Based on themonthly p recip itation data during 1951 - 2000 from 19 meteorological stations in Guizhou p rovince, the regional
characteristics of p recip itation anomalies in summer are studied by usingmethods of the fuzzy assembling analysis, emp irical orthogonal
function ( EOF) analysis and wavelet analysis on normalized anomaly of 19 stations’monthly averaged rainfall. The results show that
five obvious rainfall periods occurred in the recent 50 years, includingwet in the early 1950 s, dry from the middle of the 1950 s to the
early 1960 s, wet from the middle of 1960 s to the end of 1960 s, dry in the 1970 s and 1980 s, wet after the 1990 s, at the same time ,
the total rainfall p resented increasing trend. The coincident variability occurred in summer rainfall of Guizhou, but some regional differ2
ence also existed. The variations of p recip itation anomalies in summer can be divided into five spatial patterns, namely dry (wet) in
the whole p rovince, dry (wet) in the east and wet ( dry) in the west, dry (wet) in the south and wet ( dry) in the north, dry (wet )
in the middle and east part butwet ( dry) in the southwest, dry (wet) in the southwest and wet ( dry) in other areas. It p resented ob2
viousmultip le time scales of 10 - 12 years, 4 - 5 years and 2 - 3 years variations for different rainfall patterns.

Based on precipitation data in 126 surface stations in Northwest China, using REOF and tial characteristics of summer preclpltahon mNnlthwest China  were analyzed in the paper. Results Blur-clustering method, the spa-show that the northeastern side of Tibetan Plateau is the most tion is mostly a half of that nant.  PTPClpltatlOn  trend  In sensitive region to summer precipitation in Northwest China, the rainy 凡e月凡On (July to August) precipita-m a peal pears Wlth less Taln are  mOTe than Ta1nV  VeaTS  in the region, and dry and flood years al'e  a.Itel'-
.Tune is ascending changing in recent 40 years, changing slowly in July, and descending in August, fiom the liner trend, the summer precipitation is nnconspicnonsly.  Moreover, temperature departure and precipitation departure percent present reverse phase, and temperature is rising in recent 40 years.

Analysis of the climatological features of Jiangnan rainy season precipitation anomalies shows that the month with the most
frequent anomalies of precipitation in Jiangnan is J une and the region is located in the middle of Jiangnan. The SVD analysis points
out that the related sea surface temperature anomaly (SSTA) distributes in the pattern as a“seesaw”. The key region is just located
in the most evident correlation area. The key period of SSTA impact is from May to J uly in the previous year in which SSTA has the
best negative correlation with April to May precipitation in Jiangnan. Under the influence of SSTA“seesaw”pattern the precipitation
anomalies in the following April to J une in East China exhibit opposite phase on both sides of the Yangtze River , the most evident
correlative region is Jiangnan and the middle of which is located in the most obvious area of SSTA impacts.