ArcGIS API for Python

Dev Summit 2021: Automating coastline extraction

As the pace of coastal change accelerates, there is naturally a greater need for efficient and timely acquisition of coastline information.

At the Developer Summit 2021 plenary, Rhea Jackson demonstrated how you can use the ArcGIS API for Python and the ArcGIS Notebook Server to automate coastline extraction, which can facilitate the analysis and protection of these dynamic and fragile landforms.

Watch the plenary video below, and then read the rest of the blog for a summary of the processes that Rhea explored in her demo.

Search and visualize satellite imagery

First, Rhea used Sentinelsat, a Python package, to search for Sentinel-2 level-1C satellite images in the Copernicus Open Access Hub (SciHub). Sentinelsat provides a Python API and a command-line interface that simplifies the searching and downloading of all Sentinel products from SciHub.

Rhea’s code parsed a query to the Sentinelsat API to search for Sentinal-2 level-1C satellite images in SciHub.
Rhea’s code parsed a query to the Sentinelsat API to search for Sentinal-2 level-1C satellite images in SciHub.

She then transposed the index and columns of the resulting data frame using the Pandas Python package and visualized the footprint on a map widget provided by the ArcGIS API for Python.

Visualization of the 'footprint' on a map widget.
Visualization of the 'footprint' on a map widget.

Download and process the data

Next, Rhea used Sentinelsat to download the satellite images and retrieve their metadata.

Rhea's code leveraged Sentinelsat to download the satellite images.
Rhea's code leveraged Sentinelsat to download the satellite images.

She then used the create_image_collection method to create an image layer from the downloaded images. Note that she specified ‘Sentinel-2’ as the raster type.

Rhea used the create_image_collection method to create an image layer
'Sentinel-2' was specified as the raster type during the creation of an image layer.

Immediately after, she performed three pre-processing steps—extracting the near-infrared (NIR) band, the blue band, and the red band to delineate certain surfaces—before running the raster calculation function to apply the band ratio technique, which is a remote sensing technique for coastline extraction. She later converted the binary raster to a polyline feature during the post-processing of the data.

The application of the band ratio technique (left) generated a binary raster.
The application of the band ratio technique (left) generated a binary raster.
The binary raster was later converted to a polyline feature for coastline extraction.
The binary raster was later converted to a polyline feature for coastline extraction.

Automate the process

After successfully extracting the coastlines, Rhea used the ArcGIS Notebook Server’s task scheduling feature to configure a task that automatically executes the ArcGIS Notebook every Monday at noon.

ArcGIS Notebook Server’s task scheduling feature was used to configure a task that automatically executes the Notebook.
ArcGIS Notebook Server’s task scheduling feature was used to configure a task that automatically executes the Notebook.

Finally, she showed the results obtained after three weeks of automated execution of the notebook.

Visualizing the results side-by-side
The results accurately showed the changes in tidal height.

Try it for yourself

Rhea’s demo showed how a time-consuming and labor-intensive task such as coastline extraction can be automated with the powerful remote sensing capabilities of the ArcGIS API for Python and the ArcGIS Notebook Server’s task scheduling feature. Visit the ArcGIS API for Python and the ArcGIS Notebook Server pages to learn more about how you can automate the execution of your notebooks to facilitate your GIS work.

About the authors

Aawaj is a product engineer on the ArcGIS Enterprise team.

Rhea Jackson

Software Development Engineer on Esri's raster team.

Next Article

Harnessing the Power of Imagery: A Programmatic Approach

Read this article