Zhong-Liang Wang

Greenhouse gases emission from the sewage draining rivers

Carbon dioxide (CO2), methane (CH4) and nitrous oxide (N2O) concentration, saturation and fluxes in rivers (Beitang drainage river, Dagu drainage rive, Duliujianhe river, Yongdingxinhe river and Nanyunhe river) of Tianjin city (Haihe watershed) were investigated during July and October in 2014, and January and April in 2015 by static headspace gas chromatography method and the two-layer model of diffusive gas exchange. The influence of environmental variables on greenhouse gases (GHGs) concentration under the disturbance of anthropogenic activities was discussed by Spearman correlative analysis and multiple stepwise regression analysis. The results showed that the concentration and fluxes of CO2, CH4 and N2O were seasonally variable with >winter>fall>summer, spring>summer>winter>fall and summer>spring>winter>fall for concentrations and spring>summer>fall>winter, spring>summer>winter>fall and summer>spring>fall>winter for fluxes respectively. The GHGs concentration and saturation were higher in comprehensively polluted river sites and lower in lightly polluted river sites. The three GHGs emission fluxes in two sewage draining rivers of Tianjin were clearly higher than those of other rivers (natural rivers) and the spatial variation of CH4 was more obvious than the others. CO2 and N2O air-water interface emission fluxes of the sewage draining rivers in four seasons were about 1.20-2.41 times and 1.13-3.12 times of those in the natural rivers. The CH4 emission fluxes of the sewage draining rivers were 3.09 times in fall to 10.87 times in spring of those in the natural rivers in different season. The wind speed, water temperature and air temperature were related to GHGs concentrations. Nitrate and nitrite (NO3-+NO2--N) and ammonia (NH4+-N) were positively correlated with CO2 concentration and CH4 concentration; and dissolved oxygen (DO) concentration was negatively correlated with CH4 concentration and N2O concentration. The effect of human activities on carbon and nitrogen cycling in river is great.

Application of feature selection and regression models for chlorophyll-a prediction in a shallow lake

As a representative index of the algal bloom, the concentration of chlorophyll-a (Chl-a) is a key parameter of concern for environmental managers. The relationships between environmental variables and Chl-a are complex and difficult to establish. Two machine learning methods, including support vector machine for regression (SVR) and random forest (RF), were used in this study to predict Chl-a concentration based on multiple variables. To improve the model accuracy and reduce the input number, two feature selection methods, including minimum redundancy and maximum relevance method (mRMR) and RF, were integrated with regression models. The results showed that the RF model had a higher predictive ability than the SVR model. Furthermore, the less computational time cost and unnecessary prior data transformation also indicated a better applicability of the RF model. The comparison between ensemble models of mRMR-RF and RF-RF showed that the RF-RF yielded a better performance with fewer variables. Seven variables selected from the candidate predictors could interpret most information, and their potential implications to Chl-a were discussed based on the level of importance. Overall, the RF-RF ensemble model can be considered as a useful approach to determine the significant stressors and achieve satisfactory prediction of Chl-a concentration.