Atmospheric PM_2.5 is one of the primary pollutants affecting air quality. Therefore， how to effectively monitor and manage PM_2.5 concentrations is of great significance to the sustainable development of ecological quality in China. Based on a series of auxiliary parameters， i.e.， Top-of-Atmospheric reflectance （derived from remote sensing imageries）， meteorology， and land use， a Random Forest （RF） model was developed to estimate ground-level PM_2.5 concentrations in the contiguous Yangtze River Delta-Fujian （YRD-FJ） region located in East China in 2016， 2018 and 2020. The correlation between the spatial distribution of PM_2.5 concentrations and landscape patterns in YRD-FJ region using 3-period land classification data was carried out. The results show that （1） the R² between the PM_2.5 concentrations estimated by the RF model and the ground-level measured values in YRD-FJ region in 2016， 2018， and 2020 are 0.91， 0.89， and 0.90， respectively； the RMSE are 9.07、10.19 and 8.03 μg/m³， respectively. （2） The annual average PM_2.5 concentrations in YRD-FJ region showed a trend of year-on-year decrease from 2016 to 2020， and its spatial distribution was generally in the pattern of "Jiangsu > Shanghai > Zhejiang > Fujian". （3） Reasonable control of the landscape proportion of cropland， built-up land and water bodies， and reduction of their landscape dominance and edge density are conducive to alleviating the annual average PM_2.5 concentrations. Additionally， appropriate increase in forest occupancy， edge density， and shape complexity are beneficial to reducing PM_2.5 concentrations. Our results could provide the scientific basis and decision-making reference for the control of regional air pollution and landscape pattern planning.