第一届未来大气科学论坛

This research comprehensively assesses the suitability of MODIS07 atmospheric profile data in capturing vertical temperature distributions throughout China. Through a comparison of MOD07-derived temperatures and radiosonde-observed temperatures from 2003 to 2020, we examined the retrieval accuracy at various temporal scales (instantaneous, weekly, monthly) and spatial resolutions. The findings show that, in the upper pressure levels (400–620 hPa), the MODIS07 retrieval accuracy notably outperforms that in lower pressure levels, especially in areas with flat topography and stable weather conditions. Specifically, for the 400 hPa layer during daytime, the correlation coefficient (r) is 0.95, the bias is nearly 0°C, and the RMSE is 2.86°C. In contrast, in the lower pressure layers (920–1000 hPa), the daytime bias rises noticeably to 0.22°C at 920 hPa, with an RMSE of roughly 4.0°C. It suggests that the accuracy in upper atmospheric layers improves by approximately 30% compared to the lower layers. On extended timescales (such as monthly averages), the retrieval accuracy substantially exceeds that of instantaneous measurements. Under monthly averaging, the RMSE at the 400 hPa level drops to 1.95°C, and the correlation coefficient (r) rises to 0.98. For the lower pressure layers, the RMSE is reduced to about 3.0°C, and the correlation (r) climbs to roughly 0.98, illustrating that the smoothing effect over extended periods enhances both the stability and coherence of the retrieval outcomes. Nighttime retrieval accuracy typically surpasses daytime accuracy, particularly in lower pressure layers; the nighttime RMSE is about 3.0°C, which represents a roughly 25% reduction compared to daytime. Moreover, MODIS07 exhibits restricted capabilities for inversion layer detection. Comparative analysis shows that MODIS07 detects only about 15% of the inversion layers observed by radiosondes, dropping to approximately 9% during daytime, and reveals marked omissions and uncertainties in low-latitude and low-altitude areas. In summary, MODIS07 data hold considerable promise for monitoring vertical atmospheric temperature distributions over broad regions and long periods, but additional improvements are needed to enhance retrieval accuracy in areas with complex topography and moist climates. Subsequent investigations should employ multisource data and localized weather parameters to bolster the utility of MODIS07 in complicated environments
 

MODIS07,Vertical atmospheric temperature,Remote sensing,Temperature retrieval accuracy