Proximal and ultimate influences on planform change in Bayou Pierre, Mississippi
Cite this dataset
Stearman, Loren (2024). Proximal and ultimate influences on planform change in Bayou Pierre, Mississippi [Dataset]. Dryad. https://doi.org/10.5061/dryad.dz08kps4v
Abstract
Channel incision and associated planform changes are widespread in the Gulf Coastal Plain of the United States, following a century of sediment reduction activities. These processes cause numerous threats to infrastructure, social, and ecological systems. While the drivers of channel incision are well documented for some watersheds in the region, others remain understudied. Bayou Pierre, Mississippi, is a small watershed draining directly to the lower Mississippi River with a contentious debate regarding origins of erosional processes. The watershed is home to one listed species (Bayou Darter Nothonotus rubrus) as well as a diverse aquatic biota assemblage. Determining potential drivers of erosional activities in the watershed will be critical for effective mitigation or restoration activities. This dataset compiles information from numerous other datasets to construct histories of Mississippi River architectural change, land use change, and climate patterns to construct a history of potential erosional influences in the watershed. It further contains original measurements of planform dynamics over a sixty year period, and analyses of how changes to planform relate to recent and historical influences. In addition, a secondary assessment of both accuracy and precision of measurement methods are included.
README: Proximal and Ultimate Influences on Planform Change in Bayou Pierre, Mississippi
https://doi.org/10.5061/dryad.dz08kps4v
Description
A dataset combining original river planform measurements with literature-derived and database-derived measurements to assess patterns of planform change on Bayou Pierre, Mississippi, and potential recent and older influences on erosional behavior in the watershed.
Principle Investigator Contact Information
Name: Loren Stearman
Institution: University of Southern Mississippi
Email: Loren.Stearman@usm.edu
Alternate Contact Information
Name: Franklin Heitmuller
Institution: University of Southern Mississippi
Email: Franklin.Heitmuller@usm.edu
Data Sources and Temporal Scopes
Land Use / LandCover Data
A. Digitized US Census Records: 1840 - 2012, decadal
B. FORE-SCE backcast land cover models: 1938 - 1992
C. FORE-SCE observed land cover data: 1993 - 2005
D. FORE-SCE forecast model scenario B1: 2006 - 2021
E. National Land Cover Database land cover data: 2001, 2004, 2006, 2008, 2011, 2013, 2016, 2019, 2021
Rainfall and Hydrology
A. NOAA Climate Data Online precipitation data: 1938 - 2022
B. USGS gage 07290650 stage and discharge data: 1961 - 2022, some data gaps from original source
C. USACE Mississippi River at Vicksburg stage data: 1903 - 2022
Aerial Imagery
A. USGS Single Frames: 1949, 1953, 1969
B. USGS Aerial Photo Mosaics: 1958
C. NHAP Aerial Imagery: 1982, 1985
D. NAPP Aerial Imagery: 1992, 1995
E. NAIP Aerial Imagery: 2004, 2005, 2006, 2007, 2009, 2010, 2012, 2014, 2016, 2018, 2020, 2021
Literature Derived Data
A. Fisk 1944 Mississippi River Historic Channel Maps: pre-1600 - 1940
B. Jaeger et al. 2010 knickpoint migration rates: variable, generally 20th century
Data Access/Extraction Dates
A. Data drawn directly from source (Land use/land cover, rainfall/hydrology, literature): 2022 - 2023
E. Extraction of geomorphic measurements from aerial imagery: 2019 - 2024
Data Spatial Scope
US Census records and NOAA climate data are derived from the five counties drained by Bayou Pierre watershed (Claiborne, Copiah, Hinds, Jefferson, and Lincoln counties, MS). USGS gage data was collected at the Willows Road gage (gage 07290650), capturing approximately 61% of the watershed upstream. All other land use/land cover data was extracted from the Bayou Pierre watershed. Aerial imagery derived measurements were taken from three approximately ten kilometer focal regions on the mainstem Bayou Pierre.
Fisk's 1944 maps describe the lower Mississippi River Valley. USACE Mississippi River stage measurements were taken at Vicksburg, MS. Knickpoint migration rate estimates from Jaeger et al. 2010 were subset to records in the American southeast.
Description of the data and file structure
File Descriptions
* Spatial file collections are denoted with an asterisk, and include extensions of .cpg, .dbf, .prj, .qmd, .shp, and .shx. All spatial files are in EPSG 6509, NAD83 (2011) Mississippi West.
These files contain unvegetated channel centerlines, remeasured for precision assessment years (1953, 1982, 2004, and 2021) for the three focal reaches.
These files contain unvegetated channel centerlines, split at 1km intervals, remeasured for precision assessment years (1953, 1982, 2004, and 2021) for the three focal reaches.
These files contain unvegetated channel cross-section width transects at 200m intervals, remeasured for precision assessment years (1953, 1982, 2004, and 2021) for the three focal reaches.
These files contain the ground control points (GCPs) derived from TIGER/line road intersections, within approximately 5km of the Bayou Pierre's main channel in the three focal reaches. These points are used in assessing precision of measurement in the main analysis.
These files contain the ground control points (GCPs) derived from NAIP aerial imagery road intersections, within approximately 5km of the Bayou Pierre's main channel in the three focal reaches. These points are used in assessing precision of measurement in the main analysis.
This file contains livestock and human population density for the five counties drained by Bayou Pierre, from 1840 to 2012, at decadal intervals.
This file contains proportion of agricultural land by US census categories for the five counties drained by Bayou Pierre, from 1840 to 2012, at decadal intervals.
This file contains county metadata from US census records for the five counties drained by Bayou Pierre.
This file contains raw 15-minute hydrologic data for the Bayou Pierre at Willows, MS, gage.
This file contains raw precipitation data at 15-minute intervals, from NOAA's Climate Data Online database, for the five counties drained by Bayou Pierre.
This file contains daily mean precipitation values calculated from /hydro/counties_rain.csv, organized by water year (first day of the year is October 1 of the preceeding calendar year).
This file contains the mean and maximum annual river stage values for the Mississippi River at Vicksburg, MS.
This file contains knickpoint migration rates extracted from Jaeger et al. 2010 for knickpoints in alluvial rivers in the southeastern United States.
This file contains river kilometers of estimated knickpoint position on mainstem Bayou Pierre, derived from aerial imagery, from 1958 to 2021.
This file contains forest stand age data from the Forest Inventory Analysis, derived from raster data from the FORE-SCE models.
This file contains a key to land use categories in both FORE-SCE and NLCD data. It is not used in analyses but provided to assist end users.
This file contains the estimated percent forested land cover from FORE-SCE predictions, within a 200m buffer of the channel, upstream of the base of each of the three focal regions.
This file contains the estimated percent forested land cover from FORE-SCE predictions, for the entire upstream catchment, upstream of the base of each of the three focal regions.
This file contains land cover classes from FORE-SCE predictions, from 1938 - 2005, for the entire Bayou Pierre watershed.
This file contains land cover classes from NLCD measurements, from 2001 - 2019, for the entire Bayou Pierre watershed.
This file contains estimates of the Mississippi River's position relative to the Bayou Pierre's exit at the bluff line, as well as estimated lengths of Bayou Pierre from the bluff line, based on Fisk's (1944) historic channel maps.
These files contain unvegetated channel centerlines for the three focal reaches for all years of aerial imagery data.
These files contain unvegetated channel centerlines for the three focal reaches, split into 1km segments for displacement analyses, for the three focal reaches for all years of aerial imagery data.
These files contain unvegetated channel width transects for all three reaches for all years of aerial imagery data.
This file contains flow direction data for NHD comid values (from and to segment identifiers) in Bayou Pierre.
This file contains value-added attributes for NHD comid values in Bayou Pierre.
This file contains estimates of dates of occurrence for events important to the geomorphic history of Bayou Pierre.
Methodology
Channel planform centerlines were manually digitized using QGIS and historical aerial imagery, separately for each of the three focal reaches. Channel centerlines were digitized following the unvegetated portion of the channel (i.e., inclusive of bars). Lengths (m) were calculated using QGIS after digitization. This dataset was a replicated effort to assess precision of the original effort, using years 1953, 1982, 2004, and 2021.
Channel planform centerlines digitized in QGIS were split into approximately 1km segments, using line features drawn perpendicular to the channel, at 1km intervals as determined by the common ruler developed in the manuscript. The common ruler is derived from the NHD+ flowlines and 1km segments represent 1km segments on these lines. Line features were split using the split lines with lines tool in QGIS. Features were then merged with an associated spatial feature dataset to add kilometer ID values. Individual segment lengths were then calculated using QGIS after digitization. This dataset was a replicated effort to assess precision of the original effort, using years 1953, 1982, 2004, and 2021.
Channel planform width transects were manually digitized perpendicular to the unvegetated channel, at 200m intervals as determined by the common ruler used in the manuscript. The common ruler is derived from the NHD+ flowlines and 200m points represent 200m points on these lines. Lengths (m) were calculated using QGIS after digitization. This dataset was a replicated effort to assess precision of the original effort, using years 1953, 1982, 2004, and 2021.
Road intersections as ground control points (gcp) were first identified by finding line intersections in QGIS using the 2021 TIGER/line road dataset for Copiah and Claiborne counties, Mississippi. Point features were created using the line intersection tool. Road intersections were filtered to be within 5km of the channel in the focal reaches. Road intersections were then filtered further by visual assessment, dropping intersections which could not be seen in accuracy assessment imagery years (1949, 1953, 1969, 1982, 1985, 1992, 1995, 2004 - 2007, 2012, 2016, 2021).
Road intersection gcp points for individual aerial imagery years were created starting with the TIGER/line gcp point layer. Road intersection points were moved until they matched the appropriate center point of the intersection as described by line features in the original TIGER/line 2021 road layer, but matching the spatial pattern visible in aerial imagery. This process was repeated for all accuracy assessment imagery years separately (1949, 1953, 1969, 1982, 1985, 1992, 1995, 2004 - 2007, 2012, 2016, 2021).
This dataset was extracted from Crossley (2020, DOI: https://doi.org/10.3886/E115795V3). Data were filtered to include only counties in Mississippi drained by Bayou Pierre (Claiborne, Copiah, Hinds, Jefferson, and Lincoln). Data were further filtered to include only agricultural animals and human population density.
This dataset was extracted from Crossley (2020, DOI: https://doi.org/10.3886/E115795V3). Data were filtered to include only counties in Mississippi drained by Bayou Pierre (Claiborne, Copiah, Hinds, Jefferson, and Lincoln). Data were further filtered to include only agricultural crop land cover proportions.
This dataset was extracted from Crossley (2020, DOI: https://doi.org/10.3886/E115795V3). Data were filtered to include only counties in Mississippi drained by Bayou Pierre (Claiborne, Copiah, Hinds, Jefferson, and Lincoln). Data were further revised to remove unnecessary columns.
This dataset was extracted from the USGS gage on Bayou Pierre at Willows, MS (gage 07290650, https://waterdata.usgs.gov/nwis/uv?site_no=07290650). Data were extracted as daily mean values from 1961-06-01 to 2022-02-28.
This dataset was extracted from the Climate Data Online data archive (https://www.ncei.noaa.gov/cdo-web/). Data were extracted for rainfall daily totals by county using the search tool to identify all stations within a county, for the five counties drained by Bayou Pierre (Claiborne, Copiah, Hinds, Jefferson, and Lincoln). Data were appended into a single file following download.
This file was created using the file /hydro/counties_rain.csv. Daily totals were calculated using mean values of all gages, separately by day. Data were assigned a water year with water year starting on October 1st of the year preceeding the calendar year (e.g., October 1st, 2023 is water year 2024). Data were then organized chronologically by water year date and by water year to create a date x year matrix.
This dataset was extracted from the US Army Corps of Engineers Rivergages database (https://rivergages.mvr.usace.army.mil/WaterControl/new/layout.cfm). Stage height (ft) were extracted at daily intervals. Mean and maximum values were calculated for each year, and data were converted from feet to meters.
This dataset was extracted from multiple publications, compiled primarily in Yodis (1991) and Jaeger et al. (2010). Observations were restricted to the southeastern United States, as Jaeger et al. (2010) found a dramatic difference in migration rates between the southeastern and southwestern United States. Observations were further restricted to those with a measured upstream drainage area, and a measured migration rate. Location, upstream area, and migration rate were extracted from data sources. For observations compiled in Yodis (1991), observations were converted from US imperial units to metric units. When the name of a stream on which a knickpoint occurred was not documented, we listed it as "unknown tributary".
This dataset was compiled by identifying knickpoint locations in aerial imagery, on mainstem Bayou Pierre. Knickpoint location was identified as the point on the channel where obvious channel expansion, including riparian forest loss, had begun. Stearman and Schaefer (2023) found widening to be an appropriate proxy for channel profile breaks in DEMs when aerial imagery and DEM data years matched. Longitudinal position along the river was determined as the closest 200m point from the common ruler used in the manuscript.
This dataset was extracted from the 2011 forest age raster associated with the FORE-SCE land cover dataset, and ultimately derived from the USDA Forest Inventory and Analysis database. The original raster was clipped to the watershed boundaries of Bayou Pierre, using NHD+ catchment area to delineate Bayou Pierre watershed boundaries.
This file was created by consulting metadata for both FORE-SCE land cover databases and NLCD land cover databases. Each potential land cover class was assigned its appropriate numeric designation in both databases, with a value of NA assigned if the land cover class was absent in a database.
This file was created using data from the FORE-SCE backcast estimates from 1938 - 1992, FORE-SCE observed values from 1993 - 2005, and scenario B1 FORE-SCE forecast values from 2006 - 2023, for land cover in Bayou Pierre. The downstream-most segment of each of the three focal reaches was identified, and a 200m buffer was created around all flowline features upstream from this point. This process was done separately by reach to create three different, progressively larger, 200m usptream buffer polygons. For each raster layer, we clipped raster values within each of these polygons. We then calculated the proportion of raster cells in land use classes 8, 8, 10, and 16 (deciduous forest, evergreen forest, mixed forest, and woody wetlands respectively), separately for each year.
This file was created using data from the FORE-SCE backcast estimates from 1938 - 1992, FORE-SCE observed values from 1993 - 2005, and scenario B1 FORE-SCE forecast values from 2006 - 2023, for land cover in Bayou Pierre. The downstream-most segment of each of the three focal reaches was identified, and an upstream catchment polygon was created by merging all NHD+ catchments upstream of this segment. This process was done separately by reach to create three different, progressively larger, upstream catchment polygons. For each raster layer, we clipped raster values within each of these polygons. We then calculated the proportion of raster cells in land use classes 8, 8, 10, and 16 (deciduous forest, evergreen forest, mixed forest, and woody wetlands respectively), separately for each year.
This file was created using data from the FORE-SCE backcast estimates from 1938 - 1992 and FORE-SCE observed values from 1993 - 2005. We used the NHD+ catchment layer to create a catchment polygon for the entire Bayou Pierre. We then clipped raster layers sequentially to this polygon and extracted land use class values. We used these values to calculate a year x class total count matrix, and then used this matrix to calculate a year x class proportion of coverage matrix.
This file was created using data from the NLCD land cover databases, from 2001 - 2019. We used the NHD+ catchment layer to create a catchment polygon for the entire Bayou Pierre. We then clipped raster layers sequentially to this polygon and extracted land use class values. We used these values to calculate a year x class total count matrix, and then used this matrix to calculate a year x class proportion of coverage matrix.
This file was created using Fisk's (1944) classic Mississippi River meander belt plates. We used ImageJ to calculate distance in pixels of a given Mississippi River historical channel to the exit of Bayou Pierre at the bluffline. We then used the calibration tool and the provided scales on the maps to convert this distance to miles. We further repeated the procedure using the path tool to measure length of Bayou Pierre from the bluffline, following the retreat of Mississippi River from the bluffline. We then converted these values to kilometers. When a channel for a given stage was not visible (typically overlain by other channels), we recorded a value of NA.
Channel planform centerlines were manually digitized using QGIS and historical aerial imagery, separately for each of the three focal reaches. Channel centerlines were digitized following the unvegetated portion of the channel (i.e., inclusive of bars). Lengths (m) were calculated using QGIS after digitization. This dataset includes all aerial imagery years.
Channel planform centerlines digitized in QGIS were split into approximately 1km segments, using line features drawn perpendicular to the channel, at 1km intervals as determined by the common ruler developed in the manuscript. The common ruler is derived from the NHD+ flowlines and 1km segments represent 1km segments on these lines. Line features were split using the split lines with lines tool in QGIS. Features were then merged with an associated spatial feature dataset to add kilometer ID values. This dataset includes all aerial imagery years.
Channel planform width transects were manually digitized perpendicular to the unvegetated channel, at 200m intervals as determined by the common ruler used in the manuscript. The common ruler is derived from the NHD+ flowlines and 200m points represent 200m points on these lines. Lengths (m) were calculated using QGIS after digitization. This dataset includes all aerial imagery years.
This dataset was extracted from the NHD+ V2 attributes dataset. The original plusflow.dbf table was subset to include only comid values present in Bayou Pierre for either from or to values. We then subset the data to include only the from and two columns.
This dataset was extracted from the NHD+ V2 attributes dataset. The original vaa.dbf table was subset to include only comid values present in Bayou Pierre.
This file was created by consulting numerous literature sources and data available in this Dryad repository, to determine date ranges for various critical events in Bayou Pierre's geomorphic history.
Data-Specific Information
Projection: EPSG 6509
Units: meters
Extent: 646919.1145290870917961,257769.1563901072368026 : 686204.0752577298553661,279455.6886708496604115
Geometry: Line (MultiLineString)
Feature Count: 12
Fields: (integer) Year: Numeric value of year
(character) Section: Focal reach (MSBP, UPBP, or HWBP)
(continuous) LengthM: The length of the feature in meters
Data Type: integer, character, continuous
Missing Data Value: NA
Projection: EPSG 6509
Units: meters
Extent: 646919.1145290870917961,257769.1563901072368026 : 686204.0752577298553661,279455.6886708496604115
Geometry: Line (MultiLineString)
Feature Count: 124
Fields: (integer) Year: Numeric value of year
(character) Section: Focal reach (MSBP, UPBP, or HWBP)
(continuous) LengthM: The length of the original feature, prior to splitting, in meters
(integer) KM: The river kilometer of the segment
(continous) PieceLengt: The length of the split feature, in meters
Data Type: integer, character, continuous
Missing Data Value: NA
Projection: EPSG 6509
Units: meters
Extent: 646908.7073222999460995,257781.4477349991502706 : 686176.2090207479195669,279481.2692171948147006
Geometry: Line (MultiLineString)
Feature Count: 511
Fields: (continuous) KM: The river kilometer of the transect
(integer) Year: Numeric value of year
(character) Section: Focal reach (MSBP, UPBP, or HWBP)
(continuous) LengthM: The length of the transect feature in meters
Data Type: integer, character, continuous
Missing Data Value: NA
Projection: EPSG 6509
Units: meters
Extent: 646070.4975895385723561,259471.8185090110637248 : 688567.7694227964384481,281705.3971937397145666
Geometry: Point (Point)
Feature Count: 40
Fields: (character) Reach: Focal reach (MSBP, UPBP, or HWBP)
(integer) ID: A numeric point ID value
(character) FID: An ID value incorporating the reach and the ID values
Data Type: integer, character
Missing Data Value: NA
Projection: EPSG 6509
Units: meters
Extent: 646069.9092445196583867,259466.5717753832286689 : 688579.7359530422836542,281714.4483993565663695
Geometry: Point (Point)
Feature Count: 506
Fields: (character) Reach: Focal reach (MSBP, UPBP, or HWBP)
(integer) ID: A numeric point ID value
(integer) Year: A numeric value for year
(character) FID: An ID value incorporating the reach and the ID values
Data Type: integer, character
Missing Data Value: NA
Number of variables: 9
Number of rows: 96
Variable list: (character) County: the county in Mississippi from which the data are drawn
(integer) Year: A numeric value for year
(continuous) Cat: The density of cattle, numbers per square kilometer
(continuous) Ckn: The density of chicken, numbers per square kilometer
(continuous) Hmn: Human population density, numbers per square kilometer
(continuous) Hrs: The density of horses, numbers per square kilometer
(continuous) Shp: The density of sheep, numbers per square kilometer
(continuous) Swn: The density of swine, numbers per square kilometer
(continuous) Tky: The density of turkeys, numbers per square kilometer
Data type: character, integer, and continuous, numbers per square kilometer
Missing data value: NA
Number of variables: 20
Number of rows: 96
Variable list: (character) County: the county in Mississippi from which the data are drawn
(integer) Year: A numeric value for year
(continuous) Bkw: proportion of land under cultivation for buckwheat
(continuous) Brl: proportion of land under cultivation for barley
(continuous) Crn: proportion of land under cultivation for corn
(continuous) Ctn: proportion of land under cultivation for cotton
(continuous) Flx: proportion of land under cultivation for flax
(continuous) Hay: proportion of land under cultivation for hay
(continuous) Oat: proportion of land under cultivation for oats
(continuous) Pls: proportion of land under cultivation for pulses
(continuous) Pnt: proportion of land under cultivation for peanuts
(continuous) Ptt: proportion of land under cultivation for potatoes
(continuous) Ric: proportion of land under cultivation for rice
(continuous) Rye: proportion of land under cultivation for rye
(continuous) Sgc: proportion of land under cultivation for sugar cane
(continuous) Sgm: proportion of land under cultivation for sorghum
(continuous) Soy: proportion of land under cultivation for soybeans
(continuous) Swt: proportion of land under cultivation for sweet potatoes
(continuous) Tbc: proportion of land under cultivation for tobacco
(continuous) Wht: proportion of land under cultivation for wheat
Data type: character, integer, and continuous
Missing data value: NA
Number of variables: 5
Number of rows: 6
Variable list: (character) STATE: The US State
(character) COUNTY: The county
(continuous) AREA_KM: The county area in square kilometers
(continuous) Lon: The longitude in decimal degrees
(continuous) Lat: The latitude in decimal degrees
Data type: character and continuous
Missing data value: NA
Number of variables: 5
Number of rows: 22189
Variable list: (character) Agency: the agency responsible for data collection
(integer) Site: The USGS site number
(date) Date: The date
(continuous) Gage Height: the mean daily gage height (above reference height) in feet
(continuous) Discharge: The mean daily discharge in cubic feet per second
Data type: character, integer, date, and continuous
Missing data value: NA
Number of variables: 8
Number of rows: 388789
Variable list: (integer) STATION: The weather station ID
(character) NAME: The weather station name
(continuous) LATITUDE: The latitude in decimal degrees
(continuous) LONGITUDE: The longitude in decimal degrees
(continuous) ELEVATION: The elevation in feet
(date) DATE: The date of data collection
(continuous) PRCP: Daily rainfall totals in inches
(character) PRCP_ATTRIBUTES: Attributes or data flags for precipitation values
Data type: character, integer, date, and continuous
Missing data value: NA
Comments: Please consult the Climate Data Online archives for detailed data flag descriptions
Number of variables: 133
Number of rows: 367
Variable list: (integer) Julian: The julian date. Note order is for water year, water year begins on October 1st of previous year
(continuous, cols 2 - 133) 1891 - 2023: Mean daily precipitation for the year indicated by the column
Data type: integer and continuous
Missing data value: NA
Number of variables: 3
Number of rows: 123
Variable list: (integer) Year: The year of data collection
(continuous) Stage_Mean: The mean annual river stage at Vicksburg, MS, meters
(continuous) Stage_Max: The maximum annual river stage at Vicksburg, MS, meters
Data type: character, integer and continuous
Missing data value: NA
Number of variables: 4
Number of rows: 32
Variable list: (character) Site: The locality of knickpoint migration observation
(character) Publication: The original publication from which the observations were derived, were often cited in Yodis (1991) or Jaeger (2010)
(continuous) Basin_Area_KM2: The basin area upstream of the knickpoint, in square kilometers
(continuous) Migration_Rate_m_Y: The upstream migration rate of the knickpoint in meters per year
Missing data value: NA
Number of variables: 3
Number of rows: 19
Variable list: (character) Source: The aerial imagery source for the observed knickpoint location on mainstem Bayou Pierre
(integer) Year: The year of observation
(RKM) The river kilometer, on the common ruler station system as described in text
Missing data value: NA
Number of rows: 267
Number of columns: 323
Number of bands: 1
Projection: EPSG 5072 (NAD83(NSRS2007)/ Conus Albers)
Extent: 451630.7591000000247732,965723.2218000000575557, 532488.9910999999847263,1032545.9727000000420958
Data type: continuous, age of forest stands in 2011
Missing data value: -9999
Number of variables: 3
Number of rows: 22
Variable list: (character) Category: The description of the land use category
(integer) FORE-SCE: The FORE-SCE numeric designation for the category
(integer) NLCD: The NLCD designation for the category
Data type: character and integer
Missing data value: NA
Number of variables: 4
Number of rows: 87
Variable list: (integer) Year: A numeric value for year
(continuous) MSBP: The proportion of forested land upstream of the MSBP focal reach within a 200m buffer of the stream
(continuous) UPBP: The proportion of forested land upstream of the UPBP focal reach within a 200m buffer of the stream
(continuous) HWBP: The proportion of forested land upstream of the HWBP focal reach within a 200m buffer of the stream
Data type: integer and continuous
Missing data value: NA
Number of variables: 4
Number of rows: 87
Variable list: (integer) Year: A numeric value for year
(continuous) MSBP: The proportion of forested land upstream of the MSBP focal reach
(continuous) UPBP: The proportion of forested land upstream of the UPBP focal reach
(continuous) HWBP: The proportion of forested land upstream of the HWBP focal reach
Data type: integer and continuous
Missing data value: NA
Number of variables: 17
Number of rows: 69
Variable list: (integer) Year: the year of data
(continuous, columns 2 - 17): 1 - 16: The proportion of land surface area in the entire Bayou Pierre watershed with the specified FORE-SCE land cover class
Data Type: integer and continuous
Missing data value: NA
Number of variables: 16
Number of rows: 9
Variable list: (integer) Year: the year of data
(continuous, columns 2 - 17): 11, 21-24, 31, 41-43, 52, 71, 81-82, 90, 95: The proportion of land surface area in the entire Bayou Pierre watershed with the specified NLCD land cover class
Data Type: integer and continuous
Missing data value: NA
Number of variables: 5
Number of rows: 28
Variable list: (character) Stage: The stage designated by Fisk (1944) for a given channel, or the year of observation if known, including 2010 aerial imagery observations
(integer) PlotX: A dummy variable to facilitate plotting in figures
(continuous) Distance (Miles): The shortest distance between the edge of the Mississippi River channel and the bluff line adjacent to the Bayou Pierre's exit, in miles
(continuous) Distance (km): The shortest distance between the edge of the Mississippi River channel and the bluff line adjacent to the Bayou Pierre's exit, in kilometers
(continuous) LBP_km: The length of the Bayou Pierre's channel outside of the bluff ine, in kilometers
Data type: character, integer, and continuous
Missing data value: NA
Projection: EPSG 6509
Units: meters
Extent: 646915.9960026873741299,257682.3165051191463135 : 686471.1996930547757074,279464.4432420416269451
Geometry: Line (MultiLineString)
Feature Count: 53
Fields: (integer) Year: Numeric value of year
(character) Section: Focal reach (MSBP, UPBP, or HWBP)
(continuous) LengthM: The length of the feature in meters
Data Type: integer, character, continuous
Missing Data Value: NA
Projection: EPSG 6509
Units: meters
Extent: 646915.9960026873741299,257767.9154604672221467 : 686197.9095669961534441,279464.4432420416269451
Geometry: Line (MultiLineString)
Feature Count: 549
Fields: (integer) Year: Numeric value of year
(character) Section: Focal reach (MSBP, UPBP, or HWBP)
(continuous) LengthM: The length of the original feature, prior to splitting, in meters
(integer) KM: The river kilometer of the segment
(continuous) PieceLengt: The length of the split feature, in meters
Data Type: integer, character, continuous
Missing Data Value: NA
Projection: EPSG 6509
Units: meters
Extent: 646907.3757803656626493,257761.1463759193720762 : 686196.1347752565052360,279489.8509977775975130
Geometry: Line (MultiLineString)
Feature Count: 2468
Fields: (continuous) KM: The river kilometer of the transect
(integer) Year: Numeric value of year
(character) Section: Focal reach (MSBP, UPBP, or HWBP)
(continuous) LengthM: The length of the transect feature in meters
Data Type: integer, character, continuous
Missing Data Value: NA
Number of variables: 4
Number of rows: 846
Variable list: (integer) FROMCOMID: A COMID value for a focal flowline segment
(integer) TOCOMID: The downstream COMID for the flowline segment
Data type: integer
Missing data value: 0 (This is inherited from the original NHD+ data specifications)
Number of variables: 6
Number of rows: 12
Variable list: (integer) ComID: The line feature segment comid identifier value from NHD+
(continuous) LengthKM: The length of the line segment in kilometers
(continuous) TotDASqKM: the total upstream drainage area, in square kilometers
Data type: integer and continuous
Missing data value: NA
Number of variables: 6
Number of rows: 12
Variable list: (character) Event: The event plotted in manuscript figure 5 (knickpoint migration figure)
(integer) Start1: The start time (x-axis values) for the solid line indicating date range in the figure
(integer) Stop1: The stop time (x-axis values) for the solid line indicating date range in the figure
(integer) Start2: The start time (x-axis values) for the dashed line indicating uncertainty in date range in the figure
(integer) Stop2: The stop time (x-axis values) for the dashed line indicating uncertainty in the date range in the figure
(integer) n: The plotting order (y-axis values) for the line segments
Data type: character and integer
Missing data value: NA
Sharing/Access information
Funding
This research was funded by a grant from the US Army Corps of Engineers, Engineering Research & Development Center (Contract W912HZ21C0064).
Sharing/Access
This work is licensed under a CC0 1.0 Universial (CC0 1.0) Public Domain Dedication license.
Associated R scripts are copyright under a GNU GPLv3 license.
Related Data Sources
USGS Earth Explorer. https://earthexplorer.usgs.gov/
Mississippi Automated Resource Information System (MARIS). https://maris.mississippi.edu/
Recommended Citation
Stearman, Loren W., and Franklin T. Heitmuller. 2024. Proximal and Ultimate Influences on Planform Change in Bayou Pierre, Mississippi.
Code/Software
watershed_influences_analysis.R
This file contains R functions and a workflow to replicate analyses of external influences on erosional processes in Bayou Pierre, Mississippi.
R Environment for Statistical Computing
R 4.1.0
mgcv
moments
plotrix
raster
robustbase
focal_reach_analyses.R
This file contains R functions and a workflow to replicate analyses of planform dynamics and influences on planform dynamics in Bayou Pierre since 1949.
R Environment for Statistical Computing
R 4.1.0
colorspace
emmeans
foreign
mgcv
moments
multcomp
raster
robustbase
sf
geospatial_accuracy_assessments.R
This file contains R functions and a workflow to assess the accuracy of aerial imagery georeferencing and the precision of planform measurement methodologies.
R Environment for Statistical Computing
R 4.1.0
mgcv
plotrix
sf
Methods
Historical and contemporary watershed influence data were collected via extraction from existing datasets. Mississippi River historic planform dynamics were derived from digitized plates from Fisk (1944) and aerial imagery from 2010. Land use data were extracted from digitized US Census records at the county level (1840 - 2012), the FORE-SCE backcast, observed, and scenario B1 forecast data (1938 - 2023), and National Land Cover Database data (2001 - 2019). Secondary analysis of land use data occurred to generate estimates of land use proportions in the watershed. Percent upstream forest was calculated at catchment and 200m buffer scales, for three focal reaches in the watershed, using FOR-SCE datasets. Total percent coverage of all land use categories for the entire watershed was also calculated using FORE-SCE datasets and NLCD datasets. Hydrologic data were extracted from the Climate Data Online database for counties drained by Bayou Pierre (1901 - 2022), the USGS gage at Willows, MS (1961 - 2022), and the Mississippi River at Vicksburg from the USACE Rivergages database (1901 - 2022). Forest stand age data were extracted from the USDA Forest Inventory Analysis dataset, provided as raster data with the FORE-SCE datasets. Accessory datasets were extracted from the NHD+ attributes dataset to facilitate analyses.
Planform analysis was conducted on aerial imagery for three approximately 10km focal reaches in Bayou Pierre, Mississippi. Aerial imagery included single-frame photographs (1949, 1953, 1969), NAPP imagery (1982, 1985), NHAP imagery (1992, 1995), and NAIP imagery (2004 - 2007, 2009, 2010, 2012, 2014, 2016, 2018, 2020, 2021). All imagery prior to 1995 was georeferenced in QGIS. We manually digitized unvegetated channel centerlines for entire reaches for analysis of channel length and channel perpendicular cross-section features at 200m intervals for analysis of channel width. We further created channel centerline features split at 1km intervals for analysis of channel displacement.
As an assessment of accuracy, we digitized TIGER/line road intersection ground control points in a subset of aerial imagery years. We further repeated channel measurements for 1953, 1982, 2004, and 2021 independently, to assess precision of methodology. We provide analytical scripts to conduct all analyses and accuracy assessments pertinent to a related manuscript.
Funding
United States Army Corps of Engineers, Award: W912HZ21C0064