Air quality degradations caused by fine particulate matter (PM2.5) can lead to various health problems, and accurate PM2.5 data are critical for managing the environment and ensuring public health. Radiation signals collected by nighttime light (NTL) remote sensing satellites are influenced by PM2.5 concentrations, and thus, incorporating NTL imagery in statistical models has been widely used to predict PM2.5 concentrations. However, scarce work has been carried out with new-generation NTL data from the LJ1-01 satellite, which has a fine spatial resolution and wide measurement range. In this study, we integrated satellite observation data and meteorological data to construct five models based on the geographically weighted regression to validate the feasibility of LJ1-01/NPP-VIIRS in Moderate Resolution Imaging Spectroradiometer AOD-based PM2.5 prediction in the Beijing–Tianjin–Hebei region. The models were validated by the cross-validation method. The results showed that the addition of NTL information could improve the performance of the PM2.5 prediction model. The seasonal R2 with NTL in AOD-PM2.5 model have improved by 5.07%, 4.50%, 2.95%, and 2.56% in model fitting and 1.20%, 1.75%, 2.20%, and 4.41% in cross-validation. Furthermore, the LJ1-01 NTL data revealed additional details and improved the prediction accuracy, compared with the NPP-VIIRS in AOD-PM2.5 model, the seasonal R2 with LJ1-01 in AOD-PM2.5 model increased by 1.16%, 1.79%, 0.76%, and 1.15% in model fitting and 1.04%, 0.85%, 0.78%, 1.37% in cross-validation. Thus, our findings indicate that LJ1-01 and NTL data have the potential for predicting PM2.5 and that they could constitute a useful supplemental data source for estimating ground-level PM2.5 distributions.

中文翻译：

---

评估 LJ1-01 夜间灯光图像用于估算每月 PM2.5 浓度：与 NPP-VIIRS 夜间灯光数据的比较

细颗粒物 (PM2.5) 导致的空气质量下降会导致各种健康问题，准确的 PM2.5 数据对于管理环境和确保公众健康至关重要。夜间光 (NTL) 遥感卫星收集的辐射信号受 PM2.5 浓度的影响，因此，将 NTL 图像纳入统计模型已被广泛用于预测 PM2.5 浓度。然而，LJ1-01卫星的新一代NTL数据缺乏空间分辨率和宽测量范围的研究。在本研究中，我们综合卫星观测数据和气象数据，构建了基于地理加权回归的五个模型，以验证 LJ1-01/NPP-VIIRS 在中分辨率成像光谱仪 AOD 基 PM2.5 中的可行性。5 京津冀地区预测。模型通过交叉验证方法进行验证。结果表明，加入NTL信息可以提高PM2.5预测模型的性能。AOD-PM2.5模型中带有NTL的季节性R2在模型拟合方面分别提高了5.07%、4.50%、2.95%和2.56%，在交叉验证方面分别提高了1.20%、1.75%、2.20%和4.41%。此外，LJ1-01 NTL数据揭示了更多细节并提高了预测精度，与AOD-PM2.5模型中的NPP-VIIRS相比，AOD-PM2.5模型中LJ1-01的季节性R2增加了1.16%，模型拟合分别为 1.79%、0.76% 和 1.15%，交叉验证为 1.04%、0.85%、0.78%、1.37%。因此，我们的研究结果表明 LJ1-01 和 NTL 数据具有预测 PM2.5 的潜力。