The Identification and Changing Characteristics of Regional Low Temperature Extreme Events

When an extreme low temperature event occurs, it generally impacts a certain area and lasts for some time, which means that it is a regional extreme event. How to identify a regional extreme low temperature is the basis for studies in this area. An objective identification technique for regional low temperature extreme events (OITRLTE) is developed. This technique consists of four parts: Defining the threshold value of extreme low temperature for single station; identifying abnormality belts; distinguishing temporal continuous process of the event; an integrated index system. The index is specially developed based on the features of regional events, which includes 5 single indices: Extreme intensity, accumulated intensity, accumulated area, maximum impacted area and duration, as well as an integrated index. Case studies show that OITREE is skillful in identifying regional low temperature extreme events (RLTEs). It can objectively and automatically capture daily impacted areas of a regional event for its duration, and reasonably putting them in a "string" to shape an entire regional event. Then based on the winter daily minimum temperature from 1960 to 2009, spatial distribution and temporal changes of RLTEs is also investigated. Results show that probability distribution of lowest temperature and latitude of geometrical center of RLTEs both obey the two-peak distribution, and the center of RLTEs mainly locates at two belts of 32°N and 42°N. The annual accumulative value of the frequency, intensity and max covering area of RLTEs is decreasing, and during the end of 1980s this trend changes and the trend becomes stationary after 1990s. These characteristics might be caused by the RLTEs with long duration and wide space range that accounted for top 10% of all events. Considering the good correlation between RLTE indexes, economic loss and the number of stricken people on cold disasters, an integration index is defined based on RLTE indexes. The weighted coefficients of the first-grade integrated index are defined based on the correlation between the yearly cumulative value of first-grade index, the corresponding yearly index of economic losses and the number of stricken people on cold disasters. So, in this way the weight coefficients denote the correlation between regional low temperature extreme events and corresponding disaster losses to some extent.