Land/Sea Mask with GSHHS shapefile

[tontschman]

Hello all,

I have a problem regarding using the GSHHS_f_L1.shp shapefile to landmask data (obtained from https://www.ngdc.noaa.gov/mgg/shorelines/ ). I am working with Sentinel-1 data for arctic regions, hence I like to use the f L1 version. However, no matter what do, the shapefile does not line up with the coasts in the image. Neither with the raw-file or the terrain-corrected file to UTM or WGS84.

I use it as an via Vector->Import->ESRI shapefile.
Since it doesnt line up, using the Land/Sea Mask operator doesnt mask the landmasses properly. Am I missing something here?

My goal is to extend the the fairly simple S1 processing graph below, with the Land/Sea-Mask operator and the GSSHS shapefile.

Read
Thermal-Noise
Calibration
Speckle-Filter
Terrain-Correction (UTM Automatic)
Subset (via polygon)
Write

Am am using SNAP 10 and Sentinel-1 GRD EW data from ASF.

[MBG]

You would try raster land mask generated using GSHHG dataset, version 2.3.7, released on June 15, 2017: GitHub - AlexeyPechnikov/gmtlandmask: GMT 1-arc-second landmask for 1° tiles

[Marco_EOM]

There is also the Coastal Map I created. You could try it. Data is freely available. Coastal Map – Global High-Resolution Coastal Areas Map (zenodo.org)

Easy access from SNAP to the data is integrated in the EOMasters Toolbox Pro. A free trial is available.

Pro version of the EOMasters Toolbox released

EOMasters Toolbox Pro provides different tools to ease the daily work. Asset Library - Manage the resource you often use and have them always at hand SpeX - Execute more than 240 spectral indices easily Coastal Map Indicators - Coastal indicators to helping to analyse costal data The features are explained in more detail in the following blog: Introduction of the EOMasters Toolbox Pro and launch discount License key required & trial period You need to purchase a license key from the EOMaste…

[tontschman]

I am using this exact data as a .shp file. Why would your tiled raster landmask give me a different result?

[MBG]

Obviously, if you cannot align the vector mask, you can try using raster masks.

[tontschman]

That rasterdata still didn’t resolve this offset. The problem is GSHHG itself. The data does not appear to be accurate for Svalbard.
Using open-street-maps data resolved the issue.

[MBG]

@tontschman

Using open-street-maps data resolved the issue.

FYI, there’s no need to find and download OSM dumps. You can simply use the Python osmnx package to download OSM features, like this:

gdf = osmnx.features_from_polygon(geometry, tags=...)