基于LSTM的空气质量二次预报方法

LSTM-based Secondary Prediction Method for Air Quality

WRF(Weather Research and Forecasting Model)-CMAQ(Community Multiscale Air Quality)模型作为一种通用多尺度预测模型,被广泛应用于空气中一次污染物的预测。然而,诸如臭氧等二次污染物是由一次污染物等通过化学或光化学反应生成的。WRF-CMAQ内部机制的不完善,导致了其在二次污染物浓度方面的预测精度受到了极大的限制。为了解决该问题,我们提出了一种基于LSTM的混合预测模型。具体来说,在WRFCMAQ模型预报的基础上,使用LSTM网络对污染物进行二次预测。基于三个监测点的数据预测证明了所提出的混合模型的有效性,并成功预测了三日后的各类污染物浓度以及AQI指标。

As a general multi-scale prediction model,WRF(Weather Research and Forecasting Model)-CMAQ(Community Multiscale Air Quality)model has been widely used to predict primary pollutants in the air.However,secondary pollutants such as ozone are generated by chemical or photochemical reactions of primary pollutants.Due to the imperfect internal mechanism of WRF-CMAQ,the prediction accuracy of its secondary pollutant concentration is greatly limited.To solve this problem,we propose a hybrid prediction model based on LSTM.Specifically,on the basis of WRFCMAQ model prediction,we use LSTM network to make a second prediction of pollutants.The prediction based on the data of three monitoring points proves the effectiveness of the proposed mixed model,and successfully predicts the concentrations of various pollutants and AQI indicators three days later.