Reliability-qualified Nighttime Lights for Disaster Impact and Recovery in cloud-impacted tropical Regions

Abstract. Daily satellite-derived nighttime lights (NTL) are increasingly used to monitor electricity disruption and recovery, but their application in humid tropical regions is constrained by persistent cloud cover and irregular observation. This study addresses this limitation by treating observability as a defining condition of analysis rather than a secondary constraint. Using the VIIRS Black Marble VNP46A2 product over the Samar–Leyte sub-grid in the Philippines, we quantify how spatial completeness, temporal gaps, and radiance variability shape the availability and interpretability of daily NTL. Observability diagnostics are developed across pixel, settlement, and regional scales, showing that valid observations are intermittent, seasonally structured, and spatially heterogeneous. A valid-pixel threshold (τ ) is introduced to regulate observational support, while settlement-based masks derived from GHSL SMOD isolate stable, signal-dominant areas. Exploring these factors jointly reveals how different configurations affect signal continuity and robustness. Alignment with hourly electricity load from the National Grid Corporation of the Philippines (NGCP) is used as an external consistency check. Results show that meaningful correspondence emerges only under constrained conditions defined by sufficient observational support and urban-dominant signal extraction. Optimal configurations vary across neighbouring island grids, but consistently occur at low to mid-range thresholds and within dense settlement classes. These findings highlight a fundamental constraint: in tropical regions, periods of greatest disruption often coincide with the lowest observational availability. By formalising observability and embedding it into the workflow, this study provides a reproducible framework for identifying when daily NTL can be reliably used for disaster monitoring and recovery assessment.