PM2.5 concentration estimation using convolutional neural network and gradient boosting machine

Surface monitoring, vertical atmospheric column observation, and simulation using chemical transportation models are three dominant approaches for perception of fine particles with diameters less than 2.5 micrometers(PM2.5) concentration. Here we explored an imagebased methodology with a deep learning approach and machine learning approach to extend the ability on PM2.5 perception. Using 6976 images combined with daily weather conditions and hourly time data in Shanghai(2016), trained by hourly surface monitoring concentrations, an end-to-end model consisting of convolutional neural network and gradient boosting machine(GBM) was constructed. The mean absolute error, the root-mean-square error and the R-squared for PM2.5 concentration estimation using our proposed method is 3.56, 10.02, and 0.85 respectively. The transferability analysis showed that networks trained in Shanghai, fine-tuned with only 10% of images in other locations, achieved performances similar to ones from trained on data from target locations themselves. The sensitivity of different regions in the image to PM2.5 concentration was also quantified through the analysis of feature importance in GBM. All the required inputs in this study are commonly available, which greatly improved the accessibility of PM2.5 concentration for placed and period with no surface observation. And this study makes an exploratory attempt on pollution monitoring using graph theory and deep learning approach.

Patterns and trajectories of macrophyte change in East China’s shallow lakes over the past one century

Macrophytes play important roles in maintaining ecosystem health and stability of shallow lakes.Better understanding of their long-term dynamics has important theoretical and practical significance for both lake ecosystem restoration and eutrophication control.However,the knowledge about the historical status and changing patterns of macrophytes in China’s shallow lakes is still controversial and lacks systematic research.Here,we reviewed and synthesized the published records of submerged macrophytes from 14 typical shallow lakes in the eastern plain covering the past 100 years.The results suggest that submerged macrophytes have experienced three clear stages of change:rare period(the 1900s–the 1950s),growth period(the 1950s–the 1980s),and recession period(the 1980s–now).This finding is different from the traditional understanding that submerged macrophytes were abundant in the early 20th century and have been degrading since then.On this basis,we proposed the possible evolution pattern(less-more-less)of submerged macrophytes in the eastern plain lake region over the past 100 years,which provides new perspectives about the long-term evolution process of macrophytes in shallow lakes.Furthermore,we found that the decline of submerged macrophytes during the regime shift shows a gradual process at the interdecadal scale;this finding contradicts the classical regime shift theory that macrophytes decline sharply during the critical transition.This study has important theoretical value for the restoration of the eastern plain lakes in China from“turbid lake”to“clear lake”,especially for establishing the historical reference condition and restoration path of macrophytes.