Journal of Atmospheric and Solar-Terrestrial Physics最新文献

{"title":"A deep learning approach via multifractal detrended fluctuation analysis for PM2.5 prediction","authors":"Unjin Pak ,&nbsp;Ho Kim ,&nbsp;UnHui Jong ,&nbsp;RiGuang Hyon ,&nbsp;JangHak Kim ,&nbsp;Kukchol Kim ,&nbsp;Kwangho Kim","doi":"10.1016/j.jastp.2025.106444","DOIUrl":"10.1016/j.jastp.2025.106444","url":null,"abstract":"<div><div>Accurate prediction of PM 2.5 concentrations in Asian regions with air pollution as a hot topic is still considered as a strong premise for the public to take proactive measures for effective control of the adverse effects of PM 2.5. In this study, a hybrid model consisting of convolutional neural network (CNN)-long short-term (LSTM) with multifractal detrended fluctuation analysis (MF-DFA) for air pollution and meteorological time series data was proposed, and it was used to predict the next day's 24-h average PM 2.5 concentration in Beijing City. MF-DFA was employed to identify the fractal characteristics of the measured data series; it has been considered for the historical data of the target monitoring station and, as a result, enabling CNN-LSTM model to adopt valuable input data. CNN of the hybrid model efficiently extracted the inherent features relevant to PM 2.5 in the input data, and LSTM fully conveyed the historical information of the time series data. Air quality and meteorological data from January 1st, 2015 to February 24th, 2018 were adopted. Through the comparison of the performance indices of the proposed model with MLP and LSTM models, the proposed model showed better prediction accuracy. When the integrated data (RMSE 11.732, MAE 5.18, MAPE 0.131) were used as input of the model, it is guaranteed that the performance of the model is improved compared to the cases of using air quality data (RMSE 11.886, MAE 6.17, MAPE 0.272) and meteorological data (RMSE 29.673, MAE 25.171, MAPE 0.425) as input, respectively.</div></div>","PeriodicalId":15096,"journal":{"name":"Journal of Atmospheric and Solar-Terrestrial Physics","volume":"268 ","pages":"Article 106444"},"PeriodicalIF":1.8,"publicationDate":"2025-01-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143387403","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Large area ionospheric D region height determination method based on lightning electromagnetic pulse","authors":"Jiaquan Wang ,&nbsp;Lihua Shi ,&nbsp;Xiao Zhou ,&nbsp;Fang Xiao ,&nbsp;Shangbo Yuan ,&nbsp;Jiajun Song ,&nbsp;Qiming Ma ,&nbsp;Chaoyi Hu","doi":"10.1016/j.jastp.2025.106446","DOIUrl":"10.1016/j.jastp.2025.106446","url":null,"abstract":"<div><div>This study investigates the method of determining the height of the ionospheric D region over large areas using lightning electromagnetic pulse (LEMP) signals. Utilizing the LEMP waveforms and lightning location data from the Institute of Electrical Engineering, Chinese Academy of Sciences (IEECAS), and fully leveraging both cloud-to-ground (CG) lightning and narrow bipolar pulses (NB), a deep-learning-based method for LEMP detection and correction of ionospheric D region height calculation errors has been proposed. This approach enables the computation of the D region height in real time. For the first time, this research has obtained real-time data on the D region height for China and its surrounding areas. Also, variations in the D region height during solar flares have been observed. The results indicate that the method presented in this paper provides novel data support for the investigation of the ionosphere and illustrates the potential utility of lightning in probing solar activity.</div></div>","PeriodicalId":15096,"journal":{"name":"Journal of Atmospheric and Solar-Terrestrial Physics","volume":"268 ","pages":"Article 106446"},"PeriodicalIF":1.8,"publicationDate":"2025-01-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143176801","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Application of machine learning models for rainfall prediction and estimation of rain-induced attenuation for satellite communication in a tropical region","authors":"M.A. Sodunke ,&nbsp;J.S. Ojo ,&nbsp;Y.B. Lawal ,&nbsp;O.L. Ojo ,&nbsp;G.A. Owolabi ,&nbsp;A.I. Olateju","doi":"10.1016/j.jastp.2025.106443","DOIUrl":"10.1016/j.jastp.2025.106443","url":null,"abstract":"<div><div>The increasing demand for advanced technological equipment and high-capacity data transfer with enhanced signal quality necessitates the transition to higher frequency bands, particularly those beyond the congested Ku and Ka bands (11–40 GHz). Higher frequency signals offer longer travel distances, increased information-carrying capacity, reduced interference, and high data rates, crucial for meeting the growing demand for wireless communication services. However, these frequencies are susceptible to rain-induced attenuation, particularly in tropical regions like Nigeria, where precipitation significantly degrades signal quality and impacts satellite communication services. This study analyzes 50 years of monthly rainfall data (1962–2011) obtained from the Nigerian Meteorological Agency (NiMet) for five Nigerian locations: Port Harcourt, Calabar, Enugu, Ibadan, and Ikeja. The data was divided into five decades using a decadal scaling technique to investigate the long-term trends of precipitation effects on radio wave propagation. The Coefficient of Variation (CV) and Chebil's model were employed to analyze rainfall variability and deduce 1-min rainfall rates, respectively. CV was found to be predominant during July and August across the decades, indicating significant worst months for signal qualities over the region. The seasonal variation of rainfall rates was also examined, showing significant effects.SARIMA, ARIMA, Random Forest and SVM forecasting models have been tested toward the prediction of rainfall in Nigeria. The SVM performed better than the other tested models with the highest adjusted R² and least values of RMSE and MAE.The International Telecommunication Union's model (ITU-R P.618–14) was used to generate rain attenuation at Ku-band (12.275) GHz for the study locations. The findings confirm the reliability of Chebil's 1-min integration time rainfall rate model and validate decadal scaling as an effective method for developing mitigation measure for signal fading.</div></div>","PeriodicalId":15096,"journal":{"name":"Journal of Atmospheric and Solar-Terrestrial Physics","volume":"268 ","pages":"Article 106443"},"PeriodicalIF":1.8,"publicationDate":"2025-01-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143177673","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Low-latitude sub-ionospheric VLF radio signal disturbances due to solar flares: Effects on the attenuation and phase velocities of the waveguide modes","authors":"Kheyali Barman ,&nbsp;Supratick Adhikary ,&nbsp;Bakul Das ,&nbsp;Sujay Pal ,&nbsp;Prabir Kumar Haldar","doi":"10.1016/j.jastp.2025.106433","DOIUrl":"10.1016/j.jastp.2025.106433","url":null,"abstract":"<div><div>Solar flares are sudden bursts of X-rays and UV rays emitted from coronal magnetic loops in active regions near sunspots on the sun’s surface. Soft X-rays below 1 nm penetrate the D-region ionosphere, causing excess ionization and altering its conductivity profile. This study examines solar flares (C-class and M-class) from the 25th solar cycle using two Very Low Frequency (VLF, 3–30 kHz) radio receivers in the low-latitude Indian region, located in Cooch Behar (CHB) and Kolkata (CUB). The work focuses on the Indian VLF transmitter VTX at a frequency of 18.2 kHz to study the effects of solar flares on the D-region ionosphere and VLF signal propagation characteristics in the earth-ionosphere waveguide. The solar zenith angle at the CUB station has a significant impact on the magnitude of VLF amplitude disruption caused by solar flares, showing a positive correlation (r = +0.82, r = +0.61) with flare power during low and high solar activity, respectively. In contrast, CHB exhibits both positive and negative amplitude perturbations, with a negative correlation (r = −0.83, r = −0.78) between flare power and VLF amplitude under similar conditions. The Long Wave Propagation Capability (LWPC) code has been used to explain the differences in the observed amplitude perturbations due to solar flares in both receivers. Solar flares weaker than C2.0 at CHB reduce attenuation and phase velocity of the propagating waveguide modes in the earth-ionosphere waveguide, causing positive amplitude perturbations, while stronger flares increase these parameters, resulting in negative perturbations. In contrast, solar flares of all classes cause an increase in phase velocities and a decrease in attenuation coefficients of the propagating waveguide modes along the VTX-CUB propagation path, resulting in positive VLF amplitude responses. This study highlights distinct responses of VLF signals to solar flares in different propagation paths, emphasizing the complex interactions between solar activity and earth-ionosphere waveguide properties.</div></div>","PeriodicalId":15096,"journal":{"name":"Journal of Atmospheric and Solar-Terrestrial Physics","volume":"268 ","pages":"Article 106433"},"PeriodicalIF":1.8,"publicationDate":"2025-01-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143177675","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Seasonal asymmetries in the lag between insolation and surface air temperature over Iran during 1971–2017","authors":"Mohammad Darand ,&nbsp;Khabat Ghamari ,&nbsp;Mohammad Yasin Khaledyan ,&nbsp;Anmin Duan ,&nbsp;Jun Jian ,&nbsp;Yuepeng Pan","doi":"10.1016/j.jastp.2025.106441","DOIUrl":"10.1016/j.jastp.2025.106441","url":null,"abstract":"<div><div>The seasonal cycle in surface air temperature reflects the systematic variation in incoming solar radiation during a year. The current study focused on spatiotemporal analysis of the seasonal asymmetries in the lag between insolation and surface air temperature over Iran and the long-term trend over Iran. To do this, daily gridded surface air temperature data with spatial resolution of 0.25° × 0.25° over Iran during period 1971–2017 has been used. The results demonstrated that the seasonal cycle of surface air temperature did not coincide on annual harmonic. The time lag of maximum surface air temperature between insolation forcing and maximum surface air temperature response varies from 1 to 48 days across Iran. Maximum surface air temperature shows short lags over southeastern regions and long lags over western and northwestern regions. The time lag of minimum surface air temperature relative to winter solstice is less variable than maximum surface air temperature and differs from 26 to 48 days across Iran. The seasonal asymmetries (ASYM) which defined as the time lag of maximum surface air temperature relative to summer solstice minus the time lag of minimum surface air temperature relative to winter solstice generally showed positive values over the northwestern and western regions, indicating prolonged spring. Negative values over the southeastern, eastern and southwestern coastal regions of the Caspian Sea in the north, showing that the winter minimum surface air temperature is delayed more than the summer maximum surface air temperature. Most areas over the country have encountered negative trends in the time lag of minimum surface air temperature. These results suggest that changes in seasonal surface air temperature lags can be a potential predictor of shifting in climatic parameters at a seasonal time scale.</div></div>","PeriodicalId":15096,"journal":{"name":"Journal of Atmospheric and Solar-Terrestrial Physics","volume":"268 ","pages":"Article 106441"},"PeriodicalIF":1.8,"publicationDate":"2025-01-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143177672","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Boosting total electron content forecasting based on deep learning toward an operational service","authors":"María Graciela Molina ,&nbsp;Jorge H. Namour ,&nbsp;Claudio Cesaroni ,&nbsp;Luca Spogli ,&nbsp;Noelia B. Argüelles ,&nbsp;Eric N. Asamoah","doi":"10.1016/j.jastp.2025.106427","DOIUrl":"10.1016/j.jastp.2025.106427","url":null,"abstract":"<div><div>We present a prediction model based on deep learning able to forecast ionospheric Total Electron Content at global level 24 h in advance. It has been conceived to operate under different space weather scenarios and in an operational framework. Three different deep learning (DL) techniques have been compared: Long Short Term Memory (LSTM), Gated Recurrent Units (GRU) and Convolutional Neural Networks (CNN). The modelling approach inherits by and extends what has been proposed by Cesaroni and co-authors (2020a). Specifically, the machine learning-based approach here reported is conceived to improve the first step of Cesaroni et al. (2020a), in which TEC is forecasted on 18 selected grid points of Global Ionospheric Maps (GIMs) using the geomagnetic global index Kp index as the external input.</div><div>CNN models provide better predictive capabilities than LSTM and GRU, and it has more robust behaviour under different space weather conditions. We also show how all the proposed models outperform the two naive models: the so-called “frozen ionosphere” or recurrence model and a 27 days averaged model.</div><div>The novelty of our approach is the operational capability based on an incremental learning method to prevent the aging of the trained models by updating the weights with little computational effort adding new information immediately after the 24-h forecasting. The improvement changed from RMSE of ∼6.5 TECu to ∼2.5 TECu.We also discuss limitations and the use of other space weather inputs (e.g. solar proxies, other geomagnetic indexes, etc) and the use of complementary data science techniques (e.g. data preparation, hyperparameter tuning, better data resolution, etc.) to enhance the forecasting in future works.</div></div>","PeriodicalId":15096,"journal":{"name":"Journal of Atmospheric and Solar-Terrestrial Physics","volume":"268 ","pages":"Article 106427"},"PeriodicalIF":1.8,"publicationDate":"2025-01-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143177699","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Strong thermospheric response to the almost undetectable substorm on May 29, 2023","authors":"Yongliang Zhang ,&nbsp;Qian Wu ,&nbsp;Wenbin Wang ,&nbsp;Larry Paxton ,&nbsp;Robert Schaefer ,&nbsp;Dong Lin ,&nbsp;Lying Qian ,&nbsp;Haonan Wu ,&nbsp;Kun Wu ,&nbsp;Ying Zou ,&nbsp;Martin Connors","doi":"10.1016/j.jastp.2025.106430","DOIUrl":"10.1016/j.jastp.2025.106430","url":null,"abstract":"<div><div>A ground based FPI (Fabry Perot Interferometer) at the Athabasca Observatory detected an unusual strong and storm-like equatorward meridional wind of up to 450 m/s on May 29, 2023, a geomagnetically quiet day (AE &lt; 150 nT, Kp &lt; 1). F18 DMSP SSUSI, a Far Ultra-Violet (FUV) spectrograph imager, observed a long lasting (∼7 h) auroral substorm on the same day. TIMED/GUVI data showed a O/N₂ depletion that extended to mid/low latitudes over a limited longitude range in the northern hemisphere. Concurrent SuperDARN measurements indicated strong plasma convection around the substorm location, suggesting a strong local heating (Joule and particle precipitation heating) near the substorm location. This strong and localized heating caused the storm-like response in the thermospheric meridional wind and composition. Furthermore, the FPI also observed a strong zonal wind (up to 180 m/s), which changed its direction from westward to eastward during the substorm. Such a change is due to the competition between the pressure gradient and Coriolis forces. In the Northern Hemisphere, the Coriolis force is westward with an equatorward meridional wind during the substorm; the direction of the pressure gradient force changed from westward to eastward due to changes in the relative locations of the observatory and the substorm. A strong IMF B_y and periodic variation in the IMF likely provide a favorable upstream condition for continuous energy input from the solar wind to the magnetosphere and/or the release of the stored magnetospheric energy into the thermosphere to drive the long duration substorm and the observed thermospheric changes.</div></div>","PeriodicalId":15096,"journal":{"name":"Journal of Atmospheric and Solar-Terrestrial Physics","volume":"268 ","pages":"Article 106430"},"PeriodicalIF":1.8,"publicationDate":"2025-01-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143176802","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Localized solar radiation zoning by combining spatially continuous estimates and Gaussian mixture models","authors":"Xuecheng Wang ,&nbsp;Peiran Xie ,&nbsp;Yiyi Xie ,&nbsp;Hou Jiang","doi":"10.1016/j.jastp.2025.106432","DOIUrl":"10.1016/j.jastp.2025.106432","url":null,"abstract":"<div><div>With the increasing role of solar energy in the global decarbonization, precise geographical zoning of solar radiation becomes crucial. Traditional methods of solar radiation zoning struggle to accurately distinguish subtle spatial and temporal differences in solar radiation due to both sparse ground-based observations and the requirement for a predefined zone number, which limits their applicability for the demands of distributed photovoltaic system. This study introduces a novel method for localized solar radiation zoning, integrating spatially continuous solar radiation data with a Gaussian mixture model. High-precision spatiotemporal estimates of solar radiation are achieved by employing deep learning algorithms to analyze meteorological satellite imagery and digital elevation model data. The use of an infinite Gaussian mixture model along with variational inference allows for the adaptive determination of the number of solar radiation zones. The case study in Guangxi Province shows that incorporating Digital Elevation Model data reduces the root mean square error of global solar radiation estimates from 134.06 W/m² to 87.68 W/m² and accurately reveals temporal and spatial variability in both global and diffuse solar radiation. This approach not only prevents overfitting when the predefined upper bound surpasses the actual number of zones but also facilitates the development of zoning schemes that can range from fine-grained, capturing subtle variations, to coarse-grained, focusing on overall patterns. The outcomes lay a solid foundation for localized regional assessment and efficient utilization of solar energy resources.</div></div>","PeriodicalId":15096,"journal":{"name":"Journal of Atmospheric and Solar-Terrestrial Physics","volume":"268 ","pages":"Article 106432"},"PeriodicalIF":1.8,"publicationDate":"2025-01-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143177678","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Evaluation of water vapor from CARRA reanalysis based on GNSS and radiosonde observation in the Arctic","authors":"Ce Zhang ,&nbsp;Shuaimin Wang ,&nbsp;Yuling Zhao ,&nbsp;Yujing Xu ,&nbsp;Jiajia Zhang ,&nbsp;Yanhan Mo ,&nbsp;Hong Yu","doi":"10.1016/j.jastp.2025.106431","DOIUrl":"10.1016/j.jastp.2025.106431","url":null,"abstract":"<div><div>The Arctic has a significant impact on global climate change because of its special geographical location. Meanwhile, water vapor is one of the most important atmospheric components influencing climate change. In this study, the accuracy of precipitable water vapor (PWV) from Copernicus Arctic Regional Reanalysis (CARRA) is evaluated with 45 Global Navigation Satellite System (GNSS), 13 radiosonde stations and ERA5 reanalysis during 2020–2022 in the Arctic. The mean bias values of CARRA PWV using GNSS PWV, radiosonde PWV and ERA5 PWV are 0.30 mm, −0.00 mm and 0.21 mm, respectively. The results show that the CARRA PWV is mostly overestimated compared with GNSS and ERA5 PWV and not much different for the radiosonde PWV. The RMSE values of CARRA PWV using GNSS PWV, radiosonde PWV and ERA5 PWV are 0.71 mm, 0.40 mm and 0.67 mm, respectively. The results indicate that CARRA PWV has good agreement with GNSS PWV, radiosonde PWV and ERA5 PWV, but CARRA PWV is better agreement with radiosonde PWV. Afterwards, the seasonal bias and RMSE values of CARRA PWV using GNSS PWV, radiosonde PWV and ERA5 PWV are analyzed. The RMSE in the warm season is obviously higher than that in the cold season. Finally, the monthly RMSE values of CARRA PWV are analyzed. The results show that the RMSE values of CARRA PWV with respect to GNSS PWV, radiosonde PWV and ERA5 PWV are large in the warm season and small in the cold season, thereby indicating that the accuracy of CARRA PWV has obvious seasonality. The RMSE values of CARRA PWV and radiosonde PWV are smaller than the RMSE values of CARRA and GNSS PWV and CARRA PWV and ERA5 PWV in every month. The results illustrate that CARRA PWV has better consistency with radiosonde PWV compared with GNSS and ERA5 PWV even in every month.</div></div>","PeriodicalId":15096,"journal":{"name":"Journal of Atmospheric and Solar-Terrestrial Physics","volume":"268 ","pages":"Article 106431"},"PeriodicalIF":1.8,"publicationDate":"2025-01-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143176800","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Determinants of diffuse solar radiation in urban and peatland areas based on weather and air pollutants","authors":"Arnida L. Latifah ,&nbsp;Amandha Affa Auliya ,&nbsp;Inna Syafarina ,&nbsp;Sheila Dewi Ayu Kusumaningtyas","doi":"10.1016/j.jastp.2025.106419","DOIUrl":"10.1016/j.jastp.2025.106419","url":null,"abstract":"<div><div>Understanding solar radiation variability is essential for efficiently planning and managing solar energy systems. The transmission of solar radiation to the ground is generally affected by a variety of factors. This research deals with the impacts of weather and air pollution on the amount of diffuse solar radiation across two distinct environments, i.e., urban and peatland areas. Two stations in Jakarta, Kemayoran and Jagakarsa represent the urban area, while two stations in South Kalimantan, Pinang Habang and Jambu, represent the peatland area. Three machine learning-based models were used to estimate the amount of diffuse solar radiation related to weather and air pollution, namely Random Forest (RF), K-nearest neighbor (KNN), and Light Gradient Boosting Machine (LGBM). Five experiments were carried out using various combinations of predictor variables, including temperature, air pollutants, and cloud cover. The results of the experiments highlighted the significance of pollutants as predictive factors. All models demonstrated reliable results in capturing the variability of diffuse solar radiation in four stations, revealing that urban areas receive approximately half the amount of diffuse solar radiation compared to peatland areas, despite sharing a similar annual pattern. Among the models, RF model achieved the highest correlation coefficient with actual values, yielding the least error. Among the sites studied, the predictions from peatland areas closely aligned with the reference pattern. Furthermore, this study found that CO is the primary factor in predicting diffuse solar radiation in urban areas. Differently, PM_2.5 mostly impact the diffuse solar radiation in rural areas. This research underscores the critical role of air pollutants, particularly CO and PM_2.5, in determining solar radiation levels, which in turn affects the efficiency of solar energy systems.</div></div>","PeriodicalId":15096,"journal":{"name":"Journal of Atmospheric and Solar-Terrestrial Physics","volume":"268 ","pages":"Article 106419"},"PeriodicalIF":1.8,"publicationDate":"2025-01-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143177676","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}