Abstract: The occurrence of drought is a cumulative and gradual process. The mechanism underlying the effects of drought on crops is complicated. It is important to accurately and quantitatively identify and evaluate the occurrence and development of drought. In this study, the crop water deficit index model was improved by optimizing the calculation method of water supply and demand, with the influences of soil, crops and meteorological conditions being considered. The simulation of daily water deficit for maize beginning from emergence was achieved, realizing the dynamic and quantitative description of droughtinduced disaster process of maize. Using the data of meteorological condition, development period of maize, soil moisture and yield during 1981-2018 in Liaoning, based on crop water deficit index(CWDI_wp) and relationship between yield reduction rate and drought intensity at representative sites in typical years with spring, autumn, springtosummer droughts, twofactor (i.e., duration and CWDI_wp) dynamic threshold index of droughtinduced disaster was determined. Then, the disastercausing processes during typical drought periods were evaluated. The results showed thatCWDI_wp was negatively correlated with soil relative moisture, which could easily identify the starting and ending time, duration andintensity of drought as well as to describe the drought process including occurrence, development, persistence, mitigation and relief more accurately relative to the original model. Moreover, the improved method could be used to daily rolling assessment of regional drought development process, with the assessment result being consistent with the actual situation with a relative error of about 9% for drought areas with different levels.

Key words: maize, drought, crop water deficit index, disaster-causing process, identification, quantitative evaluation.