Scribd Logo
Upload

Welcome to Scribd!

  • 4 views

    Flash Flood Forecasting Presentation

    Uploaded by

    Nabendu Das

    Copyright:

    © All Rights Reserved
    Download as PPTX, PDF, TXT or read online from Scribd

    Flash Flood Forecasting Presentation

    Uploaded by

    Nabendu Das
    4 views10 pages

    Original Title

    Flash_Flood_Forecasting_Presentation

    Copyright

    © © All Rights Reserved

    Available Formats

    PPTX, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    Copyright:

    © All Rights Reserved
    Download as PPTX, PDF, TXT or read online from Scribd
    Download as pptx, pdf, or txt
    4 views10 pages

    Flash Flood Forecasting Presentation

    Uploaded by

    Nabendu Das

    Copyright:

    © All Rights Reserved
    Download as PPTX, PDF, TXT or read online from Scribd
    Download as pptx, pdf, or txt
    of 10

    Use of Geostationary

    Meteorological Satellite Images


    in Convective Rain Estimation for
    Flash-Flood Forecasting
    Authors: T. Wardah, S.H. Abu Bakar,
    A. Bardossy, M. Maznorizan
    Institutions: Universiti Teknologi
    MARA, Stuttgart University,
    Malaysian Meteorological
    Department
    Introduction
    • • Problem: Frequent flash floods cause severe
    damage in Malaysia’s Klang River Basin.
    • • Current Limitations: Existing flood
    forecasting systems lack sufficient lead time
    for effective warnings.
    • • Objective: Develop a rainfall estimation
    algorithm using satellite data to improve flash-
    flood forecasting accuracy.
    Convective Rain and Flash Floods
    • • Convective Rain: Rapidly developing
    rainstorms with significant vertical cloud
    development.
    • • Impact on Flooding: Convective rains are the
    main cause of flash floods in tropical areas.
    • • Forecasting Need: Early detection is crucial
    to minimize the impact on lives and property.
    Study Area – Klang River Basin
    • • Location: West coast of Peninsular Malaysia,
    includes Kuala Lumpur.
    • • Geography: Mixed terrain, with highland
    areas and flat urban regions.
    • • Flood-Prone Areas: High urban development
    increases runoff, making certain areas highly
    susceptible to flash flooding.
    Data Sources and Technology
    • • Satellite Images: Infrared data from the
    Geostationary Meteorological Satellite-5
    (GMS-5).
    • • Radar Data: Reflectivity and display data
    from local Doppler radar stations.
    • • Supplementary Data: Numerical Weather
    Prediction (NWP) from NCEP, and rain and
    streamflow data from local gauges.
    Methodology – Image Processing
    and Estimation Techniques
    • • Pixel Geolocation: Mapping each pixel of
    satellite images to specific geographic
    locations.
    • • Cloud-Top Temperature: Using infrared
    brightness temperature to infer cloud height
    and potential for rainfall.
    • • Regression Analysis: Establishing a
    relationship between cloud-top temperature
    and radar rain rates.
    ANN Model for Rainfall Estimation
    • • Model Structure: Multi-layer back-
    propagation neural network (ANN).
    • • Input Variables: Cloud-top temperature,
    surrounding temperature variance,
    temperature change rate, and NWP data.
    • • Training and Validation: Model trained on
    radar data for accuracy, achieving a correlation
    coefficient of 0.91.
    Application in Flash-Flood
    Forecasting
    • • Coupling with Rainfall-Runoff Model:
    Satellite-based rainfall estimates feed into a
    runoff model.
    • • Tracking Storm Movement: Cross-correlation
    technique applied to satellite images.
    • • Case Example: June 10, 2003 flash flood
    event successfully predicted.
    Results and Validation
    • • Accuracy Metrics:
    • - Hourly estimation accuracy: MAE of 3.8,
    RMSE of 5.6.
    • - Total event rainfall accuracy: Correlation
    coefficient of 0.91.
    • • Forecast Lead Time: Additional lead time of
    2 hours achieved.
    Conclusion and Future Directions
    • • Summary: Satellite-based ANN model
    effectively improves flash-flood forecasting.
    • • Benefits: Extra lead time allows better
    preparation for flash-flood events.
    • • Future Work: Extend model applicability,
    enhance accuracy, and integrate with real-
    time forecasting systems.
    • • Acknowledgments: Thanks to Universiti
    Teknologi MARA, Malaysian Meteorological
    Department, and others.

    You might also like