Long-term trends of ambient nitrate (NO 3) concentrations across China based on ensemble machine-learning models

Li, Rui; Cui, Lulu; Zhao, Yilong; Zhou, Wenhui; Fu, Hongbo

page2148High loadings of nitrate (inline-formula M3inlinescrollmathml chem normal NO normal 3 - 25pt16ptsvg-formulamathimga186e28964d6ae507e65dbc91f8b1f71 essd-13-2147-2021-ie00004.svg25pt16ptessd-13-2147-2021-ie00004.png ) in the aerosol over China significantly exacerbate the air quality and pose a great threat to ecosystem safety through dry–wet deposition. Unfortunately, limited ground-level observation data make it challenging to fully reflect the spatial pattern of inline-formula M4inlinescrollmathml chem normal NO normal 3 - 25pt16ptsvg-formulamathimge16cba38499a6a16cb1a10e488ec56da essd-13-2147-2021-ie00005.svg25pt16ptessd-13-2147-2021-ie00005.png levels across China. Until now, long-term monthly particulate inline-formula M5inlinescrollmathml chem normal NO normal 3 - 25pt16ptsvg-formulamathimg91b2e19ca239409a7665981c17575147 essd-13-2147-2021-ie00006.svg25pt16ptessd-13-2147-2021-ie00006.png datasets at a high resolution were still missing, which restricted the assessment of human health and ecosystem safety. Therefore, a unique monthly inline-formula M6inlinescrollmathml chem normal NO normal 3 - 25pt16ptsvg-formulamathimga33a7d42b70ca1fe513ac92c5832eec2 essd-13-2147-2021-ie00007.svg25pt16ptessd-13-2147-2021-ie00007.png dataset at 0.25inline-formula resolution over China during 2005–2015 was developed by assimilating surface observations, satellite products, meteorological data, land use types and other covariates using an ensemble model combining random forest (RF), gradient-boosting decision tree (GBDT), and extreme gradient-boosting (XGBoost) methods. The new developed product featured an excellent cross-validation inline-formulaR2 value (0.78) and relatively lower root-mean-square error (RMSE: 1.19 inline-formulaµg N m−3) and mean absolute error (MAE: 0.81 inline-formulaµg N m−3). Besides, the dataset also exhibited relatively robust performance at the spatial and temporal scales. Moreover, the dataset displayed good agreement with (inline-formulaR2=0.85, inline-formulaRMSE=0.74inline-formulaµg N m−3, and inline-formulaMAE=0.55inline-formulaµg N m−3) some unlearned data collected from previous studies. The spatiotemporal variations in the developed product were also shown. The estimated inline-formula M16inlinescrollmathml chem normal NO normal 3 - 25pt16ptsvg-formulamathimg48a6d5724cc017ced9c974ab9a81c03a essd-13-2147-2021-ie00008.svg25pt16ptessd-13-2147-2021-ie00008.png concentration showed the highest value in the North China Plain (NCP) (inline-formula3.55±1.25inline-formulaµg N m−3); followed by the Yangtze River Delta (YRD) (inline-formula2.56±1.12inline-formulaµg N m−3), Pearl River Delta (PRD) (inline-formula1.68±0.81inline-formulaµg N m−3), and Sichuan Basin (inline-formula1.53±0.63inline-formulaµg N m−3), and the lowest one in the Tibetan Plateau (inline-formula0.42±0.25inline-formulaµg N m−3). The higher ambient inline-formula M27inlinescrollmathml chem normal NO normal 3 - 25pt16ptsvg-formulamathimg54a63b90f99919b7f33388d68cde2f58 essd-13-2147-2021-ie00009.svg25pt16ptessd-13-2147-2021-ie00009.png concentrations in the NCP, YRD, and PRD were closely linked to the dense anthropogenic emissions. Apart from the intensive human activities, poor terrain condition might be a key factor for the serious inline-formula M28inlinescrollmathml chem normal NO normal 3 - 25pt16ptsvg-formulamathimgda99f0b0265c157ce21f9580f34f8fd2 essd-13-2147-2021-ie00010.svg25pt16ptessd-13-2147-2021-ie00010.png pollution in the Sichuan Basin. The lowest ambient inline-formula M29inlinescrollmathml chem normal NO normal 3 - 25pt16ptsvg-formulamathimgf406d9210c9988b6f1f99fbfd13290fc essd-13-2147-2021-ie00011.svg25pt16ptessd-13-2147-2021-ie00011.png concentration in the Tibetan Plateau was contributed by the scarce anthropogenic emission and favourable meteorological factors (e.g. high wind speed). In addition, the ambient inline-formula M30inlinescrollmathml chem normal NO normal 3 - 25pt16ptsvg-formulamathimgcbd0aa2bda73584a7a23dafea6b5761c essd-13-2147-2021-ie00012.svg25pt16ptessd-13-2147-2021-ie00012.png concentration showed a marked increasing tendency of 0.10 inline-formula M31inlinescrollmathml unit normal µ normal g nobreak0.125em normal N nobreak0.125em normal m - normal 3 nobreak0.125em normal yr - normal 1 65pt15ptsvg-formulamathimgcd87d4cff21e042dd45f82a3e3f6c046 essd-13-2147-2021-ie00013.svg65pt15ptessd-13-2147-2021-ie00013.png during 2005–2014 (inline-formulap<0.05), while it decreased sharply from 2014 to 2015 at a rate of inline-formula−0.40inline-formula M34inlinescrollmathml unit normal µ normal g 0.125emnobreak normal N 0.125emnobreak normal m - normal 3 nobreak0.125em normal yr - normal 1 65pt15ptsvg-formulamathimgf4285f72b01c6cc72cfbb9deae252f2e essd-13-2147-2021-ie00014.svg65pt15ptessd-13-2147-2021-ie00014.png (inline-formulap<0.05). The ambient inline-formula M36inlinescrollmathml chem normal NO normal 3 - 25pt16ptsvg-formulamathimg4d13138f69382cea7eca33a7bac51ba9 essd-13-2147-2021-ie00015.svg25pt16ptessd-13-2147-2021-ie00015.png levels in Beijing–Tianjin–Hebei (BTH), YRD, and PRD displayed gradual increases at a rate of 0.20, 0.11, and 0.05 inline-formula M37inlinescrollmathml unit normal µ normal g nobreak0.125em normal N nobreak0.125em normal m - normal 3 nobreak0.125em normal yr - normal 1 65pt15ptsvg-formulamathimgab9d013c508277983daa9c60aa504d33 essd-13-2147-2021-ie00016.svg65pt15ptessd-13-2147-2021-ie00016.png (inline-formulap<0.05) during 2005–2013, respectively. The gradual increases in inline-formula M39inlinescrollmathml chem normal NO normal 3 - 25pt16ptsvg-formulamathimg0ed2522f87147883b53f48851745fd62 essd-13-2147-2021-ie00017.svg25pt16ptessd-13-2147-2021-ie00017.png concentrations in these regions from 2005 to 2013 were due to the fact that the emission reduction measures during this period focused on the reduction of inline-formulaSO2 emission rather than inline-formulaNOx emission and the rapid increase in energy consumption. Afterwards, the government further strengthened these emission reduction measures and thus caused the dramatic decreases in inline-formula M42inlinescrollmathml chem normal NO normal 3 - 25pt16ptsvg-formulamathimg742f2a785243506db9e8d8e9964694e9 essd-13-2147-2021-ie00018.svg25pt16ptessd-13-2147-2021-ie00018.png concentrations in these regions from 2013 to 2015 (inline-formulap<0.05). The long-term inline-formula M44inlinescrollmathml chem normal NO normal 3 - 25pt16ptsvg-formulamathimg6c9aef46e63985e949f52ba810262b4b essd-13-2147-2021-ie00019.svg25pt16ptessd-13-2147-2021-ie00019.png dataset over China could greatly deepen the knowledge about the impacts of emission reduction measures on air quality improvement. The monthly particulate inline-formula M45inlinescrollmathml chem normal NO normal 3 - 25pt16ptsvg-formulamathimgafc29ccf929971927e195b135fd5e45c essd-13-2147-2021-ie00020.svg25pt16ptessd-13-2147-2021-ie00020.png levels over China during 2005–2015 are open access at https://doi.org/10.5281/zenodo.3988307https://doi.org/10.5281/zenodo.3988307 (Li et al., 2020c).



Li, Rui / Cui, Lulu / Zhao, Yilong / et al: Long-term trends of ambient nitrate (NO3−) concentrations across China based on ensemble machine-learning models. 2021. Copernicus Publications.


12 Monate:

Grafik öffnen


Rechteinhaber: Rui Li et al.

Nutzung und Vervielfältigung: