Joint multiscale cloud detection algorithm for ground-based lidar.

A ground-based lidar is a powerful tool for studying the vertical structure and optical properties of clouds. A layer detection algorithm is important to determine the presence and spatial position of clouds from vast lidar signals. However, current detection algorithms for ground-based lidar still involve substantial missing and false detections for tenuous layers and layer edges. Here, a joint multiscale cloud layer detection algorithm is proposed. The algorithm can effectively capture the tenuous layers and layer edges by using joint multiscale detection methods based on a trend function and the Bernoulli distribution assumption. Results show that the proposed algorithm detects 10.45% more cloud layers than the official cloud product of Micro Pulse Lidar Network (MPLNET) does. Specifically, 7.93% and 12.57% more cloud layers are detected at daytime and nighttime, respectively. The evaluation based on depolarization properties proves that the additional cloud layers detected by the joint multiscale algorithm are reliable. These additional detected clouds have important implications for cloud climatology and climate change research. The new algorithm remarkably enhances the cloud detection capability of ground-based lidar and potentially be widely used by the community.

CALIOP retrieval of droplet effective radius accounting for cloud vertical homogeneity.

Monitoring cloud droplet effective radius (re) is of great significance for studying aerosol-cloud interactions (ACI). Passive satellite retrieval, e.g., MODIS (Moderate Resolution Imaging Spectroradiometer), requires sunlight. This requirement prompted developing re retrieval using active sensors, e.g., CALIOP (Cloud-Aerosol Lidar with Orthogonal Polarization). Given the highest sensitivity of vertically homogeneous clouds to aerosols that feed to cloud base, here CALIOP profile measurements were used for the first time to quantify cloud vertical homogeneity and estimate cloud re during both day and night. Comparison using simultaneous Aqua-MODIS measurements demonstrates that CALIOP retrieval has the highest accuracy for vertically homogeneous clouds, with R2 (MAE, RMSE) of 0.72 (1.75 µm, 2.25 µm), while the accuracy is lowest for non-homogeneous clouds, with R2 (MAE, RMSE) of 0.60 (2.90 µm, 3.70 µm). The improved re retrieval in vertically homogeneous clouds provides a basis for possible breakthrough insights in ACI by CALIOP since re in such clouds reflects most directly aerosol effects on cloud properties. Global day-night maps of cloud vertical homogeneity and respective re are presented.

Approach for selecting boundary value to retrieve Mie-scattering lidar data based on segmentation and two-component fitting methods.

Fernald method is regarded as the standard method for retrieving lidar data, but the retrieval can be performed only when a boundary value is given. Generally, we can select clear atmosphere above the tropopause as a reference to determine the boundary value, but we need to use the slope method to fit the boundary value when the detecting range is lower than the tropopause. The slope method involves significant uncertainty because this algorithm is based on two hypotheses: one is that aerosol vertical distribution is homogeneous, and the other is that either molecule or aerosol components exist in the atmosphere. To reduce the uncertainty, we proposed a new approach, which segments a signal into "uniform" sub-signals to avoid the first hypothesis, and then uses nonlinear two-component fitting to avoid the second one. Compared with the approach based on the slope method, the new approach obtained more accurate boundary values and retrieving results for both of the simulated and real signals. Thus the automatic segmentation algorithm and the two-component fitting method are useful for determining the reference bin and boundary values when the effective detecting range of lidar is lower than the tropopause.

De-noising and retrieving algorithm of Mie lidar data based on the particle filter and the Fernald method.

The signal-to-noise ratio (SNR) of an atmospheric lidar decreases rapidly as range increases, so that maintaining high accuracy when retrieving lidar data at the far end is difficult. To avoid this problem, many de-noising algorithms have been developed; in particular, an effective de-noising algorithm has been proposed to simultaneously retrieve lidar data and obtain a de-noised signal by combining the ensemble Kalman filter (EnKF) and the Fernald method. This algorithm enhances the retrieval accuracy and effective measure range of a lidar based on the Fernald method, but sometimes leads to a shift (bias) in the near range as a result of the over-smoothing caused by the EnKF. This study proposes a new scheme that avoids this phenomenon using a particle filter (PF) instead of the EnKF in the de-noising algorithm. Synthetic experiments show that the PF performs better than the EnKF and Fernald methods. The root mean square error of PF are 52.55% and 38.14% of that of the Fernald and EnKF methods, and PF increases the SNR by 44.36% and 11.57% of that of the Fernald and EnKF methods, respectively. For experiments with real signals, the relative bias of the EnKF is 5.72%, which is reduced to 2.15% by the PF in the near range. Furthermore, the suppression impact on the random noise in the far range is also made significant via the PF. An extensive application of the PF method can be useful in determining the local and global properties of aerosols.

Nonlinear physical segmentation algorithm for determining the layer boundary from lidar signal.

Layer boundary (base and top) detection is a basic problem in lidar data processing, the results of which are used as inputs of optical properties retrieval. However, traditional algorithms not only require manual intervention but also rely heavily on the signal-to-noise ratio. Therefore, we propose a robust and automatic algorithm for layer detection based on a novel algorithm for lidar signal segmentation and representation. Our algorithm is based on the lidar equation and avoids most of the limitations of the traditional algorithms. Testing of the simulated and real signals shows that the algorithm is able to position the base and top accurately even with a low signal to noise ratio. Furthermore, the results of the classification are accurate and satisfactory. The experimental results confirm that our algorithm can be used for automatic detection, retrieval, and analysis of lidar data sets.

Linear segmentation algorithm for detecting layer boundary with lidar.

The automatic detection of aerosol- and cloud-layer boundary (base and top) is important in atmospheric lidar data processing, because the boundary information is not only useful for environment and climate studies, but can also be used as input for further data processing. Previous methods have demonstrated limitations in defining the base and top, window-size setting, and have neglected the in-layer attenuation. To overcome these limitations, we present a new layer detection scheme for up-looking lidars based on linear segmentation with a reasonable threshold setting, boundary selecting, and false positive removing strategies. Preliminary results from both real and simulated data show that this algorithm cannot only detect the layer-base as accurate as the simple multi-scale method, but can also detect the layer-top more accurately than that of the simple multi-scale method. Our algorithm can be directly applied to uncalibrated data without requiring any additional measurements or window size selections.

Anti-noise algorithm of lidar data retrieval by combining the ensemble Kalman filter and the Fernald method.

The lidar signal-to-noise ratio decreases rapidly with an increase in range, which severely affects the retrieval accuracy and the effective measure range of a lidar based on the Fernald method. To avoid this issue, an alternative approach is proposed to simultaneously retrieve lidar data accurately and obtain a de-noised signal as a by-product by combining the ensemble Kalman filter and the Fernald method. The dynamical model of the new algorithm is generated according to the lidar equation to forecast backscatter coefficients. In this paper, we use the ensemble sizes as 60 and the factor δ(1/2) as 1.2 after being weighed against the accuracy and the time cost based on the performance function we define. The retrieval and de-noising results of both simulated and real signals demonstrate that our method is practical and effective. An extensive application of our method can be useful for the long-term determining of the aerosol optical properties.

Spatio-temporal association mining of intercity PM2.5 pollution: Hubei Province in China as an example.

The complex interaction between emissions, meteorology, and atmospheric chemistry makes accurate predictions of particulate pollution difficult. Advanced data mining techniques can reveal potential laws, providing new possibilities for understanding the evolution and causes of air pollution. Based on the Granger method and block modeling analysis, this paper explored the intercity spillover effects of hourly PM2.5 in Hubei Province, China, to determine the specific role (i.e., overflow, limited overflow, bilateral, inflow, and limited inflow) of each city on regional pollution formation. Furthermore, a dynamic Apriori algorithm considering time-lag effects was used to mine the spatio-temporal associations of extreme PM2.5 pollution events among different cities. Results suggest that the northern and central cities with high-level PM2.5 concentration in Hubei have a significant spillover effect, whereas the eastern and southern cities generally play a role as the sink of pollutants. Based on the association rules of extreme PM2.5 pollution, four main pollutant transport channels were excavated and well matched with the trajectories extracted by the atmospheric model. This paper provides new insights for exploring the interaction of intercity particulate pollution, which is a supplement and cross-validation of the model results.

Improving PM2.5 predictions during COVID-19 lockdown by assimilating multi-source observations and adjusting emissions.

The Coronavirus Disease 2019 (COVID-19) outbreak caused a suspension of almost all non-essential human activities, leading to a significant reduction of anthropogenic emissions. However, the emission inventory of the chemistry transport model cannot be updated in time, resulting in large uncertainty in PM2.5 predictions. This study adopted a three-dimensional variational approach to assimilate multi-source PM2.5 data from satellite and ground observations and jointly adjusted emissions to improve PM2.5 predictions of the WRF-Chem model. Experiments were conducted to verify the method over Hubei Province, China, during the COVID-19 epidemic from Jan 21st to Mar 20th, 2020. The results showed that PM2.5 predictions were improved at almost all the validation sites, and the benefit of data assimilation (DA) can last for 48 h. However, the benefits of DA diminished quickly with the increase of the forecast time. By adjusting emissions, the PM2.5 predictions showed a much slower error accumulation along forecast time. At 48Z, the RMSE still has an 8.85 µg/m3 (19.49%) improvement, suggesting the effectiveness of emissions adjustment based on the improved initial conditions via DA.

Coarse sea spray inhibits lightning.

The known effects of thermodynamics and aerosols can well explain the thunderstorm activity over land, but fail over oceans. Here, tracking the full lifecycle of tropical deep convective cloud clusters shows that adding fine aerosols significantly increases the lightning density for a given rainfall amount over both ocean and land. In contrast, adding coarse sea salt (dry radius > 1 µm), known as sea spray, weakens the cloud vigor and lightning by producing fewer but larger cloud drops, which accelerate warm rain at the expense of mixed-phase precipitation. Adding coarse sea spray can reduce the lightning by 90% regardless of fine aerosol loading. These findings reconcile long outstanding questions about the differences between continental and marine thunderstorms, and help to understand lightning and underlying aerosol-cloud-precipitation interaction mechanisms and their climatic effects.

Simple multiscale algorithm for layer detection with lidar.

Lidar is a powerful active remote sensing device used in the detection of the optical properties of aerosols and clouds. However, there are difficulties in layer detection and classification. Many previous methods are too complex for large dataset analysis or limited to data with too high a signal-to-noise ratio (SNR). In this study, a mechanism of multiscale detection and overdetection rejection is proposed based on a trend index function that we define. Finally, we classify layers based on connected layers employing a quantity known as the threshold of the peak-to-base ratio. We find good consistency between retrieved results employing our method and visual analysis. The testing of synthetic signals shows that our algorithm performs well with SNRs higher than 4. The results demonstrate that our algorithm is simple, practical, and suited to large dataset applications.

Estimating hourly full-coverage PM2.5 over China based on TOA reflectance data from the Fengyun-4A satellite.

It is challenging to retrieve hourly ground-level PM2.5 on a national scale in China due to the sparse site measurements and the limited coverage of Low Earth Orbit (LEO) satellite observations. The new geostationary meteorological satellite of China, Fengyun-4A (FY-4A), provides a unique opportunity to fill this gap. In this study, the Random Forest (RF) algorithm was applied to retrieve hourly PM2.5 of China directly from FY-4A Top-of-Atmosphere (TOA) reflectance data. A one-year PM2.5 retrieval shows a strong agreement to ground-based measurements, with the averaged R2 approaching 0.92, while the RMSE was only 10.0 µg/m³. An analysis of the regional differences of the performance and the dependency on satellite Viewing Zenith Angle (VZA) show that sparse measurements, high VZA, and solar zenith angle (SZA) are the primary sources of the uncertainty. The use of the FY-4A improved 17% spatial coverage compared to the Himawari-8-based PM2.5 retrievals, enabling full-coverage, hourly PM2.5 monitoring over China, and potentially could improve PM2.5 predictions from air quality models after data assimilation.

Estimation of hourly PM1 concentration in China and its application in population exposure analysis.

Particulate pollution is closely related to public health. PM1 (particles with an aerodynamic size not larger than 1 µm) is much more harmful than particles with larger sizes because it goes deeper into the body and hence arouses social concern. However, the sparse and unevenly distributed ground-based observations limit the understanding of spatio-temporal distributions of PM1 in China. In this study, hourly PM1 concentrations in central and eastern China were retrieved based on a random forest model using hourly aerosol optical depth (AOD) from Himawari-8, meteorological and geographic information as inputs. Here the spatiotemporal autocorrelation of PM1 was also considered in the model. Experimental results indicate that although the performance of the proposed model shows diurnal, seasonal and spatial variations, it is relatively better than others, with a determination coefficient (R2) of 0.83 calculated based on the 10-fold cross validation method. Geographical map implies that PM1 pollution level in Beijing-Tianjin-Hebei is much higher than in other regions, with the mean value of ∼55 µg/m3. Based on the exposure analysis, we found about 75% of the population lives in an environment with PM1 higher than 35 µg/m3 in the whole study area. The retrieval dataset in this study is of great significance for further exploring the impact of PM1 on public health.

Improved PM2.5 predictions of WRF-Chem via the integration of Himawari-8 satellite data and ground observations.

The new-generation geostationary satellites feature higher radiometric, spectral, and spatial resolutions, thereby making richer data available for the improvement of PM2.5 predictions. Various aerosol optical depth (AOD) data assimilation methods have been developed, but the accurate representation of the AOD-PM2.5 relationship remains challenging. Empirical statistical methods are effective in retrieving ground-level PM2.5, but few have been evaluated in terms of whether and to what extent they can help improve PM2.5 predictions. Therefore, an empirical and statistics-based scheme was developed for optimizing the estimation of the initial conditions (ICs) of aerosol in WRF-Chem (Weather Research and Forecasting/Chemistry) and for improving the PM2.5 predictions by integrating Himawari-8 data and ground observations. The proposed method was evaluated via two one-year experiments that were conducted in parallel over eastern China. The contribution of the satellite data to the model performance was evaluated via a 2-week control experiment. The results demonstrate that the proposed method improved the PM2.5 predictions throughout the year and mitigated the underestimation during pollution episodes. Spatially, the performance was highly correlated with the amount of valid data.

Age- and season-specific effects of ambient particles (PM1, PM2.5, and PM10) on daily emergency department visits among two Chinese metropolitan populations.

BACKGROUND: Ambient PM2.5 has been identified as the top leading cause of risk-attributable deaths worldwide, particularly in China. Evidence suggested that PM1 contributed the most majority of PM2.5 concentrations in Chinese cities. However, epidemiologic knowledge to date is of wide lack regarding PM1-associated health effects. METHODS: We collected daily records of all-cause emergency department visits (EDVs) and ground measurements of ambient air pollutants and meteorological factors in Guangzhou and Shenzhen, China, 2015-2016. Case-crossover design and conditional logistic regression models were used to comparatively assess the short-term effects of ambient PM1, PM2.5, and PM10 on EDVs. Stratified analyses by gender, age and season were performed to identify vulnerable groups and periods. RESULTS: PM1, PM2.5 and PM10 were all significantly associated with increased EDVs in both cities. Population risks for EDVs increased by 2.2% [95% confidence interval, 1.8 to 2.6] in Guangzhou and 1.7% [1.0 to 2.4] in Shenzhen, for a 10 µg/m3 rise in PM1 at lag 0-1 days and lag 0-4 days, respectively. Relatively lower risks were found to be associated with PM2.5 and PM10. PM-EDVs associations exhibited no gender differences, but varied across age groups. Compared with adults and the elderly, children under 14 years-of-age suffered higher PM-induced risks. Results from both cities suggested greatly significant effect modification by season, with consistently stronger PM-EDVs associations during cold months. CONCLUSIONS: Our study added comparative evidence for increased EDVs risks associated with short-term exposures to ambient PM1, PM2.5 and PM10. Besides, PM-associated effects were significantly stronger among children and during cold months.

Short-term effects of ambient PM1 and PM2.5 air pollution on hospital admission for respiratory diseases: Case-crossover evidence from Shenzhen, China.

BACKGROUND: Ambient PM1 (particulate matter with aerodynamic diameter ≤1 µm) is an important contribution of PM2.5 mass. However, little is known worldwide regarding the PM1-associated health effects due to a wide lack of ground-based PM1 measurements from air monitoring stations. METHODS: We collected daily records of hospital admission for respiratory diseases and station-based measurements of air pollution and weather conditions in Shenzhen, China, 2015-2016. Time-stratified case-crossover design and conditional logistic regression models were adopted to estimate hospitalization risks associated with short-term exposures to PM1 and PM2.5. RESULTS: PM1 and PM2.5 showed significant adverse effects on respiratory disease hospitalizations, while no evident associations with PM1-2.5 were identified. Admission risks for total respiratory diseases were 1.09 (95% confidence interval: 1.04 to 1.14) and 1.06 (1.02 to 1.10), corresponding to per 10 µg/m3 rise in exposure to PM1 and PM2.5 at lag 0-2 days, respectively. Both PM1 and PM2.5 were strongly associated with increased admission for pneumonia and chronic obstructive pulmonary diseases, but exhibited no effects on asthma and upper respiratory tract infection. Largely comparable risk estimates were observed between male and female patients. Groups aged 0-14 years and 45-74 years were significantly affected by PM1- and PM2.5-associated risks. PM-hospitalization associations exhibited a clear seasonal pattern, with significantly larger risks in cold season than those in warm season among some subgroups. CONCLUSIONS: Our study suggested that PM1 rather than PM1-2.5 contributed to PM2.5-induced risks of hospitalization for respiratory diseases and effects of PM1 and PM2.5 mainly occurred in cold season.

Two-stage model for estimating the spatiotemporal distribution of hourly PM1.0 concentrations over central and east China.

Widespread and severe PM1.0 (particulate matter ≤1.0 µm) pollution in China has a significant negative influence on human health. However, knowledge of the regional spatiotemporal distribution of PM1.0 has been hindered by sparsely distributed PM1.0 concentration data. In this work, a two-stage model (linear mixed effect-bagged tree model) was proposed for estimating hourly PM1.0 pollution levels from July 2015 to June 2017 over central and east China by using Himawari-8 aerosol products and coincident geographic data, meteorology, and site-based PM1.0 concentrations from ground monitoring network. The cross-validation for the developed model displayed R2 and mean absolute error value of 0.80 and 9.3 µg/m3, respectively. Validation demonstrated that the model accurately estimated hourly PM1.0 concentrations with high R2 of 0.63-0.85 and low bias of 8.7-10.1 µg/m3. The estimated PM1.0 concentrations on daily scale showed peaks with PM1.0 of 36.9 ±â€¯8.4 µg/m3 at rush hours during daytime. Seasonal distribution displayed that summer was cleanest with an average PM1.0 of 20.9 ±â€¯6.8 µg/m3 and winter was the most polluted season with an average PM1.0 of 45.6 ±â€¯16.8 µg/m3. These results indicated that the proposed satellite-based model can estimate reliable spatial distribution of PM1.0 concentrations over a large-scale region.

Estimation of spatiotemporal PM1.0 distributions in China by combining PM2.5 observations with satellite aerosol optical depth.

Particulates smaller than 1.0 µm (PM1.0) have strong associations with public health and environment, and considerable exposure data should be obtained to understand the actual environmental burden. This study presented a PM1.0 estimation strategy based on the generalised regression neural network model. The proposed strategy combined ground-based observations of PM2.5 and satellite-derived aerosol optical depth (AOD) to estimate PM1.0 concentrations in China from July 2015 to June 2017. Results indicated that the PM1.0 estimates agreed well with the ground-based measurements with an R2 of 0.74, root mean square error of 19.0 µg/m3 and mean absolute error of 11.4 µg/m3 as calculated with the tenfold cross-validation method. The diurnal estimation performance displayed remarkable single-peak variation with the highest R2 of 0.80 at noon, and the seasonal estimation performance showed that the proposed method could effectively capture high-pollution events of PM1.0 in winter. Spatially, the most polluted areas were clustered in the North China Plain, where the average estimates presented a bimodal distribution during daytime. In addition, the quality of satellite-derived AOD, the robustness of the interpolation algorithm and the proportion of PM1.0 in PM2.5 were confirmed to affect the estimation accuracy of the proposed model.

Estimating hourly PM1 concentrations from Himawari-8 aerosol optical depth in China.

Particulate matter with diameter less than 1 µm (PM1) has been found to be closely associated with air quality, climate changes, and even adverse human health. However, a large gap in our knowledge concerning the large-scale distribution and variability of PM1 remains, which is expected to be bridged with advanced remote-sensing techniques. In this study, a hybrid model called principal component analysis-general regression neural network (PCA-GRNN) is developed to estimate hourly PM1 concentrations from Himawari-8 aerosol optical depth in combination with coincident ground-based PM1 measurements in China. Results indicate that the hourly estimated PM1 concentrations from satellite agree well with the measured values at national scale, with R2 of 0.65, root-mean-square error (RMSE) of 22.0 µg/m3 and mean absolute error (MAE) of 13.8 µg/m3. On daily and monthly time scales, R2 increases to 0.70 and 0.81, respectively. Spatially, highly polluted regions of PM1 are largely located in the North China Plain and Northeast China, in accordance with the distribution of industrialisation and urbanisation. In terms of diurnal variability, PM1 concentration tends to peak in rush hours during the daytime. PM1 exhibits distinct seasonality with winter having the largest concentration (31.5±3.5 µg/m3), largely due to peak combustion emissions. We further attempt to estimate PM2.5 and PM10 with the proposed method and find that the accuracies of the proposed model for PM1 and PM2.5 estimation are significantly higher than that of PM10. Our findings suggest that geostationary data is one of the promising data to estimate fine particle concentration on large spatial scale.

Evaluation of VIIRS Land Aerosol Model Selection with AERONET Measurements.

The Visible Infrared Imaging Radiometer Suite (VIIRS) is a next-generation polar-orbiting operational environmental sensor with a capability for global aerosol observations. Identifying land aerosol types is important because aerosol types are a basic input in retrieving aerosol optical properties for VIIRS. The VIIRS algorithm can automatically select the optimal land aerosol model by minimizing the residual between the derived and expected spectral surface reflectance. In this study, these selected VIIRS aerosol types are evaluated using collocated aerosol types obtained from the Aerosol Robotic Network (AERONET) level 1.5 from 23 January 2013 to 28 February 2017. The spatial distribution of VIIRS aerosol types and the aerosol optical depth bias (VIIRS minus AERONET) demonstrate that misidentifying VIIRS aerosol types may lead to VIIRS retrieval being overestimated over the Eastern United States and the developed regions of East Asia, as well as underestimated over Southern Africa, India, and Northeastern China. Approximately 22.33% of VIIRS aerosol types are coincident with that of AERONET. The agreements between VIIRS and AERONET for fine non-absorbing and absorbing aerosol types are approximately 36% and 57%, respectively. However, the agreement between VIIRS and AERONET is extremely low (only 3.51%). The low agreement for coarse absorbing dust may contribute to the poor performance of VIIRS retrieval under the aerosol model (R = 0.61). Results also show that an appropriate aerosol model can improve the retrieval performance of VIIRS over land, particularly for dust type (R increases from 0.61 to 0.72).