Model Ens Initialization Interval Reforecast period Source SubX (NCEP-GEFS, NASA-GEOS5, ESRL- FIM, Navy-ESPC, RSMAS-CCSM4, NCAR-CESM1) 3~20 5~7 days 1999-2016 SubX ECMWF Cy43R3 11 2/week 1997-2016 S2S UKMO/KMA GloSea5 3 7 days 1991-2010 S2S - RMM index (Wheeler and Hendon, 2004) - ERAI, NOAA CDR OLR, GPCP precipitation (1979~2017) → “OBS” - NH winter (Oct-March) SubX and S2S reforecasts MJO Propagation Processes and Mean Biases in the SubX and S2S Reforecasts Hyemi Kim (SBU), Matthew A. Janiga (NRL) and Kathy Pegion (GMU) (JGR-Atmos, under review) SubX, 07/10/2019
20
Embed
MJO Propagation Processes and Mean Biases in the SubX and …cola.gmu.edu/subx/docs/SubX_MJO_Kim_07102019.pdf · •Common biases in SubX and S2S reforecasts: - Convection starts
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
* MJO event selection- Initial Phase 1 & 2- Initially Strong MJOs- ( ): Number of events- Ensemble mean
OLR & 850hPa Wind ano
MJO Propagation
* MJO detours the Maritime Continent southward during boreal winter(e.g., Lau and Chen 1985, D. Kim et al. 2017)
Forecast DAY 10Forecast DAY 10
OLR & 850hPa Wind ano
MJO Propagation
MJO propagation skill
PCC(τ) (40-200E, 20S-20N)
DAY 10:
Q: What process impacts the fast decay of the MJO propagation signal?
Pattern Correlation Coefficient (PCC)
Bretherton et al. (2004)
Precipitable water, <q> (mm)
Pre
cip
itat
ion
(m
m d
ay-1
)
Daily mean Precipitation vs. PW (Tropical Ocean)
• Tropical deep convection occurs preferentially over the area of high column moisture
• Once the moisture (precipitable water) exceeds some threshold value, precipitation increases rapidly.
• MJO is distinguished from other waves by the strong moisture-convection coupling.
MJO: Moisture-Convection Coupling
“Moisture mode” theory for MJO
𝝏 𝒒
𝝏𝒕= − 𝑽 ∙ 𝜵𝒒 − 𝜔
𝜕𝑞
𝜕𝑝+ … .
• Growth, decay, and propagation of MJO convection are explained by those of anomalous moisture (or MSE).
• Colum-integrated moisture budget:
* Yu and Neelin (1994), Sobel et al. (2001), Raymond and Fuchs (2008), Raymond et al. (2009), Maloney (2009), Sobel and Maloney (2012, 2013), Jiang et al. (2015, 2017), Adames and Kim (2016)
− 𝑽′ ∙ 𝜵ഥ𝑸MJO wind Mean moisture
→ Advection of seasonal mean moisture( ത𝑄) by anomalous MJO circulation (V’) controls the MJO eastward propagation
Active MJOSuppressed
MJO
Winter mean moisture (Q850)
Observation
MJO Moisture advection MJO propagation
− 𝑽′ ∙ 𝜵ഥ𝑸
Moisture Advection
Day 01 Day 10
Eastward Propagation Process
OBS
Multi-model mean:Weaker
10 days meanMoisture advection (−𝑉′ ∙ 𝛻 ത𝑄)
Eastward Propagation Process
Mean State Bias: Q850
OBS
Bias (MMM-OBS)
Dry bias in the lower-
troposphere
Drier Wetter
Winter mean moisture (Q850)
(Less steep meridional Q gradient)
Mean State Bias: Q850
More rain
Drier Wetter
Precipitation (mm/day)
Mean State Bias: Precipitation
OBS
Bias (MMM-OBS)
Mean State Bias: Precipitation
Precipitation distribution
• Precip. days/Total days • 4-weeks average• [60E-180E, 15S-15N] • Control simulation • Land area excluded
→Models have too frequent light rain (drizzle)
Light Moderate Heavy
Precipitation rate (Tropical Oceans)
Moisture-Precipitation Relationship
→ Convection starts too early and occurs too frequently in the low moisture regime
Specific Humidity vs. Precipitation (Tropical Oceans)
→ Deep convection is not sufficiently inhibited when tropospheric moisture is low, which is likely due to the representation of entrainment.
Summary
• Common biases in SubX and S2S reforecasts:
- Convection starts too early in the low moisture regime
- Light precipitation occurs too frequently
→ Excess of surface precipitation & Drier lower troposphere
→Weaken the moisture advection process − 𝑉′ ∙ 𝛻 ത𝑄
→Weaken the MJO eastward propagation signal
→ Limit MJO prediction skill
Convection starts too early Dry bias MJO prediction skill