A Topology-Guided GeoAI Agentic Framework for Urban Flood Screening

Nov 3, 2026·
Yutian Tao
Haoan Feng
Haoan Feng
,
Xin Xu
,
Leila De Floriani
· 2 min read
Abstract
Open-world geographic artificial intelligence (GeoAI) agents can transform public geospatial data into accessible decision support for real-world challenges, but must balance autonomous reasoning with reliable spatial analysis and human oversight. We present a topology-guided GeoAI agentic framework for urban flood screening that takes a place name, drawn region, or uploaded boundary and returns ranked candidate basins with an interactive map and report. A deterministic controller retrieves public elevation, rainfall, soil, land-cover, exposure, and infrastructure layers, extracts terrain depressions with discrete Morse theory, and ranks basins with a relative flood-priority score. The language model acts only at recorded decision points: selecting a persistence threshold from a precomputed candidate table, drafting recommendations from computed evidence, and reviewing each recommendation against the same evidence. Three human-in-the-loop checkpoints let users revise the threshold, factor weights, or recommendation. In a College Park, Maryland case study, the selected threshold reduced 278 raw terrain minima to 70 basins, with 77% of these basins’ lowest points also identified as sinks by a separately computed eight-direction (D8) flow-routing model. Across four further study areas, agreement ranged from 61% to 83%. By combining open data, explicit spatial reasoning, traceable evidence, and human oversight, the framework enables transparent first-pass flood screening while keeping consequential recommendations inspectable and user-controlled.
Type
Publication
In The 1st ACM SIGSPATIAL Student Challenge on Open Agents with Spatial Intelligence for Social Good (OASIS 2026), to appear

Accepted at OASIS 2026, the ACM SIGSPATIAL Student Challenge on Open Agents with Spatial Intelligence for Social Good. Our team is a top-10 finalist and was selected for an on-site presentation and demo in Riverside, CA, on November 3, 2026.

Our team received a Google DeepMind Student Travel Grant to support participation.

Five-stage workflow showing data acquisition, topological analysis, ranking and validation, recommendation and reflection, and interactive outputs. Purple arrows mark three human-in-the-loop checkpoints.

Figure 1. Workflow of the framework with intermediate results. Deterministic stages retrieve and align data, extract and rank Morse basins, and compare them with D8 routing and with re-ranking under alternative weights. Separate language-model calls select the persistence threshold, draft the recommendation, review it against the recorded evidence, and answer later user questions. Purple arrows mark the three human-in-the-loop (HITL) checkpoints.

The framework starts from a place name or boundary, retrieves public geospatial layers, extracts terrain basins using discrete Morse theory, and ranks them for preliminary flood screening. Language-model calls select a persistence threshold, draft recommendations, and review them against the computed evidence. Three human checkpoints allow users to revise the threshold, ranking weights, or recommendation.

The evaluation covers five study areas. In College Park, persistence-based simplification reduced 278 raw terrain minima to 70 candidate basins; 77% of the retained basin minima also coincided with D8 sinks. Agreement ranged from 61% to 83% in the four additional areas. The highest-priority College Park basin remained first under all four tested weighting schemes. These are checks of consistency and ranking sensitivity, rather than validation against observed flood events; the scores support preliminary screening and are not calibrated flood probabilities.

The preprint is available on arXiv, posted October 22, 2026. The accepted OASIS 2026 paper has the assigned ACM DOI 10.1145/3849739.3856761.

GeoAI Spatial Intelligence Topological Data Analysis Human-in-the-Loop Urban Flood Screening
Haoan Feng
Authors
Haoan Feng
Ph.D. Candidate in Computer Science