Post on 30-Dec-2015
description
Japan ReportNational Forecasting System
(MOVEMRI.COM)and Japan Working Team- Progress in 2012-2014 -
T. Kuragano (GOVST)M. Kamachi (GOVPatron)
Y. Fujii(OSE-TT), N. Usui (COSS-TT), S. Ishizaki (ET-OOFS),
Y. Takaya, T. Toyoda, K. SakamotoJMA/MRI
Recent Progress (1/2)
• MOVE/MRI.COM-G2 and -SETO systems will be in operation early 2015.
• OSE for Argo floats and TAO/TRITON array was done (Fujii et al., 2014).
• Impact of Aquarius SSS data was examined (Toyoda et al., in revision).
• Seawater mass variation was examined to detect steric component from altimeter data (Kuragano et al., 2014).
Recent Progress (2/2)
• Analysis and predicted variables have been available since August 2014 for research and commercial users through the Japan Meteorological Business Support Center.
• Application studies are in progress in collaboration with related organizations. – Fishery, Debris, and Radiation contamination
• 18th Japan data assimilation summer school was held in August 2014 as GOV Outreach activity.
MOVE/MRI.COM-G2 for Seasonal/ENSO Forecast
• MRI.COM-G2 is an ocean part of the coupled model, JMA/MRI-CGCM2, for the next operational seasonal/ENSO forecast system in JMA
• Ocean Reanalysis was done by the assimilation system MOVE/MRI.COM-G2
• System performance was evaluated from the hindcast experiments
MRI.COM v3.2• Tripolor grid and Sea Ice model
- analysis and prediction in arctic area and sea ice• Higher latitudinal resolution
- 0.3 - 1.0 degree → 0.3 - 0.5 degree• Atmospheric forcing - JRA25/JCDAS -> JRA55
MOVE major revisions• Change statistical vertical EOF modes
- 57 regions from 40 ones in previous version - Monthly EOF modes from annual EOF mode
• Bias correction scheme• FGAT ( First-Guess at Adequate Time) • Ocean mass correction for altimeter data• T/S assimilation up to 1750 m from up to 1500 m
Lines are drawn for every 10 grids
Main Upgrades of System- MOVE/MRI.COM-G2 -
NINO3.4 SST prediction NINO3.4 RMSE (Init: Feb)
JMA/MRI-CGCM2
JMA/MRI-CGCM1(2009 experiment)
JMA/MRI-CGCM1(2014 experiment extended for LT 7-12 months)
February and November initialized predictions show remarkable improvement: ACC keeps high beyond the Spring predictability barrier
[K]
NINO3.4 ACC (Init: Feb)
NINO3 SST RMSE (Init: Nov)[K]
NINO3 SST ACC (Init: Nov)
MOVE/MRI.COM-SETO for coastal high tide
prediction
MOVE-WNP 4DVAR• 0.1x0.1 deg., 54 layers • 10-day assimilation
window• To predict short-term
open-ocean variation
SETO Coastal Model• 2km x 2km 54 layers• increment from WNP
4DVAR• To represent detailed
impact on the coastal area
Target: To predict costal high tide and rapid current caused by open ocean variation
8
Study for the unusual high tide in 2011
ObsModel
SSH on 29SEP2011
AB ST MI
TyphoonTyphoon
Assimilation experiment: 2km-coastal assimilative model 1 Aug – 31 Oct 2011
The Kuroshio took a nearshore path at the eastern flank of the Izu Ridge
Izu Ridge
Itsukushima Shrine
Unusual high tide
Kanto
Observation SST SLA In Situ TAO/TRITON ARGOType Eq Ex 0-1 2-3 4-5 6-7 8-9
OSE-XBO ○ ○ ○ OSE-XAF ○ ○ ○ ○ ○ OSE-AR2 ○ ○ ○ ○ ○ ○ OSE-AR4 ○ ○ ○ ○ ○ ○ ○ OSE-AR6 ○ ○ ○ ○ ○ ○ ○ ○ OSE-AR8 ○ ○ ○ ○ ○ ○ ○ ○ ○ OSE-XTT ○ ○ ○ ○ ○ ○ ○ OSE-TTeq ○ ○ ○ ○ ○ ○ ○ ○ REG-Exp ○ ○ ○ ○ ○ ○ ○ ○ ○ ○
Last digit of WMO number
Reference Data
Observed Data Assimilated in Each OSEs
OSE for ARGO and TAO/TRITON
9 simulation experiments are performed in the period of 2000-2012 using MOVE-G.
Profiles of Argo where the last digit of the WMO number is 8 or 9 are withheld in all simulations except for Reg-Exp, and used for the reference data.
OSE Configuration
Increase of ACC against OSE-XBO (0-300m average)
The accuracy of TS fields is generally increased with the increase in the number of assimilated Argo profiles.
The complementary impact of TRITON on T is larger than that of Argo floats in the western tropical Pacific (OSE-XAF has higher T accuracy than OSE-XTT).
The increase of ACCs for both TS with the increase of the number of Argo assimilated is closer to linear in NINO3 (Enhanced deployment of Argo floats is desirable).
SSS impact on MOVE/MRI.COM
• Impact of Aquarius SSS data on MOVE/MRI.COM-G2 are examined for future operational use of SSS
• Aquarius official release level 3 SSS standard mapped image daily data v2.0
• CTL Exp.: assimilation run for T/S profile and SLA data• ASA Exp.: SSS data are additionally assimilated
Impacts in the North Pacific• Increase of subsurface salinity and temperature especially
in winter
• Strengthen of mixing at surface layer, which is consistent with in-situ data based analysis
S 100 m
T 100 m
PV 100 m
MLD CTL
MLD ASA
In-situ basedAnalysis
Ocean mass correction for SLA
• SL variation caused by ocean mass variation is examined to steric height component from altimetric SLA.
• Local mass variation as a response to seasonal water flux, wind stress and surface pressure variation is examined using a barotropic global ocean model.
• The results are well consistent with seasonal mean variation of altimeteric SLA minus steric height.
• The results are adopted for the correction of SLA data in MOVE/MRI.COM-G2.
Altimetric SL Kuragano & Kamachi (2000)
Steric SL Ishii and Kimoto (2009)
mass
Amplitude of SL variation by mass
Phase of SL variation by mass
Cost function:
Ocean mass correction for SLA
2
22
2
)()(2
1
2ccchxdh
cJJJ trendBT
jojj
jcTSobsbackground
Model steric height
Altimetric SLA
Seasonal mass-related SSH
Trend of global mean mass-related SSH:2mm/yr
ojh:SSHAltimetric
trendBTj
oj cch :SSHcorrectedmass
2
2)(
2
1 ojj
jTSobsbackground hxdhJJJ
TSobsbackground JJJ
41010c4105c4103c
Assimilation including mass correction terms
Results of global mean SSH variation by applying mass correction
18th Data Assimilation Summer School
19 - 22 Aug. 2014, in Mutsu, Aomori Prefecture
As an Outreach of GOVIt continues from 1997 under the support of Japan Marine Science Foundation & JAMSTEC.3 days course of fundamental lectures, practices and applications
Participants
BBQ Party
Lectures