Search

Low visibility event, as a disastrous weather, has great impacts on traffic and transportation, aircraft, and people's daily life, etc. Timely and accurate forecasts of low visibility events are urgently needed and meaningful. The reasonable algorithm of atmospheric extinction in atmospheric chemistry models is the basis for quantitatively predicting low visibility. The revised IMPROVE algorithm (RIMP) of atmospheric extinction is incorporated into the chemistry-weather interacted model GRAPES_Meso5.1/CUACE CW V1 to improve the prediction of low visibility events (LVEs) in the urban agglomerations in eastern China, which is compared with the original IMPROVE algorithm (OIMP) used in this model. The study results show that the RIMP effectively reduces the overestimation of low visibility prediction by OIMP in general, leading to a decrease of root-mean-square errors (RMSEs) and an increase of Threat Score (TS) of visibility <3 km, 5 km, and 10 km overall both at regional and city scales in varying degrees due to its more detailed processing of aerosols' size, optical feature and hygroscopic growth; The improvements of visibility prediction of LVEs by RIMP depends on the combined contribution of high relative humidity (RH) and PM 2.5 instead of single high RH or PM 2.5. The relative contributions of RH and PM 2.5 concentration on different levels of low visibility are different in Beijing-Tianjin-Hebei (BTH) and Yangtze River Delta (YRD) regions due to their different RH and PM 2.5 , which leads to the different improvement of RIMP in the two regions. The larger improvements by RIMP occur for visibility <5 km in BTH, while in YRD, the larger improvements by RIMP occur for visibility <10 km and >5 km. Moreover, the improvements by RIMP were more evident with higher RH conditions in both regions. The uncertainty created by the extinction algorithm is one important factor of the multiple factors affecting LVEs prediction; accurate modeling of high RH near saturation is also very important for LVEs prediction. • The revised IMPROVE (RIMP) algorithm is applicated to the atmospheric chemistry model to improve low visibility prediction. • RIMP generally improves low visibility prediction than the original IMPROVE algorithm (OIMP). • The improvements in low visibility prediction by RIMP vary depending on different aerosol and humidity conditions. [ABSTRACT FROM AUTHOR]

Compared with climate models, the role of aerosol-cloud interaction (ACI) in mesoscale numerical weather prediction (NWP) models still needs to be better understood, especially in haze regions with relatively high aerosol concentration. Here, we perform two sensitivity experiments with and without ACI (ACI and NO-ACI) in the atmospheric chemistry model GRAPES_Meso5.1/CUACE to investigate the impact of ACI on mesoscale NWP during the low-cloud period in winter 2016 over varying haze regions (severe polluted Jing-Jin-Ji (JJJ), polluted Yangtze River Delta (YRD), and weak polluted Pearl River Delta (PRD)) in China. The study results show that the real-time ACI improves underestimated cloud optical thickness (COT) and cloud water liquid path (CLWP) in haze regions, with the mean bias of simulated COT (CLWP) decreased by 27% (3%), 60% (14%), and 55% (3%) in JJJ, YRD, and PRD, respectively. The increased COT and CLWP lead to a decrease of 6.8, 21, and 13 W m−2 in daytime surface downward shortwave radiation (SDSR) in JJJ, YRD, and PRD, helping to reduce the mean bias of daytime SDSR by 6%, 13%, and 9%. In addition, ACI mitigates the warm bias of temperature at 2 m and dry bias of relative humidity (RH) at 2 m to a certain extent in haze regions, particularly in YRD with the mean absolute bias improved by 13% and 6%. The simulated vertical structure of temperature and RH in the ACI experiment is more consistent with observations than in the NO-ACI experiment. Further investigations find that the ACI effects on mesoscale NWP strongly depend on COT and CLWP magnitude over varying haze regions. Higher COT and CLWP, hence more significant meteorology changes due to ACI, occur in YRD, followed by PRD and JJJ. This study demonstrates the importance and complexity of ACI in modifying mesoscale NWP over varying haze regions of China, which promotes the further understanding of ACI in operational NWP models and bridges the gap with climate models. •ACI increases COT and CLWP when low-cloud and haze coexist. • ACI further leads to decreased SDSR, lower temperature, and higher RH. • ACI-induced changes improve mesoscale NWP to a certain extent. • The ACI effects on mesoscale NWP strongly depend on COT and CLWP magnitude. [ABSTRACT FROM AUTHOR]