Objective This paper reviews data quality issues in integrated applications of power-system operational data and meteorological data, and analyzes how quality defects propagate to forecasting, dispatching, state estimation, reserve assessment, and secure operation.

Method The article started from the perspectives of the power side's supervisory control and data acquisition (SCADA) system, phasor measurement unit (PMU), advanced measurement system (AMI), and business data, summarizing quality enhancement methods such as bad data detection and consistency verification, missing data repair (interpolation, low-rank/graph models, and deep generation), bias correction, and probabilistic error modeling (distribution fitting, quantile regression, and ensemble forecasting). On the meteorological side, it summarized the quality assurance/quality control (QA/QC) process for observational and numerical weather prediction (NWP) data, post-processing corrections (model output statistics (MOS), quantile mapping), and uncertainty characterization (ensemble forecasting, calibration, and intervals). Furthermore, it discussed the fusion modeling and integrated quality management framework supported by cross-source spatiotemporal registration, feature and metadata governance.

Result The survey reveals persistent gaps in unified quality labels and evaluation metrics across data sources, cross-source spatiotemporal and semantic consistency, robustness to concept drift and extreme events, and reproducible benchmarks, protocols, and open datasets for end-to-end assessment.

Conclusion Future work should tightly couple grid physical constraints, probabilistic uncertainty, and online closed-loop governance to build an end-to-end pipeline from acquisition to application, providing a reliable data foundation for resilient power grids under high renewable penetration and extreme weather.