Published online by Cambridge University Press: 20 December 2006
Springtime meteorological and satellite data from 1971 to 2000 have been used to assess surface desertification and grassland deterioration to obtain a better understanding of the synoptic characteristics of dust storms observed in Inner Mongolia, the relationship between dust storms and their impact on the Beijing-Tianjin Region, and the effect of climatic change on the surface environment of dust storm source areas in Inner Mongolia. Based on the statistics and the analysis of weather charts during each dust storm event, details and characteristics of sand and dust source regions, such as cold air paths, areas of strong winds and typical synoptic types were obtained. Temperature was found to be the most significant indicator of climate change in the dust storm source regions, having increased by 1.01 °C in the Alxa source region and by 0.96 °C in Hunshdak over the past three decades. Other factors include lack of precipitation, higher evaporation, lower relative humidity, soil desertification and decrease of grasslands in the source regions. As a consequence, nearly two-thirds of the very strong dust storm events observed in Inner Mongolia (characterised by horizontal visibility below 1000 m and wind speeds of 12 m s−1 or over) led to suspended dust or dust storms in the downwind area, the Beijing-Tianjin Region. This study also analyses the upper air and surface weather charts of 58 very strong dust storm events and established the direction of three cold air paths: northwesterly, westerly and northerly. The northwesterly cold air path is seen as the major one, responsible for 58.6% of the dust storm episodes. In addition, the observed areas where strong winds occur lie mostly across the central-west and central regions of Inner Mongolia. Finally, the synoptic situations associated with dust storm events can be classified into three types: trough, northwesterly current and Mongolian vortex. The predominant synoptic system is the trough, responsible for 63.8% of the 58 events analysed in this study.