Spatiotemporal Variability and Influencing Factors of Aerosol Optical Depth over the Pan Yangtze River Delta during the 2014–2017 Period

Abstract

Large amounts of aerosol particles suspended in the atmosphere pose a serious challenge to the climate and human health. In this study, we produced a dataset through merging the Moderate Resolution Imaging Spectrometers (MODIS) Collection 6.1 3-km resolution Dark Target aerosol optical depth (DT AOD) with the 10-km resolution Deep Blue aerosol optical depth (DB AOD) data by linear regression and made use of it to unravel the spatiotemporal characteristics of aerosols over the Pan Yangtze River Delta (PYRD) region from 2014 to 2017. Then, the geographical detector method and multiple linear regression analysis were employed to investigate the contributions of influencing factors. Results indicate that: (1) compared to the original Terra DT and Aqua DT AOD data, the average daily spatial coverage of the merged AOD data increased by 94% and 132%, respectively; (2) the values of four-year average AOD were high in the north-east and low in the south-west of the PYRD; (3) the annual average AOD showed a decreasing trend from 2014 to 2017 while the seasonal average AOD reached its maximum in spring; and that (4) Digital Elevation Model (DEM) and slope contributed most to the spatial distribution of AOD, followed by precipitation and population density. Our study highlights the spatiotemporal variability of aerosol optical depth and the contributions of different factors over this large geographical area in the four-year period, and can, therefore, provide useful insights into the air pollution control for decision makers.

Estimating Urban Vegetation Biomass from Sentinel-2A Image Data

“…As it is likely that collinearity exists in the predictive variables, the variance inflation factor (VIF) [57,61] is used to examine it in this study: Although 192 vegetation quadrats were initially selected, only 140 quadrats of them (shown in Figure 1) were visited and investigated in practice-because some of the pre-selected quadrats were not accessible for various reasons (e.g., physical barriers and refusal to access). Among the 140 quadrats were 35 dominated by coniferous forest, 73 by broadleaved forest, and 32 by low vegetation.…”

“…As it is likely that collinearity exists in the predictive variables, the variance inflation factor (VIF) [57,61] is used to examine it in this study:…”

Modeling Soil Moisture from Multisource Data by Stepwise Multilinear Regression: An Application to the Chinese Loess Plateau

“…Regarding PRE, daily PRE with cumulative values were also rescaled to monthly, seasonal, and annual temporal resolutions. After interpolation by inverse distance weighting, monthly, seasonal, and annual PRE maps were generated [18,109]. The maximum and minimum total annual PRE in 2017 was 1020.15 mm and 99.80 mm, respectively (Figure 2b).…”

“…where R i represents the correlation coefficient between the i th predictive variable and the remaining predictive variables. No multicollinearity exists if VIF is less than 3 [109,113].…”

Handling Missing Data in Large-Scale MODIS AOD Products Using a Two-Step Model

“…For example, the shortcomings of the DT algorithm and DB algorithm for AOD detection in bright areas, the errors of cloud detection in some heavily polluted areas and the degradation of other sensors directly affect the detection of dark pixels in low angle areas, which leads to the loss of AOD data in some areas [26,27]. A study of the Yangtze River Delta in China found that the missing rate of MOD AOD reached 89.6% between 2014 and 2017 [28]. Because the results of AOD are affected by meteorological conditions, human activities and vegetation coverage, it is difficult to ensure the accuracy of the AOD restoration [29].…”