Flash droughts refer to those droughts that intensify rapidly in spring and summer, coupled with a strong increase in summer extreme temperatures, such as those that occurred over Texas in 2011 and the Great Plains in 2012. Climate models failed to predict these flash droughts in 2011 and 2012 and are ambiguous in projecting their future changes, largely because of models’ weaknesses in predicting summer rainfall and soil moisture feedbacks. In contrast, climate models are more reliable in simulating changes of large‐scale circulation and temperatures during winter and spring seasons. Thus, we developed and tested a physical climate indicator of the risk of “flash” droughts in summer by using the large-scale circulation and land surface conditions in winter and spring based on observed relationships between these conditions and their underlying physical mechanisms established by previous observational studies and numerical model simulations. My master research focuses on the spatial distribution of this indicator globally to see how broadly it could be applied. We also compare the different factors to see which one is the dominant contributor to drought in different area. We find that the indicator performs well at capturing the development and termination of a drought. There is much opportunity to develop and improve the indicator further. / text