A full-coverage estimation of PM2.5 concentrations using a hybrid XGBoost-WD model and WRF-simulated meteorological fields in the Yangtze River Delta Urban Agglomeration, China.

In spite of the state-of-the-art performances of machine learning in the PM2.5 estimation, the high-value PM2.5 underestimation and non-random aerosol optical depth (AOD) missing are still huge obstacles. By incorporating wavelet decomposition (WD) into the extreme gradient boosting (XGBoost), a hybrid XGBoost-WD model was established to obtain the full-coverage PM2.5 estimation at 3-km spatial resolution in the Yangtze River Delta Urban Agglomeration (YRDUA). In this study, 3-km-resolution meteorological fields simulated by WRF along with AOD derived from Moderate Resolution Imaging Spectroradiometer (MODIS) were served as explanatory variables. Model MW and Model NW were developed using XGBoost-WD for the areas with and without AOD respectively to obtain a full-coverage PM2.5 mapping in the YRDUA. The XGBoost-WD model showed good performances in estimating PM2.5 with R2 of 0.80 in the Model MW and 0.87 in the Model NW. Moreover, the K-value of Model MW increased from 0.77 to 0.79 and that of Model NM increased from 0.81 to 0.86 compared with the model without the step of WD, indicating an improvement on the problem of PM2.5 underestimation. Due to a better ability of capturing abrupt changes in the PM2.5 concentrations, the spatial evolution of PM2.5 during a typical pollution event could be mapped more accurately. Finally, the analysis of variable importance showed that the three most important variables in the estimation of the low-frequency coefficients of PM2.5 (PM2.5_A4) were temperature at 2 m (T2), day of year (DOY) and longitude (LON), while that in the high-frequency coefficients of PM2.5 (PM2.5_D) were CO, AOD and NO2. This study not only provided an effective solution to the PM2.5 underestimation and AOD missing problems in the PM2.5 estimation, but also proposed a new method to further refine the sophisticated correlations between PM2.5 and some spatiotemporal variables.

Spatiotemporal characteristics of PM2.5 concentration in the Yangtze River Delta urban agglomeration, China on the application of big data and wavelet analysis.

PM2.5 pollution has been one of the main environmental issues of concern for the Yangtze River Delta Urban Agglomeration (YRDUA) during the recent decade. In this paper, allied with big data and wavelet analysis, spatiotemporal variations of PM2.5 and its influencing factors (air pollutants and meteorological factors) are studied based on hourly concentrations of PM2.5 from 2015 to 2018 in the YRDUA. Results showed that PM2.5 presented a step-shaped decline from northwest to southeast in space and significant multi-scale temporal variations in time. On the macroscopic level, PM2.5 concentrations decreased from 2015 to 2018, showing a U-shaped pattern within a year. On the microscopic level, it had a four-stage annual variation (January to March, April to June, July to September, October to December) and the mutation events mainly occurred in winter. There were two dominant periods of PM2.5, an annual cycle on the time scale of 250-480 d and a semi-annual cycle on the time scale of 130-220 d. In addition, PM2.5 showed time scale-dependent correlations with air pollutants and meteorological factors. Among air pollutants, the correlation between PM2.5 and CO was the most consistent, and the correlation between PM2.5 and SO2/NO2 improved with the increase of time scale, while the correlation between PM2.5 and O3 was positive at shorter time scales but negative at broader time scales. Among meteorological factors, the correlations between PM2.5 and wind speed, precipitation, temperature, air pressure and relative humidity were mainly reflected at broader time scales. These findings would be helpful to improve the accuracy of prediction model and provide references for the ongoing joint prevention and control.

Biohydrogen production from enzymatic hydrolysis of food waste in batch and continuous systems.

In this study, the feasibility of biohydrogen production from enzymatic hydrolysis of food waste was investigated. Food waste (solid-to-liquid ratio of 10%, w/v) was first hydrolyzed by commercial glucoamylase to release glucose (24.35 g/L) in the food waste hydrolysate. Then, the obtained food waste hydrolysate was used as substrate for biohydrogen production in the batch and continuous (continuous stirred tank reactor, CSTR) systems. It was observed that the maximum cumulative hydrogen production of 5850 mL was achieved with a yield of 245.7 mL hydrogen/g glucose (1.97 mol hydrogen/mol glucose) in the batch system. In the continuous system, the effect of hydraulic retention time (HRT) on biohydrogen production from food waste hydrolysate was investigated. The optimal HRT obtained from this study was 6 h with the highest hydrogen production rate of 8.02 mmol/(h·L). Ethanol and acetate were the major soluble microbial products with low propionate production at all HRTs. Enzymatic hydrolysis of food waste could effectively accelerate hydrolysis speed, improve substrate utilization rate and increase hydrogen yield.