Shallow orographic heavy rainfall in the Asian monsoon region observed by TRMM PR Shoichi Shige 1 & Christian D. Kummerow 2 1 Graduate School of Science, Kyoto University 2 Dept. of Atmospheric Science, Colorado State University
Background India [Shige et al. 2014] TMI TRMM Microwave Imager PR GSMaP GPROF PR Precipitation Radar APHRODITE GSMaP-PR GPROF PR
Background (Continued) MWR rainfall algorithms assume that heavy rainfall results from deep clouds. Weak Scattering Light Rainfall cumulus stage Strong Scattering Heavy Rainfall mature stage dissipating stage Wallace & Hobbs (2006) [Adapted from Byers and Braham (1949)]
Background (Continued) However, heavy rainfall can be caused by shallow orographic convection in moist Asian monsoon regions. cumulus stage mature stage mature stage Obs. TB Original LUT@85 GHz Orographic Underestimate Wallace & Hobbs (2006) [Adapted from Byers and Braham (1949)] Est. Rain
Background India [Shige et al. 2014] PR GSMaP Ocean Land Orographically Forced Upward Motion w = Dh Dt = u h x + v h y APHRODITE GSMaP-PR
Background India [Shige et al. 2014] PR GSMaP GSMaP(revised) APHRODITE GSMaP-PR GSMaP(revised) PR
What environmental thermodynamic parameters constrain the depth of orographic rain clouds???????
Arakan Yoma Phillipines Western Ghats Bilauktaung Cardamom Annam Cordillera Murakami & Matsumoto (1994, JMSJ) Mexico
Arakan Yoma Phillipines 18 o N Western Ghats Bilauktaung Cardamom Annam Cordillera Logtitude-height sections of the 5-year unconditional mean Q 1 Q R for the June August season Mexico I II V VI
Fig. 2: Schematic of target region and upstream region. z = 7.5 km TRMM PR: 0.05 o ERA-Interim: 0.75 o SRTM30: the original 1-km resolution data averaged within the horizontal length scale of 50km Dh w oro = V Dt = H Low-level upstream region h z = 4.5 km z = 2.5 km ERA-Interim 0.75 o Target region Mid-level upstream region
The 7-yr conditional mean Q 1 Q R and RH profiles stratified with dp/dt at 500 hpa averaged over the ocean, associated with (a) all rain, for all rain averaged over the grids with SST of 28.0 29.0 C. Takayabu ( et al. 2010 JC) Q 1 Q R RH Ocean DRY MOIST
Fig. 6: Precipitation profiles with heavy near-surface rain rates (10 mm h -1 < R 40 mm h -1 ) and upward motion (w oro > 0 m s -1 ) as a function of RH low. Contours indicate values of the confidence interval for the mean at the 95 % level with Student s t test. RH low Correlation coefficients with PTH(1 mm h -1 ) -0.91-0.71-0.80-0.75-0.97-0.19
Fig. 6: Precipitation profiles with heavy near-surface rain rates (10 mm h -1 < R 40 mm h -1 ) and upward motion (w oro > 0 m s -1 ) as a function of RH mid. Contours indicate values of the confidence interval for the mean at the 95 % level with Student s t test. RH mid Correlation coefficients with PTH(1 mm h -1 ) -0.94-0.68-0.76-0.77-0.39-0.68
What determines deep or shallow? Ocean DRY MOIST Mountain DRY MOIST
Fig. 10: As in Fig. 3 but for precipitation profiles with near-surface rain rates of 10 mm h -1 < R 40 mm h -1 and upward motion (w oro > 0 m s -1 ) as a function of dt/dz low. dt/dz low Correlation coefficients with PTH(1 mm h -1 ) -0.90-0.90-0.61-0.83-0.78-0.92
Fig. 9: As in Fig. 3 but for precipitation profiles with heavy near-surface rain rates (10 mm h -1 < R 40 mm h -1 ) and upward motion (w oro > 0 m s -1 ) as a function of dt/dz mid. dt/dz mid Correlation coefficients with PTH(1 mm h -1 ) -0.95-0.71-0.82-0.83-0.97-0.89
Fig. 13: Jun-Aug climatology (2004-2008) of dt/dz mid with horizontal winds at z=5.5 km (top) and function of dt/dz low with horizontal winds at z=2.5 km (bottom) from ERA Interim. dt/dz mid dt/dz low I II V VI Low-level or midlevel static stability determines precipitation depths?
Summary Heavy rainfall can be caused by shallow convection over mountain ranges of the Asian monsoon regions, which differs from the assumptions of MWR algorithms. Depth of heavy orographic rainfall decreases with RH in opposition to what has been observed for oceanic convection. Depth of heavy orographic rainfall also decreases with the lowlevel and mid-level static stability. Mid-level static stability, which inhibit cloud growth and promotes cloud detrainment, is inferred to be a key parameter determining precipitation depths. Low-level static stability, however, might be more effective to improve microwave radiometer rainfall algorithms, because of a result of shallow convection and their strong links.
Fig. 11 Comparison of correlation coefficients between candidate thermodynamic quantities and precipitation top heights
Fig. 8: Two dimensional histogram of grid with heavy rain rates (10 mm h -1 < R 40 mm h -1 ) as function of w oro and RH mid for Western Ghats, Arakan Yoma, and Bilauktaung.
Fig. 8: Two dimensional histogram of grid with heavy rain rates (10 mm h -1 < R 40 mm h -1 ) as function of w oro and RH mid for Western Ghats, Arakan Yoma, and Bilauktaung.
Arakan Yoma Phillipines Western Ghats Bilauktaung Cardamom Annam Cordillera Mexico I II III IV V VI I V
I II V VI I: Western Ghats II: Arakan Yoma III: Bilauktaung IV: Cardamom V: Annam Cordillera VI: Phillipines I III IV V
What determines deep or shallow? Ocean DRY MOIST Mountain DRY UNSTABLE MOIST STABLE
Kato et al. (2007, JMSJ) The relatively stable atmospheric condition around the Japan Islands, found when the LNB appears at the MP, is mainly caused by the advection of a middle-level air that is warmed through convective activities over upstream regions, especially the China Continent.
Murakami & Matsumoto (1994, JMSJ)
Kummar et al. (2014, CD) Jun-Jul Deeper clouds appear more frequently over the Myanmar coast, while shallower clouds over the Western Ghats. Aug-Sep
Arakan Yoma Phillipines Western Ghats Bilauktaung Cardamom Annam Cordillera Mexico
Original Look-up table: Relationship between Tbs and rain rates Shige et al. (2013, JAMC) Obs. TB Original LUT@85 GHz Orographic Orographic Typhoon Namtheun Underestimate Est. Rain
Orographic/Non-Orographic Rainfall Classification Scheme GANAL SRTM30 Ocean Land Yes Orographic LUTs Retrieval w > 0.1 Q > 0.5 10-6 Shige et al. (2013, JAMC) Taniguchi et al. (2013, JH) No Original LUTs Retrieval Orographically Forced Upward Motion w = Dh Dt h = u x h + v y Convergence of Surface Moisture Flux ( uq) ( vq) Q = - + Ł x y ł
Typhoon Namtheun (2004) over Japan (Kubota et al. 2009, JMSJ) Radar-AMeDaS GSMaP TMI PCT85 PR GPROF PR PTH (km)
Arakan Yoma Phillipines Western Ghats Bilauktaung Cardamom Annam Cordillera Wang & LinHo (2002, JC) Mexico
Fig. 1: Location of mesoscale mountain ranges (Western Ghats, Arakan Yoma, Bilauktaung, Cardamom, Annam Cordillera and the Phillipines) of the South Asian monsoon region (Fig 1a) and Mexico for comparison (Fig. 1b). Western Ghats Arakan Yoma Bilauktaung Cardamom Annam Cordillera the Phillipines Baja California
Fig. 11-0.72-0.69-0.78-0.85-0.84-0.89
Typhoon Morakot (2009) over Taiwan: Taniguchi et al. (2013, JHM) Dh w oro = V Dt = H h
TRMM (Nov 1997~) TMI TRMM Microwave Imager (a) Background: TRMM satellite overpass of Typhoon Morakot over Taiwan (Taniguchi et al. 2013, JHM) SRTM30 (b) PR PR Precipitation Radar (c) (d) http://www.boston.com/bigpicture/2009/08/typhoon_morakot.html
TRMM (Nov 1997~) TMI TRMM Microwave Imager (a) Background: TRMM satellite overpass of Typhoon Morakot over Taiwan (Taniguchi et al. 2013, JHM) SRTM30 (b) PR PR Precipitation Radar (c) GPROF (TMI) (d) GSMaP (TMI)
PR _ Near Surface Rain Typhoon Morakot (2009) over Taiwan Taniguchi et al. (2013, JHM) GSMaP_TMI_Surface Rain (orig.) Applied orographic profile GSMaP_TMI _Surface Rain w > 0.1 (ms -1 ) Q > 0.5 10-6 (s -1 ) SRTM30_Topography Orographic Forced Lifting Convergence of Surface Moisture Flux