The Study Of Spatial And Temporal Distribution Of Drought Based On Spatial Fuzzy Cluster Algorithm Of AWTP

With the characters of a long duration, a large scope and a high frequency of occurrence,drought was one of the most serious disasters which have impact on human society. Mostindicators were chosen by taking into account of the distance between the accumulatedrainfall and the average rainfall for a long period of time. but in fact, taking into considerationof uneven spatial and temporal distribution of precipitation, when there was such a conditionof without precipitation for a long period, severe drought may still occur.On the contrary, ifthe spatial and temporal distribution of precipitation was more equably, the drought may notoccur in spite of less precipitation.It was a better way that the index of average wait time forprecipitation and index of equivalent dry period were used by analysis the successive periodof without precipitation.With the rapid development of computer science and meteorological causes,meteorological data was obtained by different ways.Due to the rapid growth ofmeteorological data, it needed us to discover potential knowledge instead of traditional dataanalysis. As GIS technology and data mining technology develop, it was possible for us tomine the massive meteorological data by means of combing GIS technology and data miningtechnology.In this paper, according to22meteorological stations daily precipitation data from1958to2008in Liaoning province, using the index of averaged waiting time for precipitation(AWTP),the values of equivalent dry period (EDP) of22meteorological stations werecalculated, temporal and spatial characteristics of drought was obtained by using FCMalgorithm in22time series of EDP values. The results showed that from variation perspective,there was an ascending trend in drought, in which the annual amplitude oscillated sharply, ithas a climbing trend of1.327/10a.From the point of space it showed that Liaoning had asignificant regional drought, which decreased from west to east; the drought of WesternChaoyang was most severe and Dandong from Eastern Liaoning was lightest. From theperspective of switching time series, it had7switching time series, they were analyseddetailedly and could be a good theoretical basis for regional drought.