.png){kind=icon}
.png){kind=icon}
Return
Deep learning-based prediction of negative air ion concentrations using TCN-attention and meteo-temporal fusion transformer
.png){kind=icon}
OA
AI
G
M
J
L
DOI：10.3389/fenvs.2026.1823488.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
The concentration of negative air ions (NAI) is a key indicator used for evaluating air quality and quantifying the health benefits that forest ecosystems provide. Accurate prediction of NAI concentration holds significant value for ecological tourism planning as well as public health service improvement. However; existing research has primarily analyzed factors influencing NAI concentration using conventional statistical methods; leaving a gap in fully capturing complex nonlinear interactions among meteorological drivers and in multi-step time series prediction. In this study; based on continuous observation data from 31 environmental monitoring stations in Jiangsu Province from 2021 to 2023; a 1-day prediction model TCN-Attention and a 7-day sliding prediction model MTFT (Meteo-Temporal Fusion Transformer) were developed. TCN-Attention achieved R2=0.917; with MAE = 167.5 ions/cm3; 2.7%–7.8% lower than those of conventional machine learning methods. MTFT achieved an overall R2=0.866 across the 7-day forecast horizon and exceeded conventional methods by 18.6%–29.8% in relative R2; indicating that the advantage of sequence-based temporal modeling increased with forecast horizon. Forecast errors of MTFT were lower in summer and autumn (seasonal MAE = 63.9 and 69.3 ions/cm3) than in winter and spring (83.1 and 85.3 ions/cm3); highlighting the temporal stability of the meteorology–NAI relationship as an important factor in seasonal predictability. To identify the effects of meteorological factors on NAI concentration; a separate SHAP analysis was conducted; revealing the inflection point of temperature effect at 25 °C; the diffusion threshold at wind speed 1.5 m/s; and optimal generation conditions under high temperature and moderate humidity (>25 °C + 70–85% RH). These meteorological response patterns were physically plausible and consistent with the descriptive results. Under the four representative meteorological scenarios; MTFT yielded MAE values of 59.5–82.1 ions/cm3; comparable in magnitude to its overall test MAE of 73.4 ions/cm3; with performance remaining comparable across scenarios. The dual-model prediction system constructed in this study has already been applied to the meteorological operational system of Jiangsu Province.
Keywords:
deep learning
time series prediction
SHAP interpretability
negative air ions
meteorological control mechanisms
Journal
IF:
3.7
Papers:
7.9K
Citations:
2.3W
Organization
Cited Papers