Research on the usability of different machine learning methods in visibility forecasting

Haze pollution, mainly characterized by low visibility, is one of the main environmental problems currently faced by China. Accurate haze forecasts facilitate the implementation of preventive measures to control the emission of air pollutants and, thereby mitigate haze pollution. However, it is not easy to accurately predict the low visibility events induced by haze, which requires not only accurate prediction for weather elements, but also refined and real-time updated source emission inventory. In order to obtain reliable forecasting tools, this paper studies the usability of several popular machine learning methods, such as support vector machine, k-nearest neighbor, random forest, as well as several deep learning methods, on the visibility forecasting. Starting from the main factors related to visibility, the relationships between wind speed, wind direction, temperature, humidity, and visibility are discussed. Training and forecasting were performed using the machine learning methods. The accuracy of these methods in visibility forecasting was confirmed through several parameters (i.e., root-mean-square error, mean absolute error, and mean absolute percentage error). The results show that: (1) Among all meteorological parameters, wind speed was the best at reflecting the visibility change patterns; (2) RNN LSTM, and GRU methods performs almost equally well on short-term visibility forecasts(i.e. 1h, 3h, and 6h); (3) A classical machine learning method (i.e. the SVM) performs well in mid- and long-term visibility forecasts; (4) The machine learning methods also have a certain degree of forecast accuracy even for long time periods (e.g. of 72h).

Operational implementation of the smoke forecasting capability in the RAP/HRRR numerical weather prediction system

<p>Since December, 2020 NOAA’s operational Rapid Refresh and High-Resolution Rapid Refresh (RAP/HRRR) numerical weather prediction modeling systems include smoke forecasting capability. In RAP/HRRR-Smoke primary aerosols (smoke) emissions from wildland fires are simulated by ingesting the fire radiative power data from the VIIRS (onboard S-NPP and NOAA-20) and MODIS (Terra and Aqua) satellite instruments in real time. I will describe the development and applications of the RAP and HRRR-Smoke models, which cover 3 domains – North America (at 13.5 km spatial gridding), CONUS and Alaska (3km resolution). I will present the applications of these models to forecast smoke distributions on regional and continental scales, and how adding the smoke direct feedback capability can improve weather and visibility forecasting. The RAP/HRRR-Smoke models are the first operational weather models in the US, which include the impact of the smoke aerosols on weather and visibility forecasting. The verification of the HRRR-Smoke model for July-August 2018 over the CONUS domain using various meteorological and aerosol measurements will be presented. For verification of the fire plume injection height simulations in HRRR-Smoke, we use the aircraft lidar and in-situ measurements from the FIREX-AQ campaign during August 6-8, 2019. Finally, I will discuss the future plans for improving forecasting of smoke-weather interactions in coupled air quality models.</p>