Marine Geospatial Ecology Tool

[ShareUser]

Hello,

I'm working on my masters thesis on identification of potential fishing zone. I need to use Cayula-Cornillon's SIED algorithm for SST front detection. I've identified the front but I want to show frontal lines as bold black lines.

1) Can you kindly guide me how to do this

2) Also for chlorophyll front detection,  is it best to us Cany algorithm? 

3) I've used GRHSST L4 data for SST front. How to combine the Daily data to monthly data for specific one year(2023)?

Because I intend to overlay monthly SST fronts and chlorophyll fronts.

[KevinButler-Analysis]

There are multiple ways to convert the output of edge detection to lines. This one has worked for me in the past. This is the output of my edge detection filter (float pixel type)

Run the Iso Cluster Unsupervised Classification tool to convert the float raster to distinct classes (integer pixel type). You may have to experiment with the Number of classes and Sample Interval value to get the level of detail you want. This is the output.

I’m interested in the front represented by the red pixels (value = 3). Reclassify the raster from above to retain only the red pixels (convert all other values to NoData).

This is the result:

Run the Raster to Polyline tool to convert the front(s) to polylines. Post-process the lines with Extend Line or Smooth Line tools if necessary.

[KevinButler-Analysis]

2) Also for chlorophyll front detection,  is it best to us Cany algorithm?  Edge detection can be challenging. It is as much an art as it is a science. Both algorithms have strengths and weaknesses, but the quality of the result will depend on the strength of the fronts and the pixel size. You may find this resource helpful: Chang, Y., & Cornillon, P. (2015). A comparison of satellite-derived sea surface temperature fronts using
two edge detection algorithms. Deep Sea Research Part II: Topical Studies in Oceanography, 119, 40-47.
doi: 10.1016/j.dsr2.2013.12.001
Available at: https://doi.org/10.1016/j.dsr2.2013.12.001

3) I've used GRHSST L4 data for SST front. How to combine the Daily data to monthly data for specific one year(2023)? GRHSST L4 data are available in netCDF format. Follow this workflow Create a multidimensional mosaic dataset to represent the daily netCDF files as a mosaic and then use Aggregate Multidimensional Raster tool to aggregate to a monthly dataset.