? However, in our opinion, there is no need to worry about the above situation. A long column of smoke appeared, and its end spread out in the cloud, just like a fly swatter. This meteorological condition is neutral atmosphere (no inversion stratification)+clear clouds. The so-called neutral atmosphere is well understood, which means that the pollutants emitted by overhead chimneys diffuse slowly in the vertical direction, but rapidly in the horizontal direction. In other words, in this case, the smoke column went to other places in the sky, and the disaster was sucked to the east and blown to the ocean (as shown below), instead of spreading to the ground, which was suitable for the dissipation of pollutants. So people who live on the ground need not worry. However, everyone who has studied physics in senior high school knows that matter and energy are conserved. Pollutants don't go into space, they still exist in the world. Contaminants are first removed in clouds or plumes. Simply put, water droplets collide with and devour pollutants emitted by industries with smaller diameters, just like giant white blood cells devour bacteria, and then return to the earth through the rainfall process. The final effect is to crush the suspended air pollutants that are easily inhaled into the human body to the ground with three fists and two feet.
? "Bodhisattva fears cause, ordinary people fear fruit". It is very difficult to convict and sentence the emission source only through observation experiments, because the determination of causality requires very strict reasoning proof. Deductive reasoning should be used without data and background conditions, that is, preliminary conclusions should be drawn according to assumptions and existing rules. We know that the pollutants in the air are mainly concentrated below 2000 meters, but in the vertical direction, the boundary between polluted air and clean air is changing, which is called the atmospheric boundary layer. Assuming that sunshine is a typical land-sea atmospheric boundary layer, according to the stratification of the atmosphere by the second generation air quality model ROM, it can be divided into:
? Although boundary layer and mixed layer are not synonyms, there is a strong convective boundary layer, and the air is violently mixed upwards, often forming a mixed layer. It should be noted that although coastal areas such as Rizhao belong to sea fog-prone areas with obvious characteristics of land-sea wind circulation, generally speaking, the characteristics of land-sea wind circulation in mid-latitude areas are only prominent in summer. Moreover, the circulation of land-sea wind is weak and the wind force is small, which is often covered by large-scale circulation such as northeast cold vortex in the south and warm and humid airflow in the southeast/southwest of the north. Therefore, the ocean layer is sometimes not obvious.
? When you have certain background information, you should use inductive reasoning to sum up the preliminary rules. When the height of the mixing layer is higher than that of the boundary layer affected by surface friction, all air masses in the turbulent boundary layer are uniformly mixed. At this time, it is assumed that the quality of pollutants discharged from the ground remains unchanged. From the relationship between density and volume, we can know that the decrease of boundary layer height will be accompanied by the accumulation of air pollutant concentration, and the ground PM 2.5 concentration will increase linearly. As shown in the following figure, the abscissa is the boundary layer height BLH, and the ordinate is the particle concentration. It can be seen that when the height of the boundary layer is less than 500 meters, the concentration of PM 2.5 begins to increase linearly. Summarize the boundary layer heights from 20 13 to 20 19 in Rizhao area, as shown in the following figure. The abscissa is the boundary layer height BLH, and the ordinate is the particle concentration. It can be seen that when the height of the boundary layer is lower than 500m, the concentration of PM 2.5 begins to increase linearly. Therefore, the height of the mixed layer in Rizhao city in winter is about 500 meters, which basically conforms to our premise assumption.
? In order to further draw a preliminary conclusion, we made probability distribution statistics of PM 2.5 concentration and ozone O 3 concentration in 20 13-20 19 in Rizhao City. As can be seen from the following figure, the concentration of PM 2.5 is a typical unimodal Gaussian normal distribution, but there is a high-value long tail phenomenon, while the ozone concentration is a non-normal bimodal distribution. The bimodal distribution of ozone shows that part of the formation process of ozone near the ground has nothing to do with photochemical reaction, but there is downward transport at high altitude, mainly before noon 12. The concentration source of fine particles is mainly linear near-surface accumulation, as mentioned above, but there is also a nonlinear explosive growth. For details, please refer to the above article of the author:/p/19085537b21e.
? The height of industrial chimneys in China is generally more than 200 meters, so the length of the plume itself is enough to break through the mixed layer and spread evenly to the outside atmosphere. However, the shape of the plume itself is changeable, which is related to the complex meteorological conditions at high altitude. Even due to the existence of local small-scale turbulence vortex, the length of smoke plume emitted by two adjacent chimneys is not consistent. We summarized the shapes of smoke plumes from various high-altitude sources, and found that the following conditions are clean weather: first, the horizontal wind speed is small, but the vertical turbulence exchange is strong, as shown in the first picture below; The other is a neutral atmosphere with high horizontal wind speed and weak vertical exchange, as shown in Figure 2 below.
? After preliminary deduction and induction. We finally decided to use image method to infer the relationship between industrial smoke flow from overhead sources and ground air pollution. When the emission source is strong, the concentration of air pollutants is sensitive to meteorological factors, that is, scarcity factor control. The meteorological factors of land-sea boundary layer are complex, so it is very difficult to accurately observe the meteorological conditions at high altitude, which is time-consuming and laborious, and plume profile is a good meteorological index.
? In order to capture the complex changes of high-altitude meteorological conditions, we use a timed automatic camera to take photos every half hour. As can be seen from the above, the outline of smoke plume can reflect complex meteorological changes, so we take the outline of smoke plume as an indicator of meteorological conditions. Because of the existence of low clouds, the automatic recognition effect of smoke plume contour in images is not ideal, and image enhancement and manual labeling are needed. The defogging algorithm used in image enhancement was developed by He, a rare artificial intelligence genius. The shape of plume changes rapidly, so the polygon length of plume section, that is, diagonal length, is used to describe its profile characteristics. It should be noted that the actual diffusion range of atmospheric particles may far exceed its visible profile, because the smoke plume profile seen by naked eyes is mainly due to the condensation of water vapor.
? The factors that control the plume length are very complicated, such as wind speed, wind direction, humidity, clouds and so on. As can be seen from the following figure, the ground particle concentration is negatively correlated with the plume profile length, but it is nonlinear. Weather system is a typical chaotic system, and the nonlinear characteristics in complex systems will only appear when certain conditions are met. Therefore, we speculate that the relationship between ground particle concentration and plume profile should be classified.
? We use a logical function to fit the particle concentration on the ground and the polygon length of the plume profile. Although logistic regression is called regression model, it is a binary classification algorithm. As shown in the figure below, where the chimney height (h) is marked in the image, the calculation shows that H=385 (pixel point). According to the national standard, the height of the chimney is 200 meters. Taking 2 times the height of the chimney as the critical point, it can be seen that when the polygon length of the smoke flow profile is greater than 2 times the height of the chimney, the ground particle concentration will be affected by the diffusion conditions. When the plume length is less than twice the height of the chimney, the influence of diffusion on local particulate matter removal is not obvious.
? We obtained the formula of ground particle concentration y and plume length x:
? When the plume length is less than 400 meters, it can be seen that the above formula can not describe the relationship between it and PM2.5 well. Therefore, we consider the meteorological factors at different heights and use the decision tree model for fitting. The following figure shows the decision tree model of ground PM 2.5 concentration at Shangang Station from 2065438 to March 2009. It can be seen that the zonal wind (U_850) and the plume length (PL) of the tropopause (about 850 hectopascals1500m) are the statistical decisive factors of the decision tree.
? Even if the decision tree has only six layers, it is still very complicated to state and explain. But in fact, we only need to care about the two trunks of the decision tree, that is, the decision branches when the ground PM 2.5 concentration is the highest and the lowest. The red circle area in the above figure is the decision branch when the concentration of PM 2.5 is the highest: U _ 850-5.4/s → PL > 796 pixels → PM2.5 <; 67 ? Square meters. It can be inferred that when the shore wind of free air above the height of1500m is more than 5 m/s (usually large-scale southeast warm and humid airflow) and the smoke plume is short (300-500m), the horizontal and vertical diffusion is weak, and the ground particles accumulate, which is prone to heavy pollution of PM2.5..
? In addition to the mass concentration of particulate matter, the number concentration of particulate matter is also a great factor affecting human health. According to the mass formula of spherical particles m=ρV= 4ρR 3 /3, it can be known that the mass of particles below 2500 nm is more than ten thousand times that of particles below 50 nm 1 10,000. Particles below 500 nm account for about 80% of the total particles and can penetrate alveolar membrane; While micron-sized particles account for about 90% by mass.
? We observed the quantitative concentration of particulate matter in rural areas of Rizhao city. When the air in the field is cleanest, the particle concentration is less than 3000 /cm 3. However, in some cases, the concentration of particulate matter is greater than 10000 /cm 3. When there are the most particles, it is a blue sky that looks good. A large number of new nano-sized particles are produced and increased, but the concentration of PM 2.5 has not changed significantly.
? On the side of urban roads, affected by the traffic of motor vehicles, the concentration of particulate matter can reach 60000 /cm 3, which is six times that of the dirtiest time in rural areas. However, the concentration of particulate matter in the industrial plume over 300 meters above the city is less than 3000 /cm 3, which is equivalent to the concentration when the outdoor air in rural areas is clean. As for the indoor air, an office of more than ten square meters is about 7000 /cm 3, which varies with the number of people. Seeing the early warning of mobile phones, the air quality index AQI exceeded 100. Although the concentration of atmospheric particulate matter is high, it is basically large particles, only a few thousand per cubic centimeter. When you see a clean blue sky, the concentration of particulate matter may reach tens of thousands per cubic centimeter.
? The reason why we can't keep our original fields is only because we are not modern enough. The roughness of industry will inevitably make us lose both field and modernity. Observing scientific experiments should gradually shift from cities to fields and extend from the center to the edge.