CAUS as a Data-Quality Layer for IoT Flood Monitoring in Uganda: Design and Preliminary Simulation

Abstract

This article examines the Context Aggregation, Validation and Unification System (CAUS) as a data-quality layer in an Internet of Things (IoT) flood monitoring architecture for data-scarce settings in Uganda. Guided by design-science research and ITIL 4 service-management principles, it synthesises a review of IoT flood frameworks, CHIRPS rainfall analysis for Bwaise, Kalerwe and Mbarara (2010–2024), available Kampala-area ERA5 Ag comparisons, and a cloud-hosted proof of concept using synthetic sensor observations. The rainfall dataset contained 5,479 daily records per location. Bwaise and Kalerwe had identical extracted CHIRPS series; Mbarara recorded a lower mean daily rainfall (2.58 mm) but a higher maximum (53.49 mm). Daily CHIRPS–ERA5 Ag correlations were approximately 0.40–0.43 and monthly correlations 0.75–0.77 for the available Kampala-area comparison. The recommends a meythodological approach to reproducible CAUS thresholds, fusion weights, imputation settings or a confidence equation. This article therefore distinguishes reported implementation from a proposed, parameterised specification covering validation, source-aware fusion, short-gap imputation, calibration metadata and a weighted quality index. The simulation demonstrated workflow execution, not field accuracy or calibrated confidence. CAUS is presented as a design and testing blueprint; controlled fault-injection, independent sensor calibration and field validation are required before operational warning use.