Development,Calibration And Field Evaluation Of Low Cost Aerosol Monitors

In recent years,frequent fog and haze in China has become an environmental and social issue of great concern.PM_2.5 is the main pollutant causing haze weather.PM_2.5 is very harmful to human health.However,the existing air quality monitoring infrastructure cannot fulfill the need to characterize urban PM_2.5 concentrations spatial and temporal variability.Data collected by low-cost aerosol monitors can be used to supplement national air pollution monitors.In this study,three low-cost monitors based on Shinyei PPD42NS aerosol sensors were established,and the BAM-1020 located in the same location was used as the calibration instrument for long-term（18-month）field calibration（December 2015 to May 2017）.Using three methods of data transformation,including linear,power law and artificial neural network（ANN）,will be to convert instrument electrical readings to ambient aerosol concentrations,and further discussing the stability and consistency of low-cost aerosol sensors in long-term field observation experiments.And the environmental impact of the sensor,including temperature and humidity,on the performance of the sensor.Research shows that:（1）The three low-cost aerosol monitors established in the study can accurately measure PM_2.5.5 concentration in the air under non-condensing conditions（RH<95%）;（2）Among the three calibration methods,the ANN approach shows the highest correlation between the estimated and BAM-1020 measured hourly PM_2.5,the correlation coefficient R²is 0.84,and the MNB and MNE are 12.67%and 29.71%.The correlation coefficients R² of linear analysis and power law analysis are 0.75 and 0.71,MNB is 22.23%and 23.8%,MNE is41.52%and 45.24%,respectively.linear fit demonstrates a slightly better agreement with BAM-1020 hourly PM_2.5 than law equation linear fit method.The better performance of ANN is mainly due to including temperature and relative humidity（RH）as input data in addition to the raw sensor output of Low-pulse Occupancy Ratio（LOR）.（3）The low-cost monitors have relatively better performance for high aerosol concentrations but have larger errors measuring concentrations under 35μg·m^-3.Both MNB and MNE are over 60%,and only 22%of data points have a relative error within±20%;High humidity（RH>75%）could cause larger errer for the low-cost monitors;When the temperature is higher than 0℃,the effect of temperature on the low-cost aerosol monitor is not obvious,and when the temperature is lower than 0℃（-5～0℃ during Nanjing research period）,the sensor performance is the best,MNB,MNE and the ratio of the relative error within±20%is0.03%,0.13%and 83%,respectively.（4）The performance of the low-cost monitor deteriorates after 18 months on continuous operation,MNB and MNE gradually increased（MNB increased from 9% to 13%,MNE increased from 19%to 30%）,and the ratio of the relative error within±20%gradually decreased（from 0.72%to 0.52%）.which indicates that regular replacement or cleaning of the PM optical sensing unit is needed to use the low-cost monitor for long-term community monitoring.（5）High consistency between different types of monitors,but to ensure monitor accuracy,different types of monitors should be individually calibrated.