In this research, an implementation of artificial deep neural networks (ANN) over outputs of 24-h multi-domain high-resolution nested real case Weather Research and Forecasting (WRF) model runs was carried out over two high-resolution simulation domains, which are tested and compared for rainfall generation in order to assess the signal fading event observed on geostationary telecommunication spacecraft in orbit for a real multiscale storm case. Our methodology of ANN, which is driven by WRF model output parameters, focuses on prediction of the rain attenuation signal impairment which is observed on the communication satellite telemetry (TM) downlink signal levels under significant aerosol presence due to dust storm which occurred on 12 September 2020. This modelling approach is then compared to rain attenuation observed on TM signal and correlated with communication satellite ground station TM signal measurements. Preliminary results from conducted error analysis (RMSE) on multiple input single output feed-forward neural network (MISO FFNN) prediction model outputs tested with several neural algorithms indicate good correlation with the TM downlink signal attenuation observations taken from the ground station TM baseband demodulator.

中文翻译：

---

气溶胶效应下使用混合数值天气预报和人工神经网络方法估计电信卫星遥测信号信号衰落的初步结果

在这项研究中，人工深度神经网络 (ANN) 在 24 小时多域高分辨率嵌套真实案例天气研究和预报 (WRF) 模型运行的输出上的实施在两个高分辨率模拟域上进行，其中对降雨产生进行了测试和比较，以评估在轨道上的地球静止通信航天器上观察到的信号衰落事件，以应对真实的多尺度风暴情况。我们的人工神经网络方法由 WRF 模型输出参数驱动，专注于预测雨衰信号损伤，该损伤是在 9 月 12 日发生的沙尘暴造成的显着气溶胶存在下，在通信卫星遥测 (TM) 下行链路信号水平上观察到的2020 年。然后将该建模方法与在 TM 信号上观察到的雨衰进行比较，并与通信卫星地面站 TM 信号测量值相关联。使用多种神经算法测试的多输入单输出前馈神经网络 (MISO FFNN) 预测模型输出的传导误差分析 (RMSE) 的初步结果表明，与从地面站 TM 基带解调器获取的 TM 下行链路信号衰减观测值具有良好的相关性。