Watershed Marching Delineation Algorithm
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing algorithms for watershed delineation from digital flow direction grids are inefficient, particularly in terms of computational complexity and storage requirements, especially for large watersheds, as they scale with the number of grid cells rather than the number of vertices, leading to significant processing time and resource challenges.
Innovation Solution
The Haag Shokoufandeh Marching (HSM) algorithm uses a modified nested set data structure to efficiently retrieve watershed boundaries by traversing the flow direction grid, recording discovery and finish times, and employing a marching algorithm that identifies boundary lattice points with a linear increase in storage costs, allowing for local determination of global watershed identity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional watershed delineation algorithms are used, then watershed boundaries can be identified, but computational complexity and storage requirements increase significantly with the number of grid cells
Solution Approach 1:
The algorithm segments the watershed delineation process into three distinct phases: (1) identifying stream cells using a modified nested set data structure, (2) extracting boundary lattice points through a marching algorithm, and (3) constructing the final boundary. This segmentation allows each phase to operate efficiently with linear complexity O(N) relative to the number of vertices N, rather than scaling with the total number of grid cells.
Solution Approach 2:
The algorithm transitions from operating on the grid cell dimension to operating on the vertex dimension by identifying boundary lattice points that define the watershed boundary. This dimensional shift allows the computational complexity to scale with the number of vertices rather than the number of grid cells, significantly improving efficiency for large watersheds.
2Reliability
If traditional watershed delineation algorithms are used, then complete watershed boundaries are obtained, but storage requirements increase with the number of grid cells
Solution Approach 1:
The algorithm extracts only the essential boundary lattice points that define the watershed boundary, rather than storing or processing all grid cells. By using a marching algorithm to identify and connect boundary points, the storage requirements scale linearly with the number of vertices rather than the number of grid cells, while maintaining complete and accurate boundary representation.
Solution Approach 2:
The algorithm applies different processing strategies to different parts of the watershed: stream cells are identified using nested set properties, boundary cells are identified through local gradient analysis, and interior cells are inferred. This local quality approach ensures accuracy where needed while reducing overall storage requirements.
3Reliability
If traditional watershed delineation algorithms are used, then watershed boundaries are delineated, but processing time increases significantly for large watersheds
Solution Approach 1:
The algorithm performs preliminary identification of stream cells and boundary cells before constructing the final watershed boundary. By pre-identifying these key cells using efficient nested set properties and local gradient analysis, the algorithm reduces the computational burden of the final boundary construction phase, achieving linear time complexity O(N) relative to the number of vertices.
Solution Approach 2:
The marching algorithm continuously traces the watershed boundary by systematically moving from one boundary lattice point to the next, maintaining continuous progress around the entire perimeter. This continuous action ensures complete boundary delineation without requiring repeated processing or backtracking, optimizing processing time.
Data Source
AI summary
The constrained watershed boundary (CWB), defined as a polygon containing all the flow direction grid cells with a surface flow distance less than a user prescribed threshold uses an algorithm that builds upon the HSM algorithm proposed and augments the data structure with a flow distance grid calculated directly from the original flow direction grid.