the high resolution global ocean forecasting system in the
TRANSCRIPT
![Page 1: The High Resolution Global Ocean Forecasting System in the](https://reader036.vdocuments.net/reader036/viewer/2022070610/62c347cf5017bc5f12264114/html5/thumbnails/1.jpg)
The High Resolution Global Ocean Forecasting System in the NMEFC and its Intercomparison with the GODAE
OceanView IV-TT Class 4 Metrics
1
Liying Wan (Group Leader)
Yu Zhang, Huier Mo, Ziqing Zu, Yinghao Qin, Sisi Qin
National Marine Environmental Forecasting Center,
State Oceanic Administration, China
System design&
Model testing&
Numerical products application
System assessment Data assimilation
![Page 2: The High Resolution Global Ocean Forecasting System in the](https://reader036.vdocuments.net/reader036/viewer/2022070610/62c347cf5017bc5f12264114/html5/thumbnails/2.jpg)
Content
➢Background
➢Development of NMEFC-NEMO
➢Assessment of NMEFC-NEMO
➢Conclusion
➢Any suggestions2
![Page 3: The High Resolution Global Ocean Forecasting System in the](https://reader036.vdocuments.net/reader036/viewer/2022070610/62c347cf5017bc5f12264114/html5/thumbnails/3.jpg)
First Generation of Global Ocean Forecasting System based on MOM4
Grid: Global 1/4°×1/4°; 50 levels
Bathymetry: OCCAM 0.2
Initialization: OMIP_NCAR
Data Assimilation: 3DVar
Parameterizations: non-Boussinesq approximation (Greatbatch et al.,
2001); Smagorinsky viscosity scheme (Griffies and Hallberg, 2000); KPP
scheme (Large et al., 1994)
Atmospheric forcing: GFS (Global Forecast System) 6-hour data
With the higher requirement for the ocean forecasting products,
we further developed a high resolution global ocean forecasting
system to provide the higher standard, more sophisticated and
qualified products.
![Page 4: The High Resolution Global Ocean Forecasting System in the](https://reader036.vdocuments.net/reader036/viewer/2022070610/62c347cf5017bc5f12264114/html5/thumbnails/4.jpg)
2. 海洋预报业务介绍Horizontal grid
Levels depth and thickness
Topography
NMEFC High Resolution Forecasting System based on NEMO
Code --NEMO3.6 + LIM3Grid-- ORCA tripolar grid (Madec and Imbard[1996])Horizontal resolution
– 4322 x 3059 horizontal grid points – Grid spacing from 10 km at equator down to 3
km at high latitudesVertical grid
– 75 levels, with a resolution of 1m near the surface and 200m in the deep ocean,0-6000m
Model configuration
4
![Page 5: The High Resolution Global Ocean Forecasting System in the](https://reader036.vdocuments.net/reader036/viewer/2022070610/62c347cf5017bc5f12264114/html5/thumbnails/5.jpg)
The Forecasting System Design
NCEP GFS ArgoECMWF GFS
D a t a c o l l e c t i o n s y s t e m
D0 D10D-3
D0D-3
D0D-3
D0D-3
Data Assimilation Forecasting
Forecast start
Restart Source
D10
D10
D10
Forecasting Start
Assimilation Forecasting
IVTTProducts
Static Forecast figures
AnimationProducts
Compress Archives
Site and Regional improved products
F o r e c a s t p r o d u c t s r e l e a s e s y s t e m
3DVAR
Data assimilation system
Nudging
Satellites Drifts Reanalysis
5
![Page 6: The High Resolution Global Ocean Forecasting System in the](https://reader036.vdocuments.net/reader036/viewer/2022070610/62c347cf5017bc5f12264114/html5/thumbnails/6.jpg)
Data Assimilation System
SLA(AVISO CMEMS)
SST(OSTIA UKMO)
ARGOTS Profiles
3 Days Hindcast
run
1st day 2nd day 3rd day
Observation information input
3DVar(developing), Nudging
AssimilationIncrement
ErrorCalculate
Forecasting
Monthly model error
covariance
Monthly vertical TS relationshi
p
Numerical model
characteristics information
SLA
SST
ARGO
Multiple sateillates(5) reconstruction
Multiple sateillates(9) reconstruction
More than 600 profiles everyday
![Page 7: The High Resolution Global Ocean Forecasting System in the](https://reader036.vdocuments.net/reader036/viewer/2022070610/62c347cf5017bc5f12264114/html5/thumbnails/7.jpg)
Artificial Neural Networks
F-60 F-59 F-2F-58 F-1 O-60 O-59 O-2O-58 O-1
History Forecast
Global Grid Mapping
Improved Forecasting Products
History Observation
DEEP Learning
Regional and site bias relationship build
![Page 8: The High Resolution Global Ocean Forecasting System in the](https://reader036.vdocuments.net/reader036/viewer/2022070610/62c347cf5017bc5f12264114/html5/thumbnails/8.jpg)
Example for SLA
SSH SDH
SLA
Wavelet analyse
mesoscale info
Large Scale info
Correct Separately
![Page 9: The High Resolution Global Ocean Forecasting System in the](https://reader036.vdocuments.net/reader036/viewer/2022070610/62c347cf5017bc5f12264114/html5/thumbnails/9.jpg)
2. 海洋预报业务介绍
Products part one:GODAE-IVTT standard productsClass 1:gridded model outputClass 2:time series of specified locations and sectionsClass 3:transport through sections and other quantitiesClass 4:metrics of forecast capability
Products pat two:Conventional static visualization products
Products part three:Animation products
Sea ice
Vertical profile Temperature, salt, current, and SSH
Forecast products release system
9
Eddies activities and variation
![Page 10: The High Resolution Global Ocean Forecasting System in the](https://reader036.vdocuments.net/reader036/viewer/2022070610/62c347cf5017bc5f12264114/html5/thumbnails/10.jpg)
NMEFC-NEMO
GLORYS
Temperature Salinity
Climatologic kinetic energy with and without eddy
SST(NEMO-GLORYS) SSS(NEMO-GLORYS)
Ten years climatologic assessment10
This high resolution ocean forecasting system can
basically reproduce the oceanic characteristics.
![Page 11: The High Resolution Global Ocean Forecasting System in the](https://reader036.vdocuments.net/reader036/viewer/2022070610/62c347cf5017bc5f12264114/html5/thumbnails/11.jpg)
Regional comparison of operational systemSST of South China Sea:RMSE 0.4℃,Bias ±0.3 ℃, against PSY Reanalysis Data
Sep, 2016 Oct, 2016 Nov, 2016 Dec, 2016 Jan, 2017
![Page 12: The High Resolution Global Ocean Forecasting System in the](https://reader036.vdocuments.net/reader036/viewer/2022070610/62c347cf5017bc5f12264114/html5/thumbnails/12.jpg)
Oct, 2
016
No
v, 2016
Jan
, 2017
Regional comparison of operational systemSST of Indian Ocean:RMSE 0.5℃, against PSY Reanalysis Data
Dec, 2
016
![Page 13: The High Resolution Global Ocean Forecasting System in the](https://reader036.vdocuments.net/reader036/viewer/2022070610/62c347cf5017bc5f12264114/html5/thumbnails/13.jpg)
Regional comparison of operational systemSST of Indian Ocean:RMSE -0.6-0.1℃, against PSY Reanalysis Data
Oct, 2
016
No
v, 2016
Jan
, 2017
Dec, 2
016
![Page 14: The High Resolution Global Ocean Forecasting System in the](https://reader036.vdocuments.net/reader036/viewer/2022070610/62c347cf5017bc5f12264114/html5/thumbnails/14.jpg)
Introduction of IV-TTBackground:
The GODAE Oceanview Intercomparison and Validation Task Team (IV-TT)
aims to coordinate and promote the development of scientific validation
and intercomparison of operational oceanographic systems. A particular
focus is a real-time intercomparison of GOV systems using the Class 4
framework.
Workshop objectives:
Review the current status of the Class4 intercomparison and discuss
outstanding issues
Discuss the future directions of this intercomparison
Discuss ongoing initiatives and other additional possible assessments
Discuss recent advances in evaluation and verification efforts, and the
development of new metrics 14
![Page 15: The High Resolution Global Ocean Forecasting System in the](https://reader036.vdocuments.net/reader036/viewer/2022070610/62c347cf5017bc5f12264114/html5/thumbnails/15.jpg)
Joint systems
15
country System name modelEnvironment Canada GIOPS-CONCEPTS NEMO 1/4Mercator Ocean PSY3 & PSY4 NEMO 1/4 &1/12Australian Bureau of Meteorology BLUEllink-OceanMAPS MOM4 10km around Australia
but low resolution further out
NOAA/NCEP/NWS RTOFS Global HYCOM 1/12UK Met Office FOAM NEMO 1/4
Five systemsand 6 models,4 of themodels arebased onNEMO.
![Page 16: The High Resolution Global Ocean Forecasting System in the](https://reader036.vdocuments.net/reader036/viewer/2022070610/62c347cf5017bc5f12264114/html5/thumbnails/16.jpg)
IVTT Class4 assessment method
• 𝐵𝑖𝑎𝑠 =1
𝑁σ𝑖=1𝑁 (𝐹𝑖 − 𝑂𝑖)
• 𝑅𝑀𝑆𝐸𝐹𝐶𝑇 =1
𝑁σ𝑖=1𝑁 (𝐹𝑖 − 𝑂𝑖)
2
• 𝑅𝑀𝑆𝐸𝑝𝑒𝑟 =1
𝑁σ𝑖=1𝑁 (𝐵𝑖 − 𝑂𝑖)
2
• 𝑃𝑆𝑆 = 1 −𝑅𝑀𝑆𝐸𝐹𝐶𝑇
𝑅𝑀𝑆𝐸𝑝𝑒𝑟
• 𝐶𝑆𝑆 = 1 −𝑅𝑀𝑆𝐸𝐹𝐶𝑇
𝑅𝑀𝑆𝐸𝑐𝑙𝑖𝑚
• 𝐴𝐶 =σ(𝐹−𝐶)(𝑂−𝐶)
σ(𝐹−𝐶)2 σ(𝐹−𝑂)2
16
𝑭—forecast 𝑶—observation C—climatology B—Best estimate
Model deviate from observation (smaller is good)
Persistence skill score: numerical forecast VS analysis data extrapolation (>0 is good)
Climatology skill score: numerical forecast VS Climatological data (>0 is good)
Anomaly correlation: weather variations capture ability
A.G. Ryan, C. Regnier, P. Divakaran, T. Spindler, A. Mehra, G.C. Smith, F.Davidson, F. Hernandez, J. Maksymczuk & Y. Liu (2015) GODAE OceanView Class 4 forecastverification framework: global ocean inter-comparison, Journal of Operational Oceanography,8:sup1, s98-s111, DOI: 10.1080/1755876X.2015.1022330
![Page 17: The High Resolution Global Ocean Forecasting System in the](https://reader036.vdocuments.net/reader036/viewer/2022070610/62c347cf5017bc5f12264114/html5/thumbnails/17.jpg)
SST
17
Bias are higher than FOAM: different assimilation methods
are used
There is no big difference in the persistence skill
Better performance of climatology skills in each system
![Page 18: The High Resolution Global Ocean Forecasting System in the](https://reader036.vdocuments.net/reader036/viewer/2022070610/62c347cf5017bc5f12264114/html5/thumbnails/18.jpg)
SLA
RMSE of SLA about 5cm to 8cm
The persistence skill is strongly negative for each system
For CSS, our system is the worst one
![Page 19: The High Resolution Global Ocean Forecasting System in the](https://reader036.vdocuments.net/reader036/viewer/2022070610/62c347cf5017bc5f12264114/html5/thumbnails/19.jpg)
RMSE of Temperature Profile
19
Comment features:
About 0.4˚C on average
Largest errors at 100m
Uncertainty in vertical position of thermocline
Significant temperature errors
![Page 20: The High Resolution Global Ocean Forecasting System in the](https://reader036.vdocuments.net/reader036/viewer/2022070610/62c347cf5017bc5f12264114/html5/thumbnails/20.jpg)
RMSE of Salinity Profile
Comment features:
About 0.1 psu on average
Largest errors at 100m-200m
Our system is on
average of all
![Page 21: The High Resolution Global Ocean Forecasting System in the](https://reader036.vdocuments.net/reader036/viewer/2022070610/62c347cf5017bc5f12264114/html5/thumbnails/21.jpg)
Conclusion
NMEFC NEMO global operational forecasting system:
runs stably and provides 5-day forecasting
working on the assimilation system and a 3-DVar assimilation schemes
Evaluation of the products
temperature and salinity is comparable to other systems
improve the forecasting of thermocline by a better 3-DVar assimilation
system
Forecasting skills:
short term forecasting: data assimilation system is more important
longer forecasting: forecasting system with an higher frequency output
plays an important role21
![Page 22: The High Resolution Global Ocean Forecasting System in the](https://reader036.vdocuments.net/reader036/viewer/2022070610/62c347cf5017bc5f12264114/html5/thumbnails/22.jpg)
Thanks for your attention
http://www.nmefc.gov.cn
22