Evolution from the physical process-based approaches to machine learning approaches to predicting urban floods: a literature review
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Urban flooding has become a growing concern for many cities due to accelerating urbanisation, changing weather, and drainage system aging. Earlier studies of floods have taken primarily the traditional process-based approach to predicting urban floods, offering limited exploration of recent advancements in AI-driven, real-time, and community-integrated approach, which this paper brings into focus. This paper reviews how flood prediction has improved over the last two decades. It begins by reviewing physical process-based models (PPBMs), which often could not handle the fast changes in cities. New tools like geographic information systems (GIS), light detection and ranging (LiDAR), and satellite images helped improve flood mapping and planning. A big shift came with the use of AI and machine learning. They have made predictions faster, smarter, and more accurately. They allow many types of data, like weather information, sensor data, and social media (crowdsourcing) data. Recently, new tools like Internet of Things devices, deep learning, and hybrid models have brought even more progress. However, there are still challenges. Many cities still do not have the data, sensors, or systems needed to use these tools. Many models work on their own, not linked with city planning or community efforts. Flood solutions must now be more than just technical. Future systems should combine AI, hydrodynamics, GIS, and real-time monitoring, adapt to city change, and include input from communities. Open-source tools, public education, and better planning are also needed to make cities safer and more resilient to costly floods.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.003 | 0.000 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.002 | 0.000 |
| Bibliometrics | 0.000 | 0.002 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.000 | 0.002 |
| Insufficient payload (model declined to judge) | 0.000 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it