Showing papers on "Haze published in 2022"

Spatial and temporal distribution characteristics of haze and pollution particles in China based on spatial statistics

Abstract: In recent years, the occurrence and frequency of haze have constantly been increasing, bringing severe threats to people's daily lives. To this end, this paper discusses the spatial and temporal distribution characteristics of severe haze in China, analyzes the interaction between haze pollution and the influence of economy and energy structure on haze in 31 provinces of China. It provides references for the treatment of haze weather and the prevention and control of air pollution in China. This paper mainly adopts the spatial autocorrelation method. The data processed mainly includes API (Air Pollution Index) and meteorological station data. Combined with the statistical yearbook data, this paper conducts multi-aspect research and exploration. By using statistical methods to study the haze distribution in China, we found that the haze and PM2.5 concentrations were mainly distributed in Beijing-Tianjin-Hebei, Shandong Province, the northern northwest, southeastern Sichuan, and Chongqing. Haze distribution has obvious seasonality, more in winter and less in summer. There are also regional differences in the concentration distribution of urban pollutants. The concentration of SO2 and absorbable particles are relatively high in northern cities. In contrast, that of southern cities is relatively low and changes with seasonal changes.

Spatial and temporal distribution characteristics of haze and pollution particles in China based on spatial statistics

Abstract: In recent years, the occurrence and frequency of haze have constantly been increasing, bringing severe threats to people's daily lives. To this end, this paper discusses the spatial and temporal distribution characteristics of severe haze in China, analyzes the interaction between haze pollution and the influence of economy and energy structure on haze in 31 provinces of China. It provides references for the treatment of haze weather and the prevention and control of air pollution in China. This paper mainly adopts the spatial autocorrelation method. The data processed mainly includes API (Air Pollution Index) and meteorological station data. Combined with the statistical yearbook data, this paper conducts multi-aspect research and exploration. By using statistical methods to study the haze distribution in China, we found that the haze and PM 2.5 concentrations were mainly distributed in Beijing-Tianjin-Hebei, Shandong Province, the northern northwest, southeastern Sichuan, and Chongqing. Haze distribution has obvious seasonality, more in winter and less in summer. There are also regional differences in the concentration distribution of urban pollutants. The concentration of SO 2 and absorbable particles are relatively high in northern cities. In contrast, that of southern cities is relatively low and changes with seasonal changes. • Focuses on the spatial and temporal distribution characteristics of severe haze in China based on pollution-related data from the Ministry of Environmental Protection. • Analyzes the interaction between haze and the economy and energy structure in 31 provinces of China, and provides reference for the treatment and the prevention and control of air pollution in China. • The haze were concentrated in Beijing-Tianjin-Hebei region, Shandong Province, the northern northwest region, southeastern Sichuan province, and Chongqing. • According to the spatial autocorrelation, haze distribution has obvious seasonality, more in winter and less in summer. • The concentration of SO 2 and absorbable particles is relatively high in northern cities, while that of southern cities is relatively low, and changes with seasonal changes

Smog prediction based on the deep belief - BP neural network model (DBN-BP)

Abstract: Smog pollution is becoming a significant problem for people worldwide, becoming an essential threat to the global environment. Many studies on haze already exist, which still need to continue in-depth research to better deal with haze problems. Due to its unique geographical environment, Sichuan has become one of the areas with severe smog pollution. Therefore, the research and prediction of smog pollution in Sichuan has become an urgent need. This paper proposes a deep learning technology based on a Deep Belief-Back Propagation neural network. It makes in-depth prediction research by using the air pollution data of PM2.5, PM10, O 3 , CO NO 2, and SO 2 in Sichuan smog to provide a decision-making basis for predicting and preventing smog polluted weather. According to the prediction results of the model, the concentrations of PM2.5 and PM10 in Chengdu were predicted. The analysis shows that the larger the number of hidden layers in the belief network, the higher the prediction accuracy. Under the same network, the prediction accuracy of PM2.5 is significantly higher than that of PM10. Compared with the traditional Back Propagation neural network, the prediction effect of the Deep Belief-Back Propagation neural network is better. • Based on the air pollution data of PM2.5, PM10, O3, Co, NO2 and SO2 in Sichuan smog, this paper makes an prediction research for predicting and preventing smog polluted weather. • Aiming at the problem of PM2.5 and PM10 concentration prediction, this paper constructs a prediction model based on Deep Belief -Back Propagation neural network. • The results show that the prediction effect of Deep Belief - BP neural network is better than BP neural network, and can quickly extract the concentration characteristics in the first 24 times.

The dynamic atmospheric and aeolian environment of Jezero crater, Mars

Abstract: Despite the importance of sand and dust to Mars geomorphology, weather, and exploration, the processes that move sand and that raise dust to maintain Mars’ ubiquitous dust haze and to produce dust storms have not been well quantified in situ, with missions lacking either the necessary sensors or a sufficiently active aeolian environment. Perseverance rover’s novel environmental sensors and Jezero crater’s dusty environment remedy this. In Perseverance’s first 216 sols, four convective vortices raised dust locally, while, on average, four passed the rover daily, over 25% of which were significantly dusty (“dust devils”). More rarely, dust lifting by nonvortex wind gusts was produced by daytime convection cells advected over the crater by strong regional daytime upslope winds, which also control aeolian surface features. One such event covered 10 times more area than the largest dust devil, suggesting that dust devils and wind gusts could raise equal amounts of dust under nonstorm conditions.

Pan-Arctic seasonal cycles and long-term trends of aerosol properties from 10 observatories

Abstract: Abstract. Even though the Arctic is remote, aerosol properties observed there are strongly influenced by anthropogenic emissions from outside the Arctic. This is particularly true for the so-called Arctic haze season (January through April). In summer (June through September), when atmospheric transport patterns change, and precipitation is more frequent, local Arctic sources, i.e., natural sources of aerosols and precursors, play an important role. Over the last few decades, significant reductions in anthropogenic emissions have taken place. At the same time a large body of literature shows evidence that the Arctic is undergoing fundamental environmental changes due to climate forcing, leading to enhanced emissions by natural processes that may impact aerosol properties. In this study, we analyze 9 aerosol chemical species and 4 particle optical properties from 10 Arctic observatories (Alert, Kevo, Pallas, Summit, Thule, Tiksi, Barrow/Utqiaġvik, Villum, and Gruvebadet and Zeppelin Observatory – both at Ny-Ålesund Research Station) to understand changes in anthropogenic and natural aerosol contributions. Variables include equivalent black carbon, particulate sulfate, nitrate, ammonium, methanesulfonic acid, sodium, iron, calcium and potassium, as well as scattering and absorption coefficients, single scattering albedo and scattering Ångström exponent. First, annual cycles are investigated, which despite anthropogenic emission reductions still show the Arctic haze phenomenon. Second, long-term trends are studied using the Mann–Kendall Theil–Sen slope method. We find in total 41 significant trends over full station records, i.e., spanning more than a decade, compared to 26 significant decadal trends. The majority of significantly declining trends is from anthropogenic tracers and occurred during the haze period, driven by emission changes between 1990 and 2000. For the summer period, no uniform picture of trends has emerged. Twenty-six percent of trends, i.e., 19 out of 73, are significant, and of those 5 are positive and 14 are negative. Negative trends include not only anthropogenic tracers such as equivalent black carbon at Kevo, but also natural indicators such as methanesulfonic acid and non-sea-salt calcium at Alert. Positive trends are observed for sulfate at Gruvebadet. No clear evidence of a significant change in the natural aerosol contribution can be observed yet. However, testing the sensitivity of the Mann–Kendall Theil–Sen method, we find that monotonic changes of around 5 % yr−1 in an aerosol property are needed to detect a significant trend within one decade. This highlights that long-term efforts well beyond a decade are needed to capture smaller changes. It is particularly important to understand the ongoing natural changes in the Arctic, where interannual variability can be high, such as with forest fire emissions and their influence on the aerosol population. To investigate the climate-change-induced influence on the aerosol population and the resulting climate feedback, long-term observations of tracers more specific to natural sources are needed, as well as of particle microphysical properties such as size distributions, which can be used to identify changes in particle populations which are not well captured by mass-oriented methods such as bulk chemical composition.

Revealing Drivers of Haze Pollution by Explainable Machine Learning

Abstract: Many places on earth still suffer from a high level of atmospheric fine particulate matter (PM2.5) pollution. Formation of a particulate pollution event or haze episode (HE) involves many factors, including meteorology, emissions, and chemistry. Understanding the direct causes of and key drivers behind the HE is thus essential. Traditionally, this is done via chemical transport models. However, substantial uncertainties are introduced into the model estimation when there are significant changes in the emissions inventory due to interventions (e.g., the COVID-19 lockdown). Here we applied a Random Forest model coupled with a Shapley additive explanation algorithm, a post hoc explanation technique, to investigate the roles of major meteorological factors, primary emissions, and chemistry in five severe HEs that occurred before or during the COVID-19 lockdown in China. We discovered that, in addition to the high level of primary emissions, PM2.5 in these haze episodes was largely driven by meteorological effects (with average contributions of 30-65 mu g m(-3) for the five HEs), followed by chemistry (similar to 15-30 mu g m(-3)). Photochemistry was likely the major pathway of formation of nitrate, while air humidity was the predominant factor in forming sulfate. Our results highlight that the machine learning driven by data has the potential to be a complementary tool in predicting and interpreting air pollution.

Evolving fusion-based visibility restoration model for hazy remote sensing images using dynamic differential evolution

Abstract: Remote sensing images taken during poor environmental conditions are degraded by the scattering of atmospheric particles, which affects the performance of many imaging systems. Hence, an efficient visibility restoration model is required to remove haze from distorted images. But, the design of visibility restoration models is an ill-posed problem as the physical information like depth information and attenuation model are usually unknown. The physical parameters computed using existing models such as dark channel prior and gradient channel prior are not accurate especially for images with large haze gradients. Therefore, in this paper, an evolving visibility restoration model is proposed for remote sensing images. Initially, the fusion-based transmission map is computed from the foreground and sky regions. The transmission map is further improved by designing a hybrid constraints-based variational model. Finally, a dynamic differential evolution is utilized to optimize the control parameters of the proposed model. The proposed model is validated on fifty synthetic benchmarks and fifty real-life remote sensing images. For comparative analysis, ten well-known restoration models are also considered. The comparative analysis demonstrates that the proposed model outperforms the existing restoration models.

Forest fire spreading and carbon concentration identification in tropical region Indonesia

Abstract: Global warming impacted the rise of temperature globally, some of the places a high risk of fire such as land and forest fire. Many efforts to prevent the occurrence of land and forest fire, but some methods are not achieved in optimum results. One of the issues is carbon emitted to the sky is in general concentration. The current sensor only detected carbon status and without the detail of the carbon concentration either from the forest fire or any other source that contribute carbon. This research identifies and detects the fog of haze emitted from a forest fire by identifying the carbon concentration. The carbon sensor detects particles flying in the skies and calculates based on number and size to identify if the fog comes from a forest fire or other fire sources. There are many other sources of haze in the skies. It can be from the pollution emitted from vehicles, fire from the garbage or rubbish, or fog emitted from the factory. The size and number of particles detected by the sensor were analyzed to identify the quantity and the size to match the type of particles emitted from the forest fire. Results show that particles from the forest fire are higher and bigger compare to other sources of fires. The fog's intensity less than from other fire sources because of forest fire the material mostly from the trees and leaves getting burn then gives less in quantity.

Gas-to-Aerosol Phase Partitioning of Atmospheric Water-Soluble Organic Compounds at a Rural Site in China: An Enhancing Effect of NH3 on SOA Formation.

Abstract: Partitioning gaseous water-soluble organic compounds (WSOC) to the aerosol phase is a major formation pathway of atmospheric secondary organic aerosols (SOA). However, the fundamental mechanism of the WSOC-partitioning process remains elusive. By simultaneous measurements of both gas-phase WSOC (WSOCg) and aerosol-phase WSOC (WSOCp) and formic and acetic acids at a rural site in the Yangtze River Delta (YRD) region of China during winter 2019, we showed that WSOCg during the campaign dominantly partitioned to the organic phase in the dry period (relative humidity (RH) < 80%) but to aerosol liquid water (ALW) in the humid period (RH > 80%), suggesting two distinct SOA formation processes in the region. In the dry period, temperature was the driving factor for the uptake of WSOCg. In contrast, in the humid period, the factors controlling WSOCg absorption were ALW content and pH, both of which were significantly elevated by NH3 through the formation of NH4NO3 and neutralization with organic acids. Additionally, we found that the relative abundances of WSOCp and NH4NO3 showed a strong linear correlation throughout China with a spatial distribution consistent with that of NH3, further indicating a key role of NH3 in WSOCp formation at a national scale. Since WSOCp constitutes the major part of SOA, such a promoting effect of NH3 on SOA production by elevating ALW formation and WSOCg partitioning suggests that emission control of NH3 is necessary for mitigating haze pollution, especially SOA, in China.

Haze Grading Using the Convolutional Neural Networks

Abstract: As an air pollution phenomenon, haze has become one of the focuses of social discussion. Research into the causes and concentration prediction of haze is significant, forming the basis of haze prevention. The inversion of Aerosol Optical Depth (AOD) based on remote sensing satellite imagery can provide a reference for the concentration of major pollutants in a haze, such as PM2.5 concentration and PM10 concentration. This paper used satellite imagery to study haze problems and chose PM2.5, one of the primary haze pollutants, as the research object. First, we used conventional methods to perform the inversion of AOD on remote sensing images, verifying the correlation between AOD and PM2.5. Subsequently, to simplify the parameter complexity of the traditional inversion method, we proposed using the convolutional neural network instead of the traditional inversion method and constructing a haze level prediction model. Compared with traditional aerosol depth inversion, we found that convolutional neural networks can provide a higher correlation between PM2.5 concentration and satellite imagery through a more simplified satellite image processing process. Thus, it offers the possibility of researching and managing haze problems based on neural networks.

120 Gbps SAC-OCDMA-OAM-based FSO transmission system: Performance evaluation under different weather conditions

Abstract: A novel 120 Gbps free space optics (FSO) transmission scheme is introduced by combining orbital angular momentum (OAM) multiplexed signals with spectral amplitude coded (SAC)-optical code division multiplexing access (OCDMA) technique. Four OAM beams, each carrying three independent channels with 10 Gbps data rate, are used for increasing the capacity of the FSO system to 120 Gbps. Enhanced double weight (EDW) codes are employed for the SAC-OCDMA system. The proposed system is simulated and its performance is compared for the twelve channels under different weather conditions including clear air (CA), varying levels of rain, haze and fog conditions. The obtained results reveal that longer propagation distances between transmitter and receiver are possible with a bit error rate (BER) of ∼ 10-5. The possible distances are, respectively, 300 m, 160 m, 200 m, and 150 m, under CA, heavy rain (HR), heavy haze (HH), and heavy fog (HF), with a system capacity of 120 Gbps.

Haze Risk Assessment Based on Improved PCA-MEE and ISPO-LightGBM Model

Abstract: With the economic development in China, haze risks are frequent. It is important to study the urban haze risk assessment to manage the haze disaster. The haze risk assessment indexes of 11 cities in Fenwei Plain were selected from three aspects: the sensitivity of disaster-inducing environments, haze component hazards and the vulnerability of disaster-bearing bodies, combined with regional disaster system theory. The haze hazard risk levels of 11 cities in Fenwei Plain were evaluated using the matter-element extension (MEE) model, and the indicator weights were determined by improving the principal component analysis (PCA) method using the entropy weight method, and finally, five haze hazard risk assessment models were established by improving the particle swarm optimization (IPSO) light gradient boosting machine (LightGBM) algorithm. It is used to assess the risk of affected populations, transportation damage risk, crop damage area risk, direct economic loss risk and comprehensive disaster risk before a disaster event occurs. The experimental comparison shows that the haze risk index of Xi’an city is the highest, and the full index can improve the evaluation accuracy by 4–16% compared with only the causative factor index, which indicates that the proposed PCA-MEE-ISPO-LightGBM model evaluation results are more realistic and reliable.

Overlooked Nonagricultural and Wintertime Agricultural NH3 Emissions in Quzhou County, North China Plain: Evidence from 15N-Stable Isotopes

Abstract: Ammonia (NH3) plays a critical role in atmospheric chemistry and can exacerbate haze formation. Agricultural emissions have been known as a primary source of global atmospheric NH3. However, with accelerating urbanization and optimized agricultural production, the dominance of agricultural emissions has become less clear. We investigated the contributions of nonagricultural and agricultural sources to atmospheric NH3 based on measurements of NH3 isotopes at nine sites in Quzhou County, a typical agricultural county in the North China Plain. We found that Quzhou had extremely high NH3 concentrations (annual average across all sites of 40.3 ± 3.3 μg m–3). We compared the sources of seasonal NH3 contributions in rural and urban areas through 15N-stable isotope analyses, which provides new insights into NH3 sources compared with the traditional emission inventories. In rural areas, agricultural sources (fertilizer application and livestock production) make significant contributions (56 ± 3%) to NH3 emissions in the winter, whereas there were larger contributions of nonagricultural sources [fossil fuel, waste, and biomass burning (56 ± 2%)] relative to agricultural sources in urban areas. More effective strategies are still needed for better manure management and vegetable/fruit production in the winter and for controlling nonagricultural sources, even in counties dominated by agriculture.

Self-augmented Unpaired Image Dehazing via Density and Depth Decomposition

Abstract: To overcome the overfitting issue of dehazing models trained on synthetic hazy-clean image pairs, many recent methods attempted to improve models' generalization ability by training on unpaired data. Most of them simply formulate dehazing and rehazing cycles, yet ignore the physical properties of the real-world hazy environment, i.e. the haze varies with density and depth. In this paper, we propose a self-augmented image dehazing framework, termed D4 (Dehazing via Decomposing transmission map into Density and Depth) for haze generation and removal. Instead of merely estimating transmission maps or clean content, the proposed framework focuses on exploring scattering coefficient and depth information contained in hazy and clean images. With estimated scene depth, our method is capable of re-rendering hazy images with different thick-nesses which further benefits the training of the dehazing network. It is worth noting that the whole training process needs only unpaired hazy and clean images, yet succeeded in recovering the scattering coefficient, depth map and clean content from a single hazy image. Comprehensive experiments demonstrate our method outperforms state-of-the-art unpaired dehazing methods with much fewer parameters and FLOPs. Our code is available at https://github.com/YaN9-Y/D4.

Performance improvement of mode division multiplexing free space optical communication system through various atmospheric conditions with a decision feedback equalizer

Abstract: Abstract Free space optics (FSO) systems use the atmosphere as a propagation medium. However, a common problem is atmospheric turbulence, including fog, rain, and haze that emerges between the transmitter and the receiver from time to time. These adverse weather conditions impose power loss on the optical signal, producing distortion and degrading bit error rate (BER) and throughput. To reduce the effect of atmospheric turbulence, this paper proposes a decision feedback equalizer (DFE) with minimum mean square error (MMSE) in mode division multiplexing (MDM) for the FSO system. The DFE with varying tap counts is investigated. The MMSE algorithm is utilized to optimize both the feedforward and feedback filter coefficients of the DFE. The proposed system consists of four parallel 2.5 Gbps channels that use Hermite-Gaussian (HG) modes. The results show that the DFE equalization scheme successfully transmits 10 Gbps over 40 m, 800 m, 1400 m, and 2 km in medium fog, medium rain, medium haze, and clear weather. Performance is analyzed in terms of BER and eye diagrams and compared with the traditional model. Based on BER and eye diagram results, DFE improves the outdoor FSO system immunity to distortion in medium fog, medium haze, medium rain, and clear weather while maintaining high throughput and desired low BER.

Corruption, market segmentation and haze pollution: empirical evidence from China

Abstract: Corruption and market segmentation generally result from inter-regional resource allocation mechanism at the level of government and market, and it is of great significance to clarify their effects on haze pollution for the healthy development of the regional economy. With theoretical analysis, this paper applies systematic GMM to examine the impact of corruption on haze pollution. The mediatory effect model is used to further evaluate the mediatory effect of market segmentation by using panel data for 30 provinces in China from 2006 to 2018. The evaluations reveal that corruption positively affects haze pollution at the 1% significance level and has a prominent “time inertia”. After alleviation of the endogenous problem and a series of robustness tests, this conclusion remains valid. Based on national samples, corruption, especially environmental corruption, not only directly provokes an increase in haze pollution, but also aggravates it through market segmentation, and, the impact of corruption on haze pollution in different regions and at different periods has significant heterogeneity. Therefore, policymakers should start from the institutional mechanism to curb haze pollution by improving the performance appraisal system. Moreover, the synergistic effect between anti-corruption and governance on the environment should be enhanced by improving the anti-corruption management system. Local protectionism should be eliminated to promote the integration of regional markets. A unified, open and organized market system should be established to form the synergy of governance on the environment.

Image Dehazing Based on Pixel Guided CNN with PAM via Graph Cut

Abstract: : Image dehazing is still an open research topic that has been under-going a lot of development, especially with the renewed interest in machine learning-based methods. A major challenge of the existing dehazing methods is the estimation of transmittance, which is the key element of haze-affected imaging models. Conventional methods are based on a set of assumptions that reduce the solution search space. However, the multiplication of these assumptions tends to restrict the solutions to particular cases that cannot account for the reality of the observed image. In this paper we reduce the number of simplified hypotheses in order to attain a more plausible and realistic solution by exploiting a priori knowledge of the ground truth in the proposed method. The proposed method relies on pixel information between the ground truth and haze image to reduce these assumptions. This is achieved by using ground truth and haze image to find the geometric-pixel information through a guided Convolution Neural Networks (CNNs) with a Parallax Attention Mechanism (PAM). It uses the differential pixel-based variance in order to estimate transmittance. The pixel variance uses local and global patches between the assumed ground truth and haze image to refine the transmission map. The transmission map is also improved based on improved Markov random field (MRF) energy functions. We used different images to test the proposed algorithm. The entropy value of the proposed method was 7.43 and 7.39, a percent increase of (cid:2) 4.35% and (cid:2) 5.42%, respectively, compared to the best existing results. The increment is similar in other performance quality metrics and this validate its superiority compared to other existing methods in terms of key image quality evaluation metrics. The proposed approach’s drawback, an over-reliance on real ground truth images, is also investigated. The proposed method show more details hence yields better images than those from the existing state-of-the-art-methods.

Haze Air Pollution Health Impacts of Breath-Borne VOCs.

Abstract: Here, we investigated the use of breath-borne volatile organic compounds (VOCs) for rapid monitoring of air pollution health effects on humans. Forty-seven healthy college students were recruited, and their exhaled breath samples (n = 235) were collected and analyzed for VOCs before, on, and after two separate haze pollution episodes using gas chromatography-ion mobility spectrometry (GC-IMS). Using a paired t-test and machine learning model (Gradient Boosting Machine, GBM), six exhaled VOC species including propanol and isoprene were revealed to differ significantly among pre-, on-, and post-exposure in both haze episodes, while none was found between clean control days. The GBM model was shown capable of differentiating between pre- and on-exposure to haze pollution with a precision of 90-100% for both haze episodes. However, poor performance was detected for the same model between two different clean days. In addition to gender and particular haze occurrence influences, correlation analysis revealed that NH4+, NO3-, acetic acid, mesylate, CO, NO2, PM2.5, and O3 played important roles in the changes in breath-borne VOC fingerprints following haze air pollution exposure. This work has demonstrated direct evidence of human health impacts of haze pollution while identifying potential breath-borne VOC biomarkers such as propanol and isoprene for haze air pollution exposure.

Forest fire spreading and carbon concentration identification in tropical region Indonesia

Abstract: Global warming impacted the rise of temperature globally, some of the places a high risk of fire such as land and forest fire. Many efforts to prevent the occurrence of land and forest fire, but some methods are not achieved in optimum results. One of the issues is carbon emitted to the sky is in general concentration. The current sensor only detected carbon status and without the detail of the carbon concentration either from the forest fire or any other source that contribute carbon. This research identifies and detects the fog of haze emitted from a forest fire by identifying the carbon concentration. The carbon sensor detects particles flying in the skies and calculates based on number and size to identify if the fog comes from a forest fire or other fire sources. There are many other sources of haze in the skies. It can be from the pollution emitted from vehicles, fire from the garbage or rubbish, or fog emitted from the factory. The size and number of particles detected by the sensor were analyzed to identify the quantity and the size to match the type of particles emitted from the forest fire. Results show that particles from the forest fire are higher and bigger compare to other sources of fires. The fog's intensity less than from other fire sources because of forest fire the material mostly from the trees and leaves getting burn then gives less in quantity.

Underwater image enhancement using adaptive color restoration and dehazing.

Abstract: Underwater images captured by optical cameras can be degraded by light attenuation and scattering, which leads to deteriorated visual image quality. The technique of underwater image enhancement plays an important role in a wide range of subsequent applications such as image segmentation and object detection. To address this issue, we propose an underwater image enhancement framework which consists of an adaptive color restoration module and a haze-line based dehazing module. First, we employ an adaptive color restoration method to compensate the deteriorated color channels and restore the colors. The color restoration module consists of three steps: background light estimation, color recognition, and color compensation. The background light estimation determines the image is blueish or greenish, and the compensation is applied in red-green or red-blue channels. Second, the haze-line technique is employed to remove the haze and enhance the image details. Experimental results show that the proposed method can restore the color and remove the haze at the same time, and it also outperforms several state-of-the-art methods on three publicly available datasets. Moreover, experiments on an underwater object detection dataset show that the proposed underwater image enhancement method is able to improve the accuracy of the subsequent underwater object detection framework.

Hierarchical Feature Fusion With Mixed Convolution Attention for Single Image Dehazing

Abstract: Single image dehazing, which aims at restoring a haze-free image from its correspondingly unconstrained hazy scene, is a fundamental yet challenging task and has gained immense popularity recently. However, the images recovered by some existing haze-removal methods often contain haze, artifacts, and color distortions, which severely degrade the visual quality and have negative impacts on subsequent computer vision tasks. To this end, we propose a network combining multi-scale hierarchical feature fusion and mixed convolution attention to progressively and adaptively enhance the dehazing performance. The haze levels and image structure information are accurately estimated by fusing multi-scale hierarchical features, thus the model restores images with less remaining haze. The proposed mixed convolution attention mechanism is capable of reducing feature redundancy, learning compact and effective internal representations and highlighting task-relevant features, thus, it can further help the model estimate images with sharper textural details and more vivid colors. Furthermore, a deep semantic loss is also proposed to highlight essential semantic information in deep features. The experimental results show that the proposed method outperforms state-of-the-art haze removal algorithms.

A Scalable Haze‐Free Antireflective Hierarchical Surface with Self‐Cleaning Capability

Abstract: The lotus effect indicates that a superhydrophobic, self‐cleaning surface can be obtained by roughening the topography of a hydrophobic surface. However, attaining high transmittance and clarity through a roughened surface remains challenging because of its strong scattering characteristics. Here, a haze‐free, antireflective superhydrophobic surface that consists of hierarchically designed nanoparticles is demonstrated. Close‐packed, deep‐subwavelength‐scale colloidal silica nanoparticles and their upper, chain‐like fumed silica nanoparticles individually fulfill haze‐free broadband antireflection and self‐cleaning functions. These double‐layered hierarchical surfaces are obtained via a scalable spraying process that permits precise control over the coating morphology to attain the desired optical and wetting properties. They provide a “specular” visible transmittance of >97% when double‐side coated and a record‐high self‐cleaning capability with a near‐zero sliding angle. Self‐cleaning experiments on photovoltaic devices verify that the developed surfaces can significantly enhance power conversion efficiencies and aid in retaining pristine device performance in a dusty environment.

Amplified role of potential HONO sources in O₃ formation in North China Plain during autumn haze aggravating processes

Abstract: Abstract. Co-occurrences of high concentrations of PM2.5 and ozone (O3) have been frequently observed in haze-aggravating processes in the North China Plain (NCP) over the past few years. Higher O3 concentrations on hazy days were hypothesized to be related to nitrous acid (HONO), but the key sources of HONO enhancing O3 during haze-aggravating processes remain unclear. We added six potential HONO sources, i.e., four ground-based (traffic, soil, and indoor emissions, and the NO2 heterogeneous reaction on ground surface (Hetground)) sources, and two aerosol-related (the NO2 heterogeneous reaction on aerosol surfaces (Hetaerosol) and nitrate photolysis (Photnitrate)) sources into the WRF-Chem model and designed 23 simulation scenarios to explore the unclear key sources. The results indicate that ground-based HONO sources producing HONO enhancements showed a rapid decrease with height, while the NO + OH reaction and aerosol-related HONO sources decreased slowly with height. Photnitrate contributions to HONO concentrations were enhanced with aggravated pollution levels. The enhancement of HONO due to Photnitrate on hazy days was about 10 times greater than on clean days and Photnitrate dominated daytime HONO sources (∼ 30 %–70 % when the ratio of the photolysis frequency of nitrate (Jnitrate) to gas nitric acid (JHNO3) equals 30) at higher layers (>800 m). Compared with that on clean days, the Photnitrate contribution to the enhanced daily maximum 8 h averaged (DMA8) O3 was increased by over 1 magnitude during the haze-aggravating process. Photnitrate contributed only ∼ 5 % of the surface HONO in the daytime with a Jnitrate/JHNO3 ratio of 30 but contributed ∼ 30 %–50 % of the enhanced O3 near the surface in NCP on hazy days. Surface O3 was dominated by volatile organic compound-sensitive chemistry, while O3 at higher altitudes (>800 m) was dominated by NOx-sensitive chemistry. Photnitrate had a limited impact on nitrate concentrations (<15 %) even with a Jnitrate/JHNO3 ratio of 120. These results suggest the potential but significant impact of Photnitrate on O3 formation, and that more comprehensive studies on Photnitrate in the atmosphere are still needed.