AbstractAccurate meteorological forecasts from the surface to troposphere layers are crucial for dust storm predictions, as even small uncertainties in meteorological conditions can influence the transportation of dust particles, thereby significantly affecting dust storm forecasts. Typically, a greater quantity and higher quality of meteorological observations result in more accurate meteorological outcomes. However, meteorological stations, especially the stations which monitor tropospheric meteorological variables, are sparsely distributed and may not be sufficient for high‐quality meteorological forecasts. To address this shortfall, this study investigates the sensitive areas for target observation to enhance meteorological forecasts for dust storm events that struck the Beijing–Tianjin–Hebei (BTH) area from 2021 to 2023, using the Conditional Nonlinear Optimal Perturbation (CNOP) method, which fully considers the impact of nonlinearity. For comparison, the First Singular Vector (FSV) method, which is widely used in operational target observation field campaigns, is also employed to identify the sensitive areas. Results show that although the sensitive areas identified by the two methods are both distributed in the northwest direction of the BTH region, the FSV‐based sensitive areas are much closer to the BTH region. By conducting observing system experiments for each dust storm event, we verified numerically and explained physically the advantages of CNOP in determining the sensitive areas in target observation. The result highlights the importance of considering nonlinearity when identifying the sensitive areas for target observation and may provide a theoretical foundation for establishing upper‐air radiosonde sites or planning practical field observation campaigns.
Read full abstract