Design an efficient data driven decision support system to predict flooding by analysing heterogeneous and multiple...

Area of Science:

• Environmental Science
• Data Science
• Hydrology

Background:

• Floods are a prevalent global natural disaster, significantly impacting human lives and economies.
• Existing flood prediction models often fail to incorporate all heterogeneous flood-causing factors due to data complexity.
• Effective flood forecasting is crucial for mitigating disaster impacts and socioeconomic losses.

Purpose of the Study:

• To address the challenge of handling heterogeneous data sources for flood prediction.
• To propose a novel big data architecture, Data Lake, for ingesting and storing diverse flood-related data.
• To enhance the accuracy and reliability of flood forecasting and warning systems.

Main Methods:

• Development of a Data Lake architecture to manage heterogeneous flood-causing data.
• Utilization of inferential statistical approaches to analyze the relevance of flood-causing factors.
• Machine learning model creation using raw, diverse data ingested into the Data Lake.

Main Results:

• The Data Lake architecture successfully ingests and stores heterogeneous flood data in its raw format.
• Inferential statistics identified key flood-causing factors and their statistical relevance to prediction outcomes.
• A foundation for creating more accurate and comprehensive flood prediction models was established.

Conclusions:

• The proposed Data Lake architecture provides a viable solution for managing diverse data in flood prediction.
• Accurate flood prediction models, informed by statistically relevant factors, can significantly improve early warning systems.
• Implementing such systems can lead to reduced socioeconomic losses and enhanced public safety during flood events.