Landslide ID specific for the database, includes the data source (311, USGS, D10, etc.) followed by a number (i.e., D10-1). ID numbers are based on ranked order of the landslides from each data source from south to north based on their latitudes. Slides with exactly the same latitude at database precision are ordered from west to east.

The name of the data source (i.e., 311, USGS, D10, etc., without an ID number).

The original labels of the landslide if they exist. This aims to enable linking between slides in the database to slides in the respective source list.

Decimal degrees north of the Equator.

Decimal degrees west of the Prime Meridian (by convention the westward direction is indicated by a negative value).

County where landslide occurred, inherited from data source or spatial query.

Municipality where landslide occurred, inherited from data source or spatial query.

Included for PennDOT data. Route number of route near landslide occurrence.

Included for PennDOT data. Start segment of route section where landslide occurred.

Included for PennDOT data. Offset from start segment of route section where landslide occurred (ft).

Describes whether the landslide is above (A) or below (B) the road in the context of local elevations. AB marks that it occurs both above and below.

Describes the landslide type following DCNR classification (Delano and Wilshusen, 2001) (Categories included in the database: Debris avalanche, Debris flow, Debris slide, Earthflow, Mudflow, Rock fall, Rock slump, Slump, and Soil creep, for more details see section 4.6 of the report describing the database).

The areal extent of slide materials (ft^2).

The magnitude of downslope displacement (ft) as recorded from the landslide’s head scarp to the nearest occurrence of run out material (see Figure 2.2.1 in the report).

The depth (ft) to the failure plain as recorded from field observations, cores or alike (see Figure 2.2.1 in the report).

Month of landslide occurrence.

Day of landslide occurrence.

Year of landslide occurrence.

Elevation (ft) above mean averaged sea level for the landslide polygon centroid or the landslide point location based on the smoothed DEM.

Local relief (ft) in each DEM pixel is computed from the smoothed DEM using TopoToolbox (a MATLAB based set of topographic analysis algorithms, (Schwanghart and Scherler, 2014)) and computes the difference between the highest and lowest elevations within a circular kernel whose radius is 820 ft (typical hillslope length) that is centered on each pixel.

Slope (degrees) was computed from the smoothed 10m DEM as the total gradient computed over a 3*3 kernel.

Roughness (ft) was computed as the standard deviation of elevation differences between the original DEM and DEM that is smoothed with a kernel of 150 ft radius.

Aspect (azimuthal degrees, clockwise from north) in each DEM pixel is computed from the smoothed DEM using TopoToolbox (a MATLAB based set of topographic analysis algorithms, (Schwanghart and Scherler, 2014)) and computes the slope orientation at the landslide location.

Drainage area (ft^2) in each DEM pixel is computed from the number of DEM pixels that drain to a pixel of interest, and is computed from the smoothed DEM using TopoToolbox (a MATLAB based set of topographic analysis algorithms, (Schwanghart and Scherler, 2014))).

Wetness index in each DEM pixel is computed from the natural logarithm of the ratio between the area (ft^2) that drains to the pixel and the topographic slope (ft/ft) of the pixel (both computed from the smoothed DEM using TopoToolbox (a MATLAB based set of topographic analysis algorithms, (Schwanghart and Scherler, 2014))).

Hillslope position is the relative location along a hillslope, where values of 0, 0.5, and 1 represent the base, middle and top of the hill, respectively. The hillslope position is approximated as (Z-Zb)/Re, where Z is the elevation at the landslide location, Zb is the lowermost elevation of the hill, and Re is the hill’s topographic relief (i.e., elevation difference between the hill’s top and base, Figure 2.2.2). Zb and Re are approximated by finding the minimum and maximum elevations within a circular area, centered on the landslide, whose radius equals a typical hillslope length (~820 ft).

Based on an ArcGIS algorithm that creates a raster of Euclidean distances from the nearest stream (i.e., the “Distance Accumulation” function), where streams are defined as landscape positions having drainage area larger than 4,500,000 ft2 (based on a typical drainage area threshold for channel initiation). Distance from nearest stream (ft) is then sampled from this raster.

Based on an ArcGIS algorithm that creates a raster of Euclidean distances from the nearest road (i.e., the “Distance Accumulation” function), where roads, including dirt roads, are based on the combination of State Roads, Local Roads, and Unpaved Roads (Penn State - Center for Dirt and Gravel Road Studies, 2012; Pennsylvania Department of Transportation, 2024a, 2024b). Distance from nearest road (ft) is then sampled from this raster.

Mean curvature (ft^-1) in each DEM pixel is computed from the smoothed DEM using TopoToolbox (a MATLAB based set of topographic analysis algorithms, (Schwanghart and Scherler, 2014)) and relies on the curvature of a two dimensional polynomial surface computed based on a 3*3 kernel centered at each DEM pixel (Schmidt, Evans and Brinkmann, 2003). Positive and negative values are indicative of concave (channels) and convex (ridges) landforms, respectively.

Planar curvature (ft^-1) in each DEM pixel is computed from the smoothed DEM using TopoToolbox (a MATLAB based set of topographic analysis algorithms, (Schwanghart and Scherler, 2014)) and relies on the curvature, perpendicular to the steepest descent of a two dimensional polynomial surface computed based on a 3*3 kernel centered at each DEM pixel (Schmidt, Evans and Brinkmann, 2003). Positive and negative values are indicative of upstream (channels) and downstream (ridges) curvature, respectively.

Profile curvature (ft^-1) in each DEM pixel is computed from the smoothed DEM using TopoToolbox (a MATLAB based set of topographic analysis algorithms, (Schwanghart and Scherler, 2014) and relies on the curvature, in the direction of steepest descent, of a two dimensional polynomial surface computed based on a 3*3 kernel centered at each DEM pixel (Schmidt, Evans and Brinkmann, 2003). Positive and negative values are indicative of decreased and increase slope downstream, respectively, which are typically associated with channels and ridges, respectively.

This is a name for the soil, generally based on historical soil series names generated for county soil surveys. This name allows linkage of experience based on soil survey names to the quantitative measures presented here.

Component proportion and horizon thickness weighted average of percent sand in SSURGO database (Soil Survey Staff, 2023) for the mapped soil unit at landslide polygon centroid or landslide point location on the smoothed parameter map.

Component proportion and horizon thickness weighted average of percent silt in SSURGO database (Soil Survey Staff, 2023) for the mapped soil unit at landslide polygon centroid or landslide point location on the smoothed parameter map.

Component proportion and horizon thickness weighted average of percent clay in SSURGO database (Soil Survey Staff, 2023) for the mapped soil unit at landslide polygon centroid or landslide point location on the smoothed parameter map.

Component proportion and horizon thickness weighted average of soil erodibility factor in SSURGO database (Soil Survey Staff, 2023) for the mapped soil unit at landslide polygon centroid or landslide point location on the smoothed parameter map.

Component proportion weighted average total thickness (in) of soil horizons in SSURGO database (Soil Survey Staff, 2023) for the mapped soil unit at landslide polygon centroid or landslide point location on the smoothed parameter map. Note that SSURGO data may not report the total depth of soils, and use of this parameter should incorporate that limitation.

Component proportion and horizon thickness weighted average of AASHTO soil classification in SSURGO database (Soil Survey Staff, 2023) for the mapped soil unit at landslide polygon centroid or landslide point location on the smoothed parameter map.

Pre-aggregated data in SSURGO database (Soil Survey Staff, 2023) for the mapped soil unit at landslide polygon centroid or landslide point location. These values are assigned to the soil map unit, so smoothing was not used for this parameter. [Class Description: Not rated, Slight, Moderate, Severe]

Component percent weighted sum of erosion class in SSURGO database (Soil Survey Staff, 2023) for mapped soil unit at landslide polygon centroid or landslide point location on the smoothed parameter map. [Class Description: Area of sediment deposition = 0, Class 1 (0-25% soil loss) = 1, Class 2 (25-75% soil loss) = 2, Class 3 (75-99% soil loss) = 3, Class 4 (complete soil loss) = 4]

Assigned to mapunit based on soil surveys (Soil Survey Staff, 2023) for mapped soil unit at landslide polygon centroid or landslide point location. These values are assigned to the soil map unit, so smoothing was not used for this parameter. [Class Description: Not rated, Not limited, Somewhat limited, Very limited].

Component percent weighted sum of drainage class in SSURGO database (Soil Survey Staff, 2023) for the mapped soil unit at landslide polygon centroid or landslide point location on the smoothed parameter map. [Class Description: Excessively drained =1, Somewhat excessively drained =2, Well drained = 3, Moderately well drained =4, Somewhat poorly drained = 5, Poorly drained =6, Very poorly drained =7]

Component percent weighted sum of runoff class in SSURGO database (Soil Survey Staff, 2023) for mapped soil unit at landslide polygon centroid or landslide point location on the smoothed parameter map. [Class Description: Negligible =1, Very low =2, Low =3, Medium =4, High= 5, Very high =6]

Component percent weighted sum of soil slip potential in SSURGO database (Soil Survey Staff, 2023) for mapped soil unit at landslide polygon centroid or landslide point location on the smoothed parameter map. [Class Description: Low=1, Medium =2, High=3]

Component proportion and horizon thickness weighted average of soil porosity (in3/in3) in SSURGO database (Soil Survey Staff, 2023) for mapped soil unit at landslide polygon centroid or landslide point location on the smoothed parameter map.

Horizon thickness and component proportion weighted average of horizon vertical conductivities (micrometers per second) in SSURGO database (Soil Survey Staff, 2023) for mapped soil unit at landslide polygon centroid or landslide point location on the smoothed parameter map.

Geologic formation or group (as defined by Pennsylvania Bureau of Topographic and Geologic Survey, Department of Conservation and Natural Resources (2001)) at the landslide polygon centroid or landslide point location.

This is based on the elevation (ft) difference between the topographic surface and the marker bed elevation at the location of each landslide (Northern and Central Appalachian Basin Coal Regions Assessment Team, 2001). In areas that are not underlain by a marker bed, no value is assigned.

This is based on the elevation (ft) difference between the topographic surface and the marker bed elevation at the location of each landslide (Northern and Central Appalachian Basin Coal Regions Assessment Team, 2001). In areas that are not underlain by a marker bed, no value is assigned.

The dip (degrees) was computed from the marker bed raster (Northern and Central Appalachian Basin Coal Regions Assessment Team, 2001) as the total gradient computed over a 3*3 kernel.

The dip (degrees) was computed from the marker bed raster (Northern and Central Appalachian Basin Coal Regions Assessment Team, 2001) as the total gradient computed over a 3*3 kernel.

Aspect (Measured in azimuthal degrees, clockwise from north) in each marker bed (Northern and Central Appalachian Basin Coal Regions Assessment Team, 2001) pixel is computed using TopoToolbox (a MATLAB based set of topographic analysis algorithms, (Schwanghart and Scherler, 2014)) and reflects the orientation of the marker bed dip the landslide location.

Aspect (Measured in azimuthal degrees, clockwise from north) in each marker bed (Northern and Central Appalachian Basin Coal Regions Assessment Team, 2001) pixel is computed using TopoToolbox (a MATLAB based set of topographic analysis algorithms, (Schwanghart and Scherler, 2014)) and reflects the orientation of the marker bed dip the landslide location.

Based on an ArcGIS algorithm that creates a raster of Euclidean distances (ft) from the nearest underground mine (i.e., the “Distance Accumulation” function), where underground mines are defined by the PADEP “Digitized Mined Areas” (Pennsylvania Department of Environmental Protection, 2023b). A distance of zero implies that a mine is mapped directly below the landslide.

Elevation (ft) of landslide above uppermost mapped coal contours (i.e., Pittsburgh or Upper Freeport coal (Northern and Central Appalachian Basin Coal Regions Assessment Team, 2001)) if mining occurred directly below the landslide polygon centroid or point.

The mean average precipitation (in) was computed over a period of 1, 2, 7, 30, 90, and 180 days prior to landslides that have an associated date of landslide occurrence.

The mean average precipitation (in) was computed over a period of 1, 2, 7, 30, 90, and 180 days prior to landslides that have an associated date of landslide occurrence.

The mean average precipitation (in) was computed over a period of 1, 2, 7, 30, 90, and 180 days prior to landslides that have an associated date of landslide occurrence.

The mean average precipitation (in) was computed over a period of 1, 2, 7, 30, 90, and 180 days prior to landslides that have an associated date of landslide occurrence.

The mean average precipitation (in) was computed over a period of 1, 2, 7, 30, 90, and 180 days prior to landslides that have an associated date of landslide occurrence.

The mean average precipitation (in) was computed over a period of 1, 2, 7, 30, 90, and 180 days prior to landslides that have an associated date of landslide occurrence.

The number of freeze thaw alterations over periods of 7, 30 and 365 days were counted prior to landslides that have timing information.

The mean length (days) of freeze thaw cycles over periods of 7, 30 and 365 days prior to landslides that have timing information.

The mean temperature (degrees F) of freezes over periods of 7, 30 and 365 days prior to landslides that have timing information.

The minimum temperature (degrees F) of freezes over periods of 7, 30 and 365 days prior to landslides that have timing information.

The mean temperature (degrees F) of thaws over periods of 7, 30 and 365 days prior to landslides that have timing information.

The mean temperature (degrees F) of entire period over periods of 7, 30 and 365 days prior to landslides that have timing information.

The number of freeze thaw alterations over periods of 7, 30 and 365 days were counted prior to landslides that have timing information.

The mean length (days) of freeze thaw cycles over periods of 7, 30 and 365 days prior to landslides that have timing information.

The mean temperature (degrees F) of freezes over periods of 7, 30 and 365 days prior to landslides that have timing information.

The minimum temperature (degrees F) of freezes over periods of 7, 30 and 365 days prior to landslides that have timing information.

The mean temperature (degrees F) of thaws over periods of 7, 30 and 365 days prior to landslides that have timing information.

The mean temperature (degrees F) of entire period over periods of 7, 30 and 365 days prior to landslides that have timing information.

The number of freeze thaw alterations over periods of 7, 30 and 365 days were counted prior to landslides that have timing information.

The mean length (days) of freeze thaw cycles over periods of 7, 30 and 365 days prior to landslides that have timing information.

The mean temperature (degrees F) of freezes over periods of 7, 30 and 365 days prior to landslides that have timing information.

The minimum temperature (degrees F) of freezes over periods of 7, 30 and 365 days prior to landslides that have timing information.

The mean temperature (degrees F) of thaws over periods of 7, 30 and 365 days prior to landslides that have timing information.

The mean temperature (degrees F) of entire period over periods of 7, 30 and 365 days prior to landslides that have timing information.