Global Preconditioned DEM Workflow

Overview

This document outlines the process for creating a global preconditioned Digital Elevation Model (DEM) using ALOS data from Google Earth Engine (GEE). The workflow includes:

1. Extracting the DEM from GEE
2. Downloading the DEM tiles
3. Stitching the DEM into a global dataset
4. Parallelized pit-filling using HydroBASINS subwatersheds
5. Multi-Flow Direction (MFD) routing on the processed DEM

Steps

1. Extract ALOS DEM from Google Earth Engine

A script is needed to retrieve ALOS DEM data from GEE. The script should:

• Access ALOS DEM from GEE.
• Define the global extent of the DEM.
• Export the DEM in manageable tiles.

2. Download DEM Tiles

Once exported, a process is required to:

• Copy the tiles from GEE to local storage or cloud storage.
• Maintain metadata for merging and processing.

3. Stitch the DEM into a Global Dataset

To create a seamless global DEM:

• Merge the downloaded tiles.
• Ensure proper alignment and resolution consistency.
• Handle any missing data or artifacts.

4. Parallelized Pit-Filling using HydroBASINS Subwatersheds

Pit-filling is required to remove sinks and depressions. This can be parallelized using HydroBASINS subwatersheds:

• Divide the global DEM into subwatershed regions.
• Assign each region to a processing unit.
• Apply pit-filling algorithms (e.g., Wang & Liu or FillDepressions).
• Merge the processed outputs while ensuring hydrological consistency.

5. Multi-Flow Direction (MFD) Routing

Once pit-filled, the DEM should be hydrologically conditioned using MFD:

• Compute flow directions using MFD.
• Generate flow accumulation layers.
• Verify consistency across boundaries.

Next Steps

• Implement the GEE script to extract ALOS DEM.
• Develop a download automation script.
• Identify a tool or library to merge tiles.
• Determine best practices for parallel pit-filling.
• Set up MFD routing and validation.