Best Practices for Analyzing Global Forest Change (Hansen Dataset) in ArcGIS Pro (Basic License)

[AdrianLain]

Hi everyone,

I'm doing some research on a project to analyze Global Forest Change first at country level, and then using that output also to fit different smaller areas. As an example, I was trying over Indonesia using the Hansen dataset in ArcGIS Pro. I'm using the basic license, so I don't have access to advanced analysis tools.

My current approach involves:

1. Mosaicking: Creating a mosaic of the relevant raster images.
2. Clipping: Clipping the mosaic to the extent of Indonesia.
3. Conversion: Converting the clipped raster to a shapefile.
4. Calculation: Adding a field to calculate forest cover and loss areas.

However, this process is extremely time-consuming.

I'm looking for suggestions on how to improve efficiency. Here are some ideas I'm considering:

• Smaller Areas of Interest (AOIs): Processing smaller AOIs and merging the results.
• Python Automation: Using Python scripts to automate the process (Suggestions of materials to learn appreciated).

Do you have any experience with similar projects? What techniques have you found to be most effective?

I'm eager to learn from the community and find the best approach for my analysis.

Thanks, Adrian

P.S. No GIS expert here, but eager to learn.
Computer Hardware is quite recent, so it should not be a limitation.

• 12 Core 5GHz
• 64 GB RAM DDR5
• 8 GB VRAM 4060
• SSD ~4.000MB/s

[BobBooth1 · Accepted solution]

You might experiment with automating the workflow with ModelBuilder.

https://pro.arcgis.com/en/pro-app/latest/help/analysis/geoprocessing/modelbuilder/modelbuilder-quick...

If you can build out the workflow from one country, you can make that model a tool and run it via another model that iterates over each country.

https://pro.arcgis.com/en/pro-app/latest/help/analysis/geoprocessing/modelbuilder/working-with-itera...

You can also automate workflows with Python. Here is an example tutorial:

https://learn.arcgis.com/en/projects/automate-a-geoprocessing-workflow-with-python/

There are a bunch of Python tutorial resources:

https://learn.arcgis.com/en/gallery/#?q=python

I would suggest looking at this series, to get familiar with Python in ArcGIS Pro. It starts very basic and works up:

https://learn.arcgis.com/en/paths/learn-python-in-arcgis-pro/

For change detection, you might look at this one:

https://learngis.maps.arcgis.com/home/item.html?id=24e13f2f4fe8453b9d1107e1f5b1d3cf

Best,

Bob

[AdrianLain]

Thanks!

Do you know if ModelBuilder works faster/ more efficient than using GUI of geoprocessing tools?

[BobBooth1]

The tools each take the same amount of time to run, but if you build a model containing a set of tools to automate a process, you only need to do that part once, then you can run the tool multiple times on different inputs to get different outputs. You can also specify that this process have input and output parameters, making it, in effect, a new tool. This tool could then be included in another model.

[clt_cabq]

another advantage of using ModelBuilder (or a python script) is that you can schedule it to run at a particular time; if your processes take a long time you can let it run overnight for instance.