Estimating monthly wet sulfur (S) deposition flux over China using an ensemble model of improved machine learning and geostatistical approach
Estimating monthly wet sulfur (S) deposition flux over China using an ensemble model of improved machine learning and geostatistical approach
复制标题
使用改进的机器学习和地统计方法的集成模型估算中国每月的湿硫 (S) 沉降通量
DOI：
10.1016/j.atmosenv.2019.116884
复制
发表时间：
2019
影响因子：
5
通讯作者：
Fu Hongbo
中科院分区：
文献类型：
--
作者：
Li Rui;Cui Lulu;Zhao Yilong;Meng Ya;Kong Wang;Fu Hongbo
湿硫沉降对土壤酸化、生物多样性丧失和全球气候变化具有负面影响，因此被认为是一个关键问题。然而，有限的地面监测站点，很难完全澄清湿S沉降在中国的时空变化。因此，基于排放清单、气象因子和其他地理协变量，建立了一种改进的机器学习和地统计学方法的集成模型--果蝇优化算法-随机森林-时空克里格（FOA-RF-STK）模型，用于估算全国范围内的硫沉降。与原始RF模式（R2 = 0.52，RMSE = 8.99 kg ha− 1 yr −1）相比，集合模式能更好地捕捉预报因子与S沉降通量之间的关系（R2 = 0.68，RMSE = 7.51 kg ha − 1 yr − 1）。基于改进的模型，预测了S沉降通量的最高值和最低值主要集中在中国东南部（69.57 kg S ha− 1 yr −1）和内蒙古（42.37 kg S ha− 1 yr −1）。湿S沉降通量具有明显的季节变化，夏季最高（22.22 kg S ha− 1 sea −1），秋季次之（18.30 kg S ha− 1 sea −1），春季最低（16.27 kg S ha− 1 sea −1），冬季最低（14.71 kg S ha− 1 sea −1），与降水量密切相关。该研究为全国尺度的硫沉降估算提供了一种新的方法。
The wet S deposition was treated as a key issue because it played the negative on the soil acidification, biodiversity loss, and global climate change. However, the limited ground-level monitoring sites make it difficult to fully clarify the spatiotemporal variations of wet S deposition over China. Therefore, an ensemble model of improved machine learning and geostatistical method named fruit fly optimization algorithm-random forest-spatiotemporal Kriging (FOA-RF-STK) model was developed to estimate the nationwide S deposition based on the emission inventory, meteorological factors, and other geographical covariates. The ensemble model can capture the relationship between predictors and S deposition flux with the better performance (R2= 0.68, root mean square error (RMSE) = 7.51 kg ha−1yr−1) compared with the original RF model (R2= 0.52, RMSE = 8.99 kg ha−1yr−1). Based on the improved model, it predicted that the highest and lowest S deposition flux were mainly concentrated on the Southeast China (69.57 kg S ha−1yr−1) and Inner Mongolia (42.37 kg S ha−1yr−1), respectively. The estimated wet S deposition flux displayed the remarkably seasonal variation with the highest value in summer (22.22 kg S ha−1sea−1), follwed by ones in autumn (18.30 kg S ha−1sea−1), spring (16.27 kg S ha−1sea−1), and the lowest one in winter (14.71 kg S ha−1sea−1), which was closely associated with the rainfall amounts. The study provides a novel approach for the S deposition estimation at a national scale.