Point Flood Query Based on Fast Binary Merge Tree

Abstract

In recent years, flood-risk analysis has played an important role in the evaluation of submergence and flooding simulation. In this paper, we study a flood point query problem: Given a terrain T and total volume of rainfall in the area, determine if a given point is flooded. Existing methods build a contour tree by exacting all data points from T, and then get a merge tree, which is complicated with increasing resolution of terrains. Given the volume of rain, this paper proposes a flood analysis algorithm based on a fast binary merge tree generation. By eliminating the invalid saddle vertices in the data, the algorithm directly establishes the merger tree according to the corresponding contour hierarchy. Besides, the area and volume are also attached to enrich the merge tree. We describe a suite of experimental results showing the performance of our algorithm in practice and the running time of the preprocessing step is greatly reduced.