Trends in daytime and nighttime dust aerosols over the Dust Belt revealed by IASI

The "dust belt", extending from North Africa through the Middle East to Asia, is the dustiest region in the world. While many studies have examined the trends in dust aerosols in these regions, little is known about their long-term variations in the diurnal cycle. Using dust products from the Infrared Atmospheric Sounder Interferometer (IASI) instrument aboard the MetOP-A satellite, this study examines the interannual variations and trends in dust optical depth (DOD) and dust layer altitude (DLA) at day and night overpass times of IASI (i.e., 9:30 a.m. and 9:30 p.m. local solar equator-crossing time (ECT)) and their day-night differences (daytime minus nighttime) over the dust belt. The latter allows a direct comparison of dust variability at daytime and nighttime. Statistically significant positive trends are detected in DOD over the Sahara at nighttime in summer (JJA), along with a negative trend in the day-night differences. Over the Indian Peninsula, statistically significant decreasing trends in DOD at daytime, nighttime, and in the day-night differences are observed in the spring (MAM). Over the Arabian Peninsula, significantly negative trends are detected at daytime, nighttime, and in the day-night differences in both MAM and JJA. Significant trends are also observed in DLA over the Sahara in JJA for daytime, nighttime, and day-night differences, as well as in the Arabian Peninsula at both daytime and nighttime. The trends in MAM and JJA are the strongest among all seasons and dominate the annual mean trend. Ensemble Random Forest (RF) regression is then used to examine the drivers of the trends in DOD and DLA in MAM and JJA. Key controlling factors include surface temperature, boundary layer height, precipitation, and vegetation cover. Overall, the ensemble RF models can capture 85 -98% of the interannual variances in DOD and DLA in our focused regions from 2007 to 2021. These results complement dust trend analysis and attribution based on DOD at visible-wavelength, advancing the current understanding of dust variability, especially diurnal variations.