Global aeolian dust variations and trends: a revisit of dust event and visibility observations from surface weather stations

This study revisits the use of horizontal visibility and manually reported present weather (ww) records from the NOAA Integrated Surface Database (ISD) for characterizing the aeolian dust variability and recent trends over the globe and three largest source regions (North Africa, Middle East, and East Asia). Due to its qualitative nature, ww is combined with visibility to derive a new variable, VI, which has higher correlations with the dust emission and burden from satellite observations and global aerosol reanalyses than does the dust event frequency (FR) derived from ww only. Both FR and VI capture the intensive dust activity associated with the prolonged North American drought during the 1950s and Sahelian drought during the 1980s. Correlation analysis suggests soil moisture has a lagged effect on the global dustiness, with a maximum r = −0.3 when soil moisture leads VI by 14 months. Through a critical assessment of the ww continuity and ww-visibility consistency of various report types in ISD, the SYNOP data are used for global dust trend detection from 1986 to 2019. Globally, FR and VI decreased at a rate of −0.23 % yr−1 and −8.0 × 10−4 km−1 yr−1, respectively, from 1986 to 1996/1997 when dust reached a minimum, followed by a slower rebound at a rate of 0.085 % yr−1 and 1.9 × 10−4 km−1 yr−1, respectively. The nonlinear behavior of global dustiness is qualitatively consistent with satellite observations and global aerosol reanalyses. Regionally, North Africa experienced increased dust activity during the past decade after staying below average for most of the 1990s–2000s, in response to reduced soil moisture and increased wind speed following the transition of North Atlantic Oscillation (NAO) from strong negative to recurring positive phases since 2011. In the Middle East, dust has been increasing since 1998 due to a prolonged drought in the Tigris-Euphrates basin associated with strong negative Pacific Decadal Oscillation (PDO) phases. As PDO turned positive and weak negative after 2015, the amelioration of drought has led to decreased dust activity in recent years. The dust variability in East Asia is primarily driven by wind speed, which explains the dust decline from 1986 to 1997, and the absence of dust trends during the past two decades. This study constitutes an initial effort of creating a homogenized weather station-based dust-climate dataset in support of wind erosion monitoring, dust source mapping, and dust-climate analysis at local to global scales