10


Fre

qu

en

cy

Frequencies of catchment drainage densities

High Res.

Medium Res.



High Resolution 10290107Reach Summary

0

1

2

3

4

5

6

7

8

9

0 1000 2000 3000 4000

Arbolate Sum (km)

Catchment area (sq km)

Upstream drainage density

Medium Resolution 10290107Reach Summary

0

2

4

6

8

10

12

14

0 500 1000 1500 2000 2500

Arbolate Sum (km)

Catchment area (sq km)

Upstream drainage density

Catchment area and upstream drainage density vs. arbolate sum

High

Medium



Thiessen-estimated catchments derived for hydrographic network features do not precisely follow the ridgelines of surface models.

Results: Flowline Network PruningResults: Flowline Network Pruning


Gasconade-Osage subregion.

Relationship between elevation-derived drainage density and minimum drainage area for stream formation.

y = 0.7181x-0.4417

R2 = 0.9988

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

0 20 40 60 80 100

Minimum Drainage Area (sq km)

Ele

vati

on

Der

ived

Dra

inag

e D

ensi

ty (

km/s

q

km)




Linear relationship predicting elevation-derived drainage density from the inverse of the square root of minimum drainage area for stream formation.

y = 0.683xR2 = 0.9961

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

0 0.5 1 1.5 2 2.5 3

1/(square root of minimum drainage area) (1/km)

Ele

vati

on

Der

ived

Dra

inag

e D

ensi

ty

(km

/sq

km

)

derived

Linear (derived)



Estimation of drainage density from desired output scale for generalization.


y = 257.33xR2 = 0.9659

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

0 0.001 0.002 0.003 0.004 0.005 0.006 0.007

Square Root of Scale

Dra

inag

e D

ensi

ty

Vector Hydrography(Network)

Linear (VectorHydrography(Network))



Colorado Headwaters and Gunnison subregions

Relationship between elevation-derived drainage density and minimum drainage area for stream formation.

y = 0.7505x-0.4767

R2 = 0.9994

0

0.5

1

1.5

2

2.5

0 20 40 60 80 100

Minimum Drainage Area (sq km)

Dra

ina

ge

De

ns

ity

(k

m/s

q k

m)




Linear relationship predicting elevation-derived drainage density from the inverse of the square root of minimum drainage area for stream formation.

y = 0.7174xR2 = 0.9955

0

0.5

1

1.5

2

2.5

0 0.5 1 1.5 2 2.5 3

1/(square root of minimum drainage area) (1/km)

Dra

inag

e D

ensi

ty (

km/s

q k

m)

Derived (1401 1402)

Linear (Derived (1401 1402))



Estimation of drainage density from desired output scale for generalization.


y = 236.5xR2 = 0.97

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

0 0.002 0.004 0.006 0.008

Square root of scale

Dra

inag

e D

ensi

ty (

km/s

q k

m)

Vector Hydrography(Network)

Linear (Vector Hydrography(Network))



Pomme De Terre network generalization

Tested process to remove less significant network features.

Goal: Prune high resolution network to resemble medium resolution network (drainage density 0.942 km/km2)

Prune using minimum upstream drainage area.

Iteratively increase upstream drainage area threshold until desired output basin drainage density is achieved.

Results: Pruned high resolution to minimum upstream drainage area of 2.158 km2 to yield 0.939 km/km2 basin drainage density.




Medium Resolution Generalized


Reach Statistics

Number of Network Reaches 1079 1648

Average Length of Network Reaches (km) 1.91 1.25

Catchment Areas (sq. km) 0.031 12.846 0.00135 8.263

Arbolate Sum (km) 0.077 2064 0.0169 2056

Proportion of Total Drainage Flowing Through Main Outflow Reach (%) 100 100

Upstream Drainage Areas (sq. km) 0.254 2190.3 0.0943 2189.97

Upstream Drainage Densities (km/sq. km) 0.2164 2.4 0.0261 2.722

Upstream Drainage Density for Main Outflow Reach (km/sq. km) 0.9421 0.9387




Medium Resolution Generalized


Basin Statistics

Number of Confluence-to-Confluence Segments 1014 1332

Length of Network Features (km) 2063.6 2056.7

Area of Subbasin (sq. km) 2190.3 2190.3

Basin Drainage Density (km/sq. km) 0.9421 0.939

Stream Order on Confluence-to-Confluence Segments 1 6 1 14

Bifurcation Ratios 1 8.67 1 4.31

Basin Bifurcation Ratio 4.28 2.13




Graduated symbols of arbolate sum values by network reach for Pomme De Terre subbasin. High resolution (green), Medium resolution (purple), Generalized (pink).




Medium resolution (orange) over generalized (pink)

Generalized (pink) over medium resolution (orange)




Generalized (pink) over high resolution (green)

To summary




0

50

100

150

200

250

300

350

400

450

500

0 5 10 15 20


Fre

qu

en

cy

Drainage density distribution of reach catchments for Pomme De Terre subbasin of high-resolution NHD network generalized to 0.942 km/km2 basin drainage density. Maximum frequency at 1.32 km/km2




Catchment area and upstream drainage density relationships with arbolate sum for high-resolution Pomme De Terre subbasin generalized to 0.942 km/sq km basin drainage density

0

1

2

3

4

5

6

7

8

9

0 500 1000 1500 2000 2500

Arbolate Sum (km)

Catchment area (sqkm)

Upstream drainagedensity (km/sq km)




Summary of stream length by stream order for high resolution, medium resolution, and generalized high resolution Pomme De Terre subbasin.

Medium Resolution

0

500

1000

1500

2000

2500

0 5 10 15

Stream Order

Su

m o

f L

en

gth

(k

m)

Generalized High Resolution

0

500

1000

1500

2000

2500

0 5 10 15

Stream Order

Su

m o

f L

en

gth

(k

m)

High Resolution

0

500

1000

1500

2000

2500

0 5 10 15 20

Stream Order

Su

m o

f L

en

gth

(k

m)




Summary of stream frequency by stream order for high resolution, medium resolution, and generalized high resolution Pomme De Terre subbasin.

High Resolution

0

500

1000

1500

2000

0 3 6 9 12 15

Stream Order

Fre

qu

en

cy

Generalized High Resolution

0

500

1000

1500

2000

0 3 6 9 12 15

Stream Order

Fre

qu

en

cy

Medium Resolution

0

500

1000

1500

2000

0 3 6 9 12 15

Stream Order

Fre

qu

en

cy

SummarySummary


Identified and automated computations for quantities that characterize an NHD flowline network.

Reach and basin statistics, along with visual inspections assist analysis and comparison of the network drainage.

Identified a linear relationship to estimate watershed drainage density of elevation-derived streams from the minimum upstream drainage area for stream formation.

Identifed a linear relationship to estimate drainage density appropriate for a user-specified map scales, to be used for network pruning.

Developed and tested a network pruning strategy that provides a user-specified network drainage density based on upstream drainage area, which also maintains reaches and network connectivity.

Network pruning process furnishes a network that is structurally similar to a target dataset. However, when compared to the target dataset, a larger number of shorter, confluence-to-confluence segments and reaches were maintained in the generalized network because that was a pre-existing condition of the source dataset.

SummarySummary


Upstream drainage area quickly rises to overwhelming sizes. We may need to implement a numeric scaling technique to the pruning strategy.

Additional work is required to tailor the pruning strategy to be affected by terrain or topographic variations.

Next steps in developing an automated generalization include defining a pruning strategy for area features and rules for feature simplification, and also implementing a line simplification process.

Aside from enhancing our understanding of the relationship between terrain and surface hydrography, it is possible that similar automated pruning strategies can be developed for other vector layers of The National Map.

ReferencesReferences


Gold, C. M. and D. Thibault. 2001. Map Generalization by Skeleton Retraction, In Proceedings, International Cartographic Association, Beijing,China, pp.2072-2081.

Horn, R.C., L. McKay, and, S.A. Hanson. 1994. " History of the U.S. EPA's River Reach File: A National Hydrographic Database Available for ArcInfo Applications ," in Proceedings of the Fourteenth Annual ESRI User Conference, Environmental Systems Research Institute, Redlands, CA.

Horton, R.E. 1945. Erosional development of streams and their drainage basins: Hydrophysical approach to quantitative morphology. Geological Society of America Bulletin, 56, 275-370.

Maidment, D. R. 1996. GIS and hydrologic modeling – an assessment of progress. Third International Conference on GIS and Environmental Modeling. January 22-26, 1996. Sante Fe, New Mexico.

Northcott, W. 2002. Lecture notes to Biosystems Engineering 481: Agricultural and Small Watershed Hydrology. College of Engineering, Michigan State University.

Pidwirny, M. 2004. Stream Morphometry. Accessed June 9, 2005, at URL: http://www.physicalgeography.net/fundamentals/10ab.html

Sanchez, R.R. 2002. GIS-Based Upland Erosion Modeling, Geovisualization, and Grid Size Effects on Erosion Simulations with CASC2D-SED. Ph.D. Dissertation, Civil Engineering Dept., Colorado State University.

http://www.voronoi.com/pdfs/2000-2003/Map_Generalization_by_Skeleton_Retraction.pdf

http://www.epa.gov/OWOW/NPS/rf/esripapr.html.bak



http://www.physicalgeography.net/fundamentals/10ab.html

ReferencesReferences


Smith, P.N.H, and D.R. Maidment. 1995. Hydrologic Data Development System. Report 95-1 Center for Research in Water Resources. University of Texas at Austin.

Tarbotom, D. G. and D.P. Ames. 2001. Advances in the mapping of flow networks from digital elevation data. World Water and Environmental Resources Congress, May 20-24,2001, Orlando, Florida.

U.S. Geological Survey. 2005. National Elevation Dataset. Accessed June 7, 2005, at URL: http://ned.usgs.gov/.

U.S. Geological Survey. 2004. Physiographic divisions of conterminous U.S. Accessed June 14, 2005, at URL: http://aa179.cr.usgs.gov/metadata/wrdmeta/physio.htm.

U.S. Geological Survey. 2003. Streams and Waterbodies of the United States. Accessed June 14, 2005, at URL: http://nationalatlas.gov/atlasftp.html.

U.S. Geological Survey. 2000. The National Hydrography Dataset: Concepts and Contents. Accessed June 8, 2005, at URL http://nhd.usgs.gov/chapter1/index.html.

U.S. Geological Survey. 1955. Map Publication Scales. Book 1, Part B, Chapter 1 of the Geological Survey Topographic Instructions. March 1955. 13pp.

Verdin, K. L. 1997. A system for topologically coding global drainage basins and stream networks. 1997 ESRI International GIS User Conference Proceedings.

http://ned.usgs.gov/

http://aa179.cr.usgs.gov/metadata/wrdmeta/physio.htm

http://nationalatlas.gov/atlasftp.html

http://nhd.usgs.gov/chapter1/index.html

Questions?Questions?

Generalization for Generalization for The National Map The National Map wwith emphasis onith emphasis onthe National Hydrography Datasetthe National Hydrography Dataset

Lawrence Stanislawski, Lawrence Stanislawski, [email protected]@usgs.govMichael Starbuck, U.S. Geological SurveyMichael Starbuck, U.S. Geological SurveyMichael Finn, U.S. Geological SurveyMichael Finn, U.S. Geological SurveyE. Lynn Usery, U.S. Geological SurveyE. Lynn Usery, U.S. Geological SurveyPat Turley, U.S. Geological SurveyPat Turley, U.S. Geological SurveyJosh Wilkerson, U.S. Geological SurveyJosh Wilkerson, U.S. Geological Survey


Pruning test on Gasconade-Osage subregionPruning test on Gasconade-Osage subregion

Green: 1:100,000Green: 1:100,000

Blue: 1:500,000Blue: 1:500,000

Red: 1:2,000,000Red: 1:2,000,000


Difference in Network Difference in Network DistributionDistribution

Green: original high Green: original high resolution resolution

data.data.

Blue: pruned data Blue: pruned data

(1:100,000)(1:100,000)


Difference in Difference in Network DistributionNetwork Distribution

Green: original high Green: original high resolution resolution

data.data.Blue: pruned data Blue: pruned data

(1:500,000) (1:500,000)


Future Work: (Truman Future Work: (Truman Reservoir Area)Reservoir Area)

Tan: WaterbodyTan: Waterbody

Thick Blue: 1:100,000Thick Blue: 1:100,000

Thin Blue: Original dataThin Blue: Original data

Red: Pruned network Red: Pruned network (1:100,000)(1:100,000)


Documents

Generalization for The National Map with emphasis on the National Hydrography Dataset Lawrence Stanislawski, Science Applications International Corporation