Abstract:
Based on the data of MAM (March,April,and May) precipitation from 1963 to 2015 and monthly 144 circulation characteristics from 1962 to 2014 in 81 meteorological stations,as well as the GEV drought index,we have applied the methods of Mann-Kendall(MK)and Morlet wavelet to analyze temporal and spatial variation of spring drought and building up one drought forecast model for Guizhou province.The results demonstrate that the position parameter can properly show the spatial characteristic of spring precipitation in Guizhou gradually growing from west to east,and the scale parameter can well indicate rainless northwest areas and the rainy east areas,and the shape parameters reflecte the Weibull uniform distribution of spring precipitation in Guizhou.We've also found that the GEV drought index and its grade division can well describe actual distribution of the spring drought in Guizhou,and it is better to use seasonal GEV drought index in continuous drought process.In addition,there exists the oscillation cycles of 4—6,8—10,24—28 years of spring drought in Guizhou,and the frequency of serious spring drought is during 6%—15%.The serious spring drought occurred in 1979,1986,1987,1988,1991 and 2011 respectively.Moreover,statistical analysis shows that preceding south oscillation,the west ridge and intensity of northwest Pacific subtropical high,and India-Myanmar trough correlate well with the trend of Guizhou spring drought to some extent.By applying forecast signals and taking a period from 1963 to 2007 as a swatch,a short-term climatic forecast model of spring drought in Guizhou has been established.With this model,an independent 8-year test forecast of 2008—2015 has been made and the results are satisfactory.