Pradeepthi et al. (2026) Application of Machine Learning in Rainfall Disaggregation and Flood Inundation Mapping: A Case Study

Identification

• Journal: Lecture notes in civil engineering
• Year: 2026
• Date: 2026-01-01
• Authors: N. Pradeepthi, Akkera Hinduja, K. Vaishnavi, K. Veerendra Gopi, R. Vinay Ashok, K. Navya, M. Rohith Reddy
• DOI: 10.1007/978-981-95-3775-4_15

Research Groups

• Department of Civil Engineering, VNRVJIET, Hyderabad, India
• Department of Civil Engineering, PCCOE, Pune, India

Short Summary

This study applies Gaussian Process Regression for temporal rainfall disaggregation and integrates hydrological (SWMM) and hydraulic (HEC-RAS) models to map flood-inundated regions in Zone XV of Hyderabad city, India, addressing urban flood challenges.

Objective

• To identify and delineate flood inundated regions in Zone XV of Hyderabad city.
• To apply machine learning for temporal rainfall disaggregation and integrate it with hydrological and hydraulic modeling for urban flood inundation mapping.

Study Configuration

• Spatial Scale: Zone XV of Hyderabad city, India.
• Temporal Scale: Sub-hourly rainfall data disaggregated to 15-minute intervals.

Methodology and Data

• Models used:
  • Gaussian Process Regression (GPR) for rainfall disaggregation.
  • Stormwater Management Model (SWMM) for runoff estimation.
  • 2D–HEC RAS model for flood inundation delineation.
• Data sources:
  • Sub-hourly rainfall data.
  • Information on the existing drainage system (from GHMC authorities).

Main Results

• Sub-hourly rainfall data was successfully disaggregated to 15-minute intervals using Gaussian Process Regression.
• Intensity Duration Frequency (IDF) curves were developed from the disaggregated rainfall data.
• Runoff generated in the existing storm drainage system was estimated using the SWMM model.
• Flood inundated areas in Zone XV of Hyderabad city were delineated by integrating SWMM runoff output into the 2D–HEC RAS model.

Contributions

• Demonstrates a comprehensive methodology combining machine learning (GPR) for rainfall disaggregation with coupled hydrological-hydraulic modeling (SWMM-HEC-RAS) for urban flood inundation mapping.
• Provides a practical case study application in a densely populated Indian urban area, offering insights into managing recurring urban flood phenomena.
• Highlights the utility of disaggregated rainfall data for improving the accuracy of urban flood modeling and risk assessment.

Funding

• Not explicitly mentioned in the provided text.

Citation

@article{Pradeepthi2026Application,
  author = {Pradeepthi, N. and Hinduja, Akkera and Vaishnavi, K. and Gopi, K. Veerendra and Ashok, R. Vinay and Navya, K. and Reddy, M. Rohith},
  title = {Application of Machine Learning in Rainfall Disaggregation and Flood Inundation Mapping: A Case Study},
  journal = {Lecture notes in civil engineering},
  year = {2026},
  doi = {10.1007/978-981-95-3775-4_15},
  url = {https://doi.org/10.1007/978-981-95-3775-4_15}
}