Ozone has become one of the most serious air pollutants in China in recent years. Since people spend most of their time indoors, the ozone in the indoor environment could be a major factor affecting the occupants’ health. The indoor ozone in residential buildings mainly comes from two sources: outdoor atmosphere and indoor ozone produced by electrical devices. In this study, a typical residence in Nanjing was taken as an example to calculate and compare the contributions of indoor and outdoor sources to ozone in the building. A questionnaire survey about the type, the placement, and the frequency of use of the ozone emission devices was performed to provide the basis for the settings of indoor ozone sources. The multi-zone software CONTAM was used hourly to simulate the ozone concentration in summer and in winter with inner doors either closed or open, and it was noted whether there were ozone emission devices indoors or not. Source contribution was quantified and compared by three methods in this paper: (1) the average indoor/outdoor (I/O) ratio, (2) the I/O ratio frequency, and (3) the ratio of indoor ozone concentration without ozone sources to that with ozone sources. The results showed that the contribution of outdoor sources was much greater than that of indoor sources in summer, but in winter, the frequency of I/O > 1 could reach 55.8% of the total seasonal time, and the ratio of indoor ozone concentration without sources to that with sources could reach as high as 74.3%. This meant that the indoor concentration had the potential to exceed the outdoor. Furthermore, human respiratory exposure in different ages and genders was calculated. It was found that teenagers aged 10–18 years old and female adults had a higher respiratory exposure level.

The accurate identification of the characteristics of pollutant sources can effectively prevent the loss of human life and property damage caused by the sudden release of harmful chemicals in emergency situations. Machine learning algorithms, artificial neural network (ANN), support vector machine (SVM), k-nearest neighbour (KNN) and naive Bayesian (NB) classification can be used to identify the location of pollutant sources with limited sensor data inputs. In this study, the identification accuracy of the four above-mentioned machine learning algorithms was investigated and compared, considering the different sensor layouts, eigenvector inputs, meteorological parameters and number of samples. The results show that the collection of pollutant concentrations over an extended period of time could improve identification accuracy. Additional sensors were required to reach the same identification accuracy after the introduction of distributed meteorological parameters. Increasing the number of trained samples by a factor of five improved the identification accuracy of KNN by 22% and that of SVM by 1.7%; however, ANN and NB classification remained basically unchanged. When identifying the release mass of the pollutant source, multiple linear, ANN and SVM regression models were adopted. Results show that ANN performs best, whereas SVM provides the least optimal performance.

Many studies have investigated subjective responses to noise, but few concerned about the influence of age on the annoyance (discomfort) caused by noise. It is difficult to get a quantitative model featuring the relationship between noise-induced annoyance and age from one or several laboratory studies due to relatively small samples and limited age groups. This paper investigated recent studies (published after the year 2000) on noise-induced annoyance by the literature review method. We classified the studies according to their employed noise types and summarized the quantified subjective values and the ranges of age. The quantitative values of annoyance obtained from variable rating scales were transferred to a uniform scale and normalized. A probability density function then figured out the corresponding annoyance of a certain age under the small sample -distribution assumption. A predicting model of noise-induced annoyance from the age of 7-55 was proposed, which fitted previous data well.

At present, the raindrop size distribution (DSD) fitting method and traditional neural network method are the main quantitative precipitation estimation (QPE) algorithms. The former is not enough to express the spatiotemporal variability of the DSDs using generalized $Z-R$ parameter relationship, and the latter is limited by the number of network layers, network structure and computing power. Their performance cannot both meet actual needs. In this research, we propose an alternative approach to dual-polarization radar (DPR) QPE. In particular, a radar-gauges dataset (RGD) is constructed for QPE based deep learning using radar raw data and gauges measurements for landfall typhoon in South China, and a deep learning model is designed and trained using this dataset. The model is applied to radar data to produce rainfall estimates. Preliminary results show the promising performance of this novel method compared to traditional QPE estimators.