Global coastal transect type classification associated to "Global distribution and dynamics of muddy coasts", Nat. Coms 2023

This dataset is associated with the scientific article "Global distribution and dynamics of muddy coasts"  (https://www.nature.com/articles/s41467-023-43819-6)

The dataset contains the following items:

VisualizeResults.qgz: QGIS project with the formatted datasets ready for viewing (tested in QGIS v3.28.2.)

PredictedClasses_MuddyBeaches.shp (and associated files): Shapefile containing predicted classes of coastal transects differentiating the following types:

0: Sandy coastal transect

1: Muddy coastal transect

2: Rocky coastal transect

3: Vegetated coastal transect

4: Other undefined type

MuddyZonesFleming.shp (and associated files): Shapefile containing hand-drawn polygons based on the description of muddy coastline concentrations by (Flemming, B. W. Geographic distribution of muddy coasts, 2002).

TransectTrainingDataframe.csv : comma separated value file containing 1867 hand-labelled coastal type transects randomly distributed around the world. Each transect contains the following entries:

transect_id: individual transect ID

box_id: tiled ID

country_name

continent

New_Start_lon: longitude transect start

New_Start_lat: latitude transect start

New_End_lon: longitude transect end

New_End_lat: latitude transect end

Center_lon

Center_lat

Length: transect length

label: hand-labeled coastal type, Str (Sandy, Muddy, Rocky, Vegetated, Other)

abs lat: latitude

flag_sandy: bool, classified as a sandy area (Luiijendijk et al, 2018)

sand: % of pixels classified as sand from Sentinel 2 Multispectral land-type classification

mud: % of pixels classified as muddy from Sentinel 2 Multispectral land-type classification

water: % of pixels classified as water from Sentinel 2 Multispectral land-type classification

vegetation: % of pixels classified as vegetation from Sentinel 2 Multispectral land-type classification

dry: % of pixels classified as dry-vegetation from Sentinel 2 Multispectral land-type classification

other: % of pixels classified as other from Sentinel 2 Multispectral land-type classification

turbid % of pixels classified as turbid-water from Sentinel 2 Multispectral land-type classification

height max: max height in transect (MERIT-DEM)

height var: variance height in transect (MERIT-DEM)

mangrove: bool, mangrove presence (Global Mangrove Forest Distributions)

intertidal: intertidal range Murray et al 2019.

gsw: Global Surface Water dataset, Pekel et al 2016

maxtemp: max temperature WorldClim V1 Bioclim dataset

mintemp: min temperature WorldClim V1 Bioclim dataset

mhhw: Tide MHHW Deltares GTSM model

mllw: Tide MLLW Deltares GTSM model

tidal range: Tidal range Deltares GTSM model

PixelTrainingDataframe.csv : comma separated value file containing 3240 hand-labelled land-type pixel randomly distributed around the world. These are derived from a Sentinel 2 TOA images as a 2020 composite. Bands (BX) refer to the Sentinel-2 MSI Multispectral Instrument, level-2A spectral bands names. Each entry contains:

Aerosols: B1

Blue: B2

Green: B3

Red: B4

Red Edge 1: B5

Red Edge 2: B6

Red Edge 3: B7

NIR: B8

Red Edge 4: B8A

SWIR 1: B11

SWIR 2: B12

NDWI: Normalized Differential Water Index ((B3-B8)/(B3+B8)

NDVI ((B8-B4)/(B8+B4)

num: hand-label type differentiating the following types (1) sandy beaches, (2) mudflats, (3) clear water, (4) 111 turbid (brown) water, (5) green vegetation, (6) dry vegetation, and (7) other (containing clouds, snow, buildings, and all other indefinable points)