L t t d fAi d Long term trends of Asian summer monsoon and Long‐term trends of Asian summer monsoon and Long term trends of Asian summer monsoon and B Tt Y i 1 Y hi ki K jik By Tetsuzo Yasunari 1 Yoshiyuki Kajikawa By Tetsuzo Yasunari 1 , Yoshiyuki Kajikawa By Tetsuzo Yasunari , Yoshiyuki Kajikawa 1 1 HyARC Nagoya Univ Japan 1 HyARC, Nagoya Univ. Japan, HyARC, Nagoya Univ. Japan, Issues Issues 1 H dt ht t th Ai R 1. How, and to what extent has Asian Ra 1. How, and to what extent has Asian Ra monsoon changed since the late ha monsoon changed since the late ha 20 h ? 20th century? Th 20th century? Th 2 How are its regionality and 2. How are its regionality and ra 2. How are its regionality and ra seasonality? Th seasonality? Th seasonality? 3 What are dynamics of these m 3. What are dynamics of these m h ? Th changes in time‐space? Th changes in time space? Th ra ra Data set / Method w Data set / Method w Rainfall Rainfall 1 CMAP ( ) 1 CMAP (Xie and Arkin 1997) 1. CMAP (Xie and Arkin, 1997) 2 GPCP (H ff t l 1997) 2. GPCP (Huffman et al., 1997) ( ) 3 APHRODITE (Yatagai et a al 2009) 3. APHRODITE (Yatagai et a al., 2009) 4 Ui it fD l 4. University of Delaware 4. University of Delaware (http://www esrl noaa gov/psd/) (http://www.esrl.noaa.gov/psd/ ) Water vapor flux Water vapor flux NCEP DOE l i II( ) NCEP DOE reanalysis II (Kanamitsu et al 2002) NCEP DOE reanalysis II (Kanamitsu et al., 2002) H dl C SST d ( ) Hadley Center SST dataset (Rayner 2003) Hadley Center SST dataset (Rayner , 2003) Th tiM K d llt t The non‐parametric Mann‐Kendall test The non parametric Mann Kendall test was applied for assessment of the trends was applied for assessment of the trends. Fi 1( ) Li t d f i f ll (CMAP Sh di /d ) d ti li t td Fi Figure 1: (a) Linear trends of rainfall (CMAP: Shading: mm/day) and vertical integrated Fi Figure 1: (a) Linear trends of rainfall (CMAP: Shading: mm/day) and vertical integrated sa water vapor flux (kg m ‐1 s ‐1 ) in May All values were multiplied by 30 (years) for the period sa water vapor flux (kg m s ) in May. All values were multiplied by 30 (years) for the period CM 1979 2008 Only those that are significant at the 95% level are plotted (b) Same as (a) but CM 1979–2008. Only those that are significant at the 95% level are plotted. (b) Same as (a) but CM f J ( ) Cli l i l diff i i f ll d fl b J dM Pr for June. (c) Climatological difference in rainfall and water vapor flux between June and May. Pr for June. (c) Climatological difference in rainfall and water vapor flux between June and May. () f Figure 6: (a) Longitude‐time section of the trend and Figure 6: (a) Longitude time section of the trend and atmospheric thickness (200 500 hPa) between land atmospheric thickness (200–500 hPa) between land lti li d b 30 ( )f th i d 1979 2008 O multiplied by 30 (years) for the period 1979–2008. O multiplied by 30 (years) for the period 1979 2008. O level are plotted (b) Same as (a) but for SST along 5 level are plotted. (b) Same as (a) but for SST along 5 Fi 5 Dt f li tl i l td i 1979 Figure 5: Date of climatological monsoon onset during 1979– Figure 5: Date of climatological monsoon onset during 1979 1993 (Digit) The onset pentad was determined by the method of 1993 (Digit). The onset pentad was determined by the method of Wang and Linho (2002) The shading denotes the difference in Wang and Linho (2002). The shading denotes the difference in th tb t 1994 2008 d 1979 1993 the monsoon onset between 1994–2008 and 1979–1993. the monsoon onset between 1994 2008 and 1979 1993. R lt II Results II Results II The monsoon onset dates over the BoB Indochina Peninsula and The monsoon onset dates over the BoB, Indochina Peninsula, and h ifi h hif d li f b 10 1d i western‐north Pacific have shifted earlier for about 10‐15days in western north Pacific have shifted earlier for about 10 15days in the recent decades (Fig 5) the recent decades (Fig.5) the recent decades (Fig.5) Warming trend over the continent in May (Fig 6c) has induced Warming trend over the continent in May (Fig. 6c) has induced li l i fl d h (Fi 6) earlier seasonal overturning of land‐sea heat contrast (Fig 6a) earlier seasonal overturning of land sea heat contrast (Fig.6a). i d Overall weakening (intensified) moisture convergence has become Figure 7: Latitude‐time se Overall weakening (intensified) moisture convergence has become Figure 7: Latitude time se trend averaging between significant over land (sea) in the main monsoon months except in trend averaging between significant over land (sea) in the main monsoon months except in i ifi t t 95% l l M dS t b (Fi 7) significant at 95% level ar May and September (Fig. 7) significant at 95% level ar May and September (Fig. 7) d it lit d i th t ld d d its seasonality during the recent several decades d its seasonality during the recent several decades d its seasonality during the recent several decades 1 Sh h i Y hid 2 d Ht ki F ji i 1 a 1 Shuhei Yoshida 2 and Hatsuki Fujinami 1 a 1 , Shuhei Yoshida 2 and Hatsuki Fujinami 1 a , Shuhei Yoshida and Hatsuki Fujinami 2 2 Hitachi Solutions Ltd Japan 2 Hitachi Solutions, Ltd., Japan Hitachi Solutions, Ltd., Japan Results I Results I Results I i f ll i M d i h d d ainfall pattern in May during the recent decades ainfall pattern in May during the recent decades as become close to that in June (Fig 1c) as become close to that in June (Fig.1c) he monthly rainfall trends are robust in different he monthly rainfall trends are robust in different i f lld t t (Fi 2) ainfall datasets (Fig 2) ainfall datasets (Fig.2) he Asian monsoon rainfall has a significant trend he Asian monsoon rainfall has a significant trend ( ) mainly during the transitional season (Fig 3) mainly during the transitional season (Fig.3) h l t t d fb l he long‐term trend of boreal summer mean he long term trend of boreal summer mean ainfall changes significantly depending upon ainfall changes significantly depending upon h h ii ld i f lli (i ) hether it includes rainfall in May (Fig 4) hether it includes rainfall in May (Fig.4). Figure 3: (Left) Latitude time section of trend Figure 3: (Left) Latitude‐time section of trend ( h di ) d li tl ( t ) f CMAP (shading) and climatology (contour) of CMAP (shading) and climatology (contour) of CMAP rainfall along 10–15N Shadings are the same as in rainfall along 10–15N. Shadings are the same as in Fig 1 (Right) Time series of trend (shading top Fig 1. (Right) Time series of trend (shading: top i) d li l (li b i) f i f ll axis) and climatology (line: bottom axis) of rainfall axis) and climatology (line: bottom axis) of rainfall averaged over the area (10 15N 60 140E) averaged over the area (10–15N, 60–140E). MJJAS MJJAS MJJAS igure 2: Linear trends of rainfall in May (mm/day) as igure 2: Linear trends of rainfall in May (mm/day) as ame as Fig 1a but for 4 different rainfall data set: (a) ame as Fig. 1a but for 4 different rainfall data set: (a) JJA MAP (b) GPCP ( ) APHRODITE’ W R JJA MAP, (b) GPCP, (c) APHRODITE’s Water Resources JJA MAP, (b) GPCP, (c) APHRODITE s Water Resources roject and (d) University of Delaware Precipitation roject , and (d) University of Delaware Precipitation. Figure 4: Linear trends of rainfall in (a) May Sep Figure 4: Linear trends of rainfall in (a) May – Sep (MJJAS) d (b) J A (JJA) ( /d ) (MJJAS) and (b) Jun‐Aug (JJA) (mm/day) (MJJAS) and (b) Jun Aug (JJA) (mm/day) Summary Summary Summary We have elucidated the significant seasonality We have elucidated the significant seasonality in long‐term trends in the Asian monsoon in long‐term trends in the Asian monsoon. 1 Increasing rainfall trend in May along 10N 1. Increasing rainfall trend in May along 10N d h l correspond to the earlier monsoon onset correspond to the earlier monsoon onset. 2 R i f llt di Jl dA th b l 2. Rainfall trends in July and August have been less f ff 2. Rainfall trends in July and August have been less d climatology of the difference in significant d climatology of the difference in significant. (30N) and ocean (5N) areas Trends are 3 Th li d k i fh (30N) and ocean (5N) areas. Trends are 3 The earlier monsoon onset and weakening of the Ol th th t i ifi t t 95% 3. The earlier monsoon onset and weakening of the Only those that are significant at 95% monsoon are most likely to be attributed to the Only those that are significant at 95% monsoon are most likely to be attributed to the N (c) Same as (a) but for along 30N N. (c) Same as (a) but for along 30N. stronger heating over land in May and stronger stronger heating over land in May, and stronger h ti th t i l i J th hA t heating over the tropical sea in June through August heating over the tropical sea in June through August. 4 These changes of seasonal land sea heating contrast 4. These changes of seasonal land‐sea heating contrast may be presumed to be related to changes of may be presumed to be related to changes of th i lf i d SST anthropogenic aerosol forcing and SST. anthropogenic aerosol forcing and SST. References References Tetsuzo Yasunari: yasunari@hyarc nagoya u ac jp Tetsuzo Yasunari: [email protected]‐u.ac.jp Yoshiyuki Kajikawa: ykaji@hyarc nagoya‐u ac jp Yoshiyuki Kajikawa: [email protected]‐u.ac.jp Kajikawa Y T Yasunari S Yoshida and H Fujinami 2011: Kajikawa, Y., T. Yasunari, S. Yoshida and H. Fujinami, 2011: d d Advanced Asian monsoon onset in recent Advanced Asian monsoon onset in recent d d G h R L tt b itt d decades. Geophys. Res. Lett., submitted. decades. Geophys. Res. Lett., submitted. Ak ld t f fl Acknowledgement ection of water vapor flux converngece Acknowledgement ection of water vapor flux converngece Thi ki db h Gl b lE i R h 60 100E Only those that are This work is supported by the Global Environment Research 60‐100E. Only those that are This work is supported by the Global Environment Research l tt d Fund of the Ministry of the Environment Japan (A0902) re plotted. Fund of the Ministry of the Environment, Japan (A0902). re plotted.