Search

A three-dimensional variational (3DVAR) lidar data assimilation method is developed based on the Community Radiative Transfer Model (CRTM) and Weather Research and Forecasting model coupled to Chemistry (WRF-Chem) model. A 3DVAR data assimilation (DA) system using lidar extinction coefficient observation data is established, and variables from the Model for Simulating Aerosol Interactions and Chemistry (MOSAIC) mechanism of the WRF-Chem model are employed. Hourly lidar extinction coefficient data from 12:00 to 18:00 UTC on March 13, 2018 at four stations in Beijing are assimilated into the initial field of the WRF-Chem model; subsequently, a 24 h PM2.5 concentration forecast is made. Results indicate that assimilating lidar data can effectively improve the subsequent forecast. PM2.5 forecasts without using lidar DA are remarkably underestimated, particularly during heavy haze periods; in contrast, forecasts of PM2.5 concentrations with lidar DA are closer to observations, the model low bias is evidently reduced, and the vertical distribution of the PM2.5 concentration in Beijing is distinctly improved from the surface to 1200 m. Of the five aerosol species, improvements of NO3− are the most significant. The correlation coefficient between PM2.5 concentration forecasts with lidar DA and observations at 12 stations in Beijing is increased by 0.45, and the corresponding average RMSE is decreased by 25 μg·m−3, which respectively compared to those without DA.

The performances of Beijing Lightning Network (BLNET) operated in Beijing-Tianjin-Hebei urban cluster area have been evaluated in terms of detection efficiency and relative location accuracy. A self-reference method has been used to show the detection efficiency of BLNET, for which fast antenna waveforms have been manually examined. Based on the fast antenna verification, the average detection efficiency of BLNET is 97.4% for intracloud (IC) flashes, 73.9% for cloud-to-ground (CG) flashes and 93.2% for the total flashes. Result suggests the CG detection of regional dense network is highly precise when the thunderstorm passes over the network; however it changes day to day when the thunderstorms are outside the network. Further, the CG stroke data from three different lightning location networks across Beijing are compared. The relative detection efficiency of World Wide Lightning Location Network (WWLLN) and Chinese Meteorology Administration - Lightning Detection Network (CMA-LDN, also known as ADTD) are approximately 12.4% (16.8%) and 36.5% (49.4%), respectively, comparing with fast antenna (BLNET). The location of BLNET is in middle, while WWLLN and CMA-LDN average locations are southeast and northwest, respectively. Finally, the IC pulses and CG return stroke pulses have been compared with the S-band Doppler radar. This type of study is useful to know the approximate situation in a region and improve the performance of lightning location networks in the absence of ground truth. Two lightning flashes occurred on tower in the coverage of BLNET show that the horizontal location error was 52.9 m and 250 m, respectively.

Beijing Lightning Network (BLNET), a regional multi-frequency-band lightning detection and location network, has been developed for both research and operational purposes. The network consisted of 15 stations in 2014, covering an area of about 100 km × 85 km in Beijing area. In this paper, we present the configuration of the network and the detailed description on the sensors for the first time. The Chan algorithm and Levenberg–Marquardt method are adopted jointly in the lightning location algorithm. The Monte Carlo simulation reveals that the horizontal and the vertical location errors within BLNET coverage are less than 300 m and 600 m, respectively. The comparison of total lightning location results with the corresponding radar reflectivity for a thunderstorm case indicates that BLNET has a good capability to track thunderstorms in the Beijing area. Besides, we examine the electric field waveforms of the preliminary breakdown process (PBP) of 111 − CG flashes, 92 + CG flashes, 22 normal-polarity IC flashes and 22 inverted-polarity IC flashes. Statistics on the time interval between the largest PBP pulse and the corresponding first RS pulse for CG flash ( PB–RS interval), the ratio of the largest PBP pulse to the corresponding first RS pulse for CG flash ( PB / RS ratio), the pulse width for individual pulse in the PBP train ( T w ) and the time interval between successive pulses in the PBP train ( T i ) are presented. The 3D location result shows that the PBP of a − CG flash initiates at an altitude of ~ 6 km and propagates downward to ~ 3 km with a vertical speed of 5.9 × 10 5 m/s. The initial 10 ms of an IC flash is found to start at ~ 7.2 km and develop upward to ~ 10 km. The vertical propagation speed is estimated to be 4.8 × 10 5 m/s during the first 6 ms. A PBP with relatively clear branched 3D structure is presented and discussed in the end of the paper. The average 3D propagation speed of two branches is estimated to be 2.1 × 10 6 m/s and 2.6 × 10 6 m/s, respectively.