Enhancing Forecasting Capabilities Through Data Assimilation: Investigating the Core Role of WRF 4D-Var in Multidimensional Meteorological Fields

As climate change intensifies, enhancing numerical weather prediction (NWP) accuracy has been increasingly critical. While data assimilation optimizes NWP initial conditions, its effectiveness over complex terrain requires further systematic evaluation. This study implemented a high-resolution WRF/4D-Var data assimilation framework, overcoming its inherent limitation of not supporting two-layer nested assimilation across domains by designing a two-layer nested “assimilation-forecast” workflow. Representative winter and summer cases from February and June 2019 were selected to evaluate improvements in near-surface and upper-air meteorological parameters. The results indicated that the 4D-Var data assimilation significantly improved the correlation coefficients of near-surface variables during winter by 2.9% (temperature), 14.5% (relative humidity), 6.6% (wind speed), and 10.4% (wind direction), with even greater improvements observed in summer reaching 13.3%, 5.8%, 35.3%, and 42.3%, respectively. Meanwhile, 4D-Var considerably enhanced the atmospheric vertical profiling, with the middle troposphere (300–700 hPa) exhibiting the most pronounced improvement. Among different surface types, water bodies exhibited the strongest assimilation response. Results also revealed systematic corrections to the background fields, with February exhibiting more uniform adjustments in contrast to June’s complex spatiotemporal patterns. Positive effects persisted throughout the 24-h forecasts, with the maximum benefit occurring within the first 12 h. These results demonstrate the effectiveness of 4D-Var in regional meteorological forecasting, highlighting its value for constructing high-precision multidimensional meteorological fields to support both weather and air quality simulations.