Improving Solar Data Quality in Rio Grande do Norte Stations Through Nighttime Anomaly Detection Tests
Abstract
This work presents a data validation strategy to improve the quality of solar radiation measurements from stations located in Rio Grande do Norte, Brazil. We focus on nighttime anomaly detection using physical limit and rare extreme tests. A real-world case in the Pau dos Ferros station is discussed, where ants damaged the datalogger input channel, introducing a permanent offset that corrupted both nighttime and daytime readings. By placing the physical and statistical anomaly tests at the beginning of the validation workflow, the proposed methodology enabled early fault detection and prevented its influence on downstream analyses. The study also introduces a modification to the original test structure, improving its sensitivity to rare events. During the presentation, we briefly outline the validation workflow and highlight how this adjustment enhances detection efficiency. Overall, this procedure improves the consistency and reliability of solar monitoring networks, providing a stronger foundation for solar energy modeling, forecasting, and performance assessment in Rio Grande do Norte and other regions.
Keywords
solar radiation, data quality control, solar resource assessment, measurement validation