Satellite measurements from the thermal Infrared Atmospheric Sounding Interferometer (IASI), aircraft data from the MOZAIC/IAGOS project, as well as observations from ground-based stations, are used to assess the tropospheric ozone (O3) variability during the East Asian Summer Monsoon (EASM). Six years 2008–2013 of IASI data analysis reveals the ability of the instrument to detect the onset and the progression of the monsoon seen by a decrease in the tropospheric 0–6 km O3 column due to the EASM, and to reproduce this decrease from one year to the other. The year-to-year variability is found to be mainly dependent on meteorology. Focusing on the period of May-August 2011, taken as an example year, IASI data show clear inverse relationship between tropospheric 0–6 km O3 on one hand and meteorological parameters such as cloud cover, relative humidity and wind speed, on the other hand. Aircraft data from the MOZAIC/IAGOS project for the EASM of 2008–2013 are used to validate the IASI data and to assess the effect of the monsoon on the vertical distribution of the tropospheric O3 at different locations. Results show good agreement with a correlation coefficient of 0.73 (12 %) between the 0–6 km O3 column derived from IASI and aircraft data. IASI captures very well the inter-annual variation of tropospheric O3 observed by the aircraft data over the studied domain. Analysis of vertical profiles of the aircraft data shows a decrease in the tropospheric O3 that is more important in the free troposphere than in the boundary layer and at 10–20° N than elsewhere. Ground station data at different locations in India and China show a spatiotemporal dependence on meteorology during the monsoon, with a decrease up to 22 ppbv in Hyderabad, and up to 5 ppbv in the North China Plain.
24 March 2024 in Proceedings of the AAAI Conference on Artificial Intelligence
Mixed-Modal Image Retrieval (MMIR) as a flexible search paradigm has attracted wide attention. However, previous approaches always achieve limited performance, due to two critical factors are seriously overlooked. 1) The contribution of image and text modalities is different, but incorrectly treated equally. 2) There exist inherent labeling noises in describing users' intentions with text in web datasets from diverse real-world scenarios, giving rise to overfitting. We propose a Dynamic Weighted Combiner (DWC) to tackle the above challenges, which includes three merits. First, we propose an Editable Modality De-equalizer (EMD) by taking into account the contribution disparity between modalities, containing two modality feature editors and an adaptive weighted combiner. Second, to alleviate labeling noises and data bias, we propose a dynamic soft-similarity label generator (SSG) to implicitly improve noisy supervision. Finally, to bridge modality gaps and facilitate similarity learning, we propose a CLIP-based mutual enhancement module alternately trained by a mixed-modality contrastive loss. Extensive experiments verify that our proposed model significantly outperforms state-of-the-art methods on real-world datasets. The source code is available at https://github.com/fuxianghuang1/DWC.
24 April 2024 in International Journal of Computer Vision
Northern China was hit by a severe dust storm on 15 March 2021, covering a large area and bring devastating impact to a degree that was unprecedented in more than a decade. In the study, we carried out a day-and-night continuous monitoring to the path of the moving dust, using multi-spectral data from the Chinese FY-4A satellite combined with the Japanese Himawary-8 from visible to near-infrared, mid-infrared and far-infrared bands. We monitored the whole process of the dust weather from the occurrence, development, transportation and extinction. The HYSPLIT(Hybrid Single Particle Lagrangian Integrated Trajectory) backward tracking results showed the following two main sources of dust affecting Beijing during the north China dust storm: one is from western Mongolia; the other is from arid and semi-arid regions of northwest of China. Along with the dust storm, the upper air mass, mainly from Siberia, brought a significant decrease in temperature. The transport path of the dust shown by the HYSPLIT backward tracking is consistent with that revealed by the satellite monitoring. The dust weather, which originated in western Mongolia, developed into the “3.15 dust storm” in north China, lasting more than 40 h, with a transport distance of 3900 km, and caused severe decline in air quality in northern China, the Korean peninsula and other regions. It is the most severe dust weather in the past 20 years in east Asia.
16 September 2023 in Journal of Geophysical Research: Atmospheres