Search

The atmosphere exhibits variability across different time scales. Currently, in the field of atmospheric science, statistical filtering is one of the most widely used methods for extracting signals on certain time scales. However, signal extraction based on traditional statistical filters may be sensitive to missing data points, which are particularly common in meteorological data. To address this issue, this study applies a new type of temporal filters based on a one-dimensional convolution neural network (1D-CNN) and examines its performance on reducing such uncertainties. As an example, we investigate the advantages of a 1D-CNN bandpass filter in extracting quasi-biweekly-to-intraseasonal signals (10–60 days) from temperature data provided by the Hong Kong Observatory. The results show that the 1D-CNN achieves accuracies similar to a 121-point Lanczos filter. In addition, the 1D-CNN filter allows a maximum of 10 missing data points within the 60-point window length, while keeping its accuracy higher than 80% (R2 > 0.8). This indicates that the 1D-CNN model works well even when missing data points exist in the time series. This study highlights another potential for applying machine learning algorithms in atmospheric and climate research, which will be useful for future research involving incomplete time series and real-time filtering. [ABSTRACT FROM AUTHOR]

Initialization and the spin‐up effect are critical to model performance in the four‐dimensional variational data assimilation (4DVar) system. The incremental analysis update (IAU) technique is able to combat these problems. This study introduces the IAU technique into the 4DVar framework (IAU‐4DVar) of the operational China Meteorological Administration Global Forecast System (CMA‐GFS) and evaluates the performance of the IAU‐4DVar. In this IAU‐4DVar framework, the initial increments optimally estimated are split up into equally weighted forcing terms and integrated into the tangent model and its adjoint model. Then, the optimal estimation of the analysis increment is gradually introduced into the model integration during the assimilation window. A set of two‐month assimilation and forecast experiments were established and executed in this study, and the preliminary results show that the IAU technique can reduce the spin‐up effect of the initial analysis increment from the operational 4DVar system with digital filter initialization (DFI‐4DVar) by retaining the balance in dynamics and thermodynamics. In the optimal minimization process, the IAU‐4DVar converges faster than the DFI‐4DVar with a reduction of the computational cost of up to ca 40%. In addition, the introduction of IAU‐4DVar into the CMA‐GFS model could have a positive impact on the analysis and forecast skills relative to the operational DFI‐4DVar. [ABSTRACT FROM AUTHOR]

Tropical cyclone (TC) genesis forecasting is essential for daily operational practices during the typhoon season. The updated version of the Tropical Regional Atmosphere Model for the South China Sea (CMA-TRAMS) offers forecasters reliable numerical weather prediction (NWP) products with improved configurations and fine resolution. While traditional evaluation of typhoon forecasts has focused on track and intensity, the increasing accuracy of TC genesis forecasts calls for more comprehensive evaluation methods to assess the reliability of these predictions. This study aims to evaluate the effectiveness of the CMA-TRAMS for cyclogenesis forecasts over the western North Pacific and South China Sea. Based on previous research and typhoon observation data over five years, a set of localized, objective criteria has been proposed. The analysis results indicate that the CMA-TRAMS demonstrated superiority in cyclogenesis forecasts, predicting 6 out of 22 TCs with a forecast lead time of up to 144 h. Additionally, over 80% of the total could be predicted 72 h in advance. The model also showed an average TC genesis position error of 218.3 km, comparable to the track errors of operational models according to the annual evaluation. The study also briefly investigated the forecast of Noul (2011). The forecast field of the CMA-TRAMS depicted thermal and dynamical conditions that could trigger typhoon genesis, consistent with the analysis field. The 96-hour forecast field of the CMA-TRAMS displayed a relatively organized threedimensional structure of the typhoon. These results can enhance understanding of the mechanism behind typhoon genesis, fine-tune model configurations and dynamical frameworks, and provide reliable forecasts for forecasters. [ABSTRACT FROM AUTHOR]