Sinkholes Modeling, and Soil Survey. Objectives Examine a few techniques and efforts to model karst features for map unit and/or component delineation.

Sinkholes

Modeling, and Soil Survey

Objectives

• Examine a few techniques and efforts to model karst features for map unit and/or component delineation.

• Consider some case studies of mapping and updating karst areas:– Lawrence co, Indiana– Springfield plateau, Missouri

Karst: Dade county, MO

GIS techniques to enhance identification of karst

Same area: 10m DEM, hillshade relief, hypso lines

Sinkholes: 3x vertical exaggeration, 5m display resolution (10m DEM)

Flowaccumulation (multi-path): karst area

Didn’t work that well…

Karst: wetness index (multi-path)

Works fairly well for some sinkholes.

However, not all upland wet areas are sinkholes.

Can we model these sinkholes?

• Moderately high elevationAvg about 325m; range about 315 to 335m

• High Flow accumulationAbove about 200

• High Wetness IndexAbove about 9

• Not on the stream networkNot so easy to model

Fuzzy Membership Model for SinksHigher-elevation valleys confounded w sinks

What about using streams in the model?

USGS streams layer is incomplete (missing tributaries)

Modeled stream network runs into sinks

Fuzzy Membership Model w more restrictive lower elevation criteria

(lost the lowest elevation sinkhole)

Karst plain: Pierpont area, Boone county, MO

Simple Fuzzy Membership model for Pierpont sinkhole areas:

High elevation (full membership above 225m)

High wetness index (full membership above 5)

Does an OK job of identifying areas where the sinkholes occur

Mapping karst in SSURGO

• How can we identify karst features in…– Spatial data– Attribute data

• How might this affect interpretations?• Case studies: karst in soil survey– Lawrence county, IN– Springfield Plateau (MLRA 116B), MO

Karst mapping:

Lawrence county, IN

Acknowledgements:Dena Marshall (now Lexington KY)Genny Helt (Indianapolis IN)

Physiographic Features of southern Indiana (Schneider, 1966)

A closer look at features of the Crawford Upland & Mitchell Plain

Highland Rim & Pennyroyal

KY & IN Sandstone & Shale Hills & Valleys

KY & IN Sandstone & Shale Hills & ValleysLawrence county

- 30m DEM - MLRAs

How do the lines look?

Where’s the karst?

Mitchell Plain

Norman Upland

Crawford Upland

Physiographic Regions

Mitchell Plain:middle Mississippian- aged limestones

Karst map units in Lawrence county, IN

Note: no sinkhole spot symbols

The karst soils in Lawrence county

• “Terra rossa”• Loess over “residuum” (in situ?)• Crider– Fi-si Typic Paleudalfs– 10R w depth

• Knobcreek (Frederick correlated to this in IN)– Fi-si / clayey Typic Paleudalfs– Not as red

Mapping patterns & rules

• Limestone underlain w siltstone: little/no karst.

• Limestone thickens westerly; sinkholes appear.• IN Geomorphology Tour “some years ago”:– Karst map unit:• Areas w sinkhole density of > 1 per 5 acres

– Spot symbol:• Sinkhole density < 1 per 5 acres

Ref: Pers. Comm., Dena Marshall, SS, currently Lexington KY

Attributes (NASIS soil properties)

• Bedrock in a karst component within 80”– Higher Ksat than corresponding non-karst unit

• Entry for Ksat in R layer

– E.g., Navilleton components, Floyd co IN:• Karst component: R Ksat 1.42 – 141.14• Non-karst component: R ksat 0.43 – 42.34

• Bedrock below 80”– No known differences in soil properties between karst

and non-karst map units• If we could populate soil property data below 80”, there would be differences.

MLRA Soil Survey Project:Updating Karst Areas on the Springfield Plateau (MLRA 116B)

• Ignored in some,• Spot symbols in some,• Karst map units in some.• Inconsistent use of all of

the above, both within and among counties.

County-based soil surveys (over 40+ years) handled karst in various ways.

Springfield Plateau

Karst Project: Methodology

• Primarily a GIS office project• Utilize:

– Elevation data• DEMs, Hypsography lines

– Existing SSURGO– Orthophotos– Sinkholes layer (statewide; USGS)

• On-screen digitizing/editing of SSURGO• Establish new map units

– “Karst” in name, or– Soil series that occur only in sinkhole basins.

Results: 20 new map units in 7 counties• “Karst” map units; e.g.,

– Goss grsil, karst, 8-15%

• Sinkhole basin map units; e.g.,– Lowassie sil, 0-3%, freq.

ponded

Zoom in to this area (next slide)

Sinkholes point layer over 10m DEM, with SSURGO polygons

Map units added in this area

Blue = sinkhole basin units; Red = “Karst” in MU Name

Karst discussion “answers”• Spatial identification of karst:

– Map unit phases (i.e., “karst” in muname).– Spot symbols.– Map units named by components that only occur in karst (i.e., no need for

“karst” in muname).• e.g., in Missouri: Lowassie, Grandgulf series only occur in sinkholes.

• Attribute interpretation– Muname contains “karst”– Component geomorphic description table: landform = “sinkhole”– Differences in soil properties?

• Interpretations– Possibility for groundwater contamination– Onsite wastewater disposal; lagoons; others?– Not keyed in standard National Interps (?)– Custom, state-created interps (e.g., MO onsite wastewater interp)

• Keys on “karst” in the muname, and “sinkhole” landform