Upload
others
View
7
Download
0
Embed Size (px)
Citation preview
1
Isaac Ginis1, Richard Yablonsky1, Biju Thomas1, Morris Bender2, and Vijay Tallapragada3
1University of Rhode Island 2NOAA/OAR/GFDL 3NOAA/NWS/NCEP/EMC
68th Interdepartmental Hurricane ConferenceTuesday, March 4, 2014
2
MPIPOM‐TC uses MPI software to run efficiently on multiple processors, allowing for both higher grid resolution and a larger ocean domain than POM‐TC
MPIPOM‐TC accepts flexible initialization options MPIPOM‐TC is an adaptation of sbPOM, which has community support and includes 18 years of physics updates and bug fixes
MPIPOM‐TC is a modernized code with NetCDF I/O MPIPOM‐TC uses a single prognostic code in all worldwide GFDL, GFDN, and HWRF ocean basins
GFDL/N & HWRF Ocean Model Upgrade: Replacing POM‐TC with MPIPOM‐TC
3
POM‐TC Atlantic Domains:“United” and “East Atlantic”
225
157
47.5°N
10.0°N30°W60°W
~18-km gridspacing
254 50°W98.5°W
4
URI’s MPIPOM‐TC transatlantic domain47.5°N
449
10.0°N
~9 kmhorizontal
grid spacing
15.3°W98.5°W 869
5
MPIPOM‐TC vs. POM‐TC: Ocean response to Hurricane Katia (through 2011090800)
MPIPOM-TC SST (full domain) POM-TC SST (full domain) MPIPOM-TC – POM-TC SST
MPIPOM-TC SST (wake only) POM-TC SST (wake only) 20110908 TMI 3-day SST
More cooling in MPIPOM-TC
MPIPOM-TC agrees with TMI TMI is “observed” SST product
6
Hurricane Sandy 18L: 2012102506
HWRF/MPIPOM-TC (blue)& HWRF/POM-TC (red)tracks are similar, but missthe left recurvature late
HWRF/MPIPOM-TC (blue)vs. HWRF/POM-TC (red)max wind: noisy, but bestfit is HWRF/MPIPOM-TC
HWRF/MPIPOM-TC (blue)vs. HWRF/POM-TC (red)pressure: best forecast isHWRF/MPIPOM-TC
HWRF/MPIPOM-TC& HWRF/POM-TCtrack standard deviations:2012102506 forecastis highlighted red
HWRF/MPIPOM-TC& HWRF/POM-TCmax wind std dev
HWRF/MPIPOM-TC& HWRF/POM-TCpressure std dev
7
MPIPOMTC SST + Currents POM-TC SST + Currents MPIPOMTC – POM-TC SST
48h 48h 48h
Hurricane Sandy 18L: 2012102506
More cooling in MPIPOM-TC:Better intensity?
8
Hurricane Sandy 18L: 2012102612
HWRF/MPIPOM-TC (blue)& HWRF/POM-TC (red)tracks: HWRF/MPIPOM-TCcaptures recurvature better
HWRF/MPIPOM-TC (blue)vs. HWRF/POM-TC (red)max wind: noisy with noclear “winner”
HWRF/MPIPOM-TC (blue)vs. HWRF/POM-TC (red)pressure: very similaruntil landfall at the end
HWRF/MPIPOM-TC& HWRF/POM-TCtrack standard deviations:2012102612 forecastis highlighted red
HWRF/MPIPOM-TC& HWRF/POM-TCmax wind std dev
HWRF/MPIPOM-TC& HWRF/POM-TCpressure std dev
9
Hurricane Sandy 18L: 2012102612
Better MPIPOM-TC Gulf Streamevolution improved track?
MPIPOMTC SST + Currents POM-TC SST + Currents MPIPOMTC – POM-TC SST
10
Hurricane Katia 12L: 2011083018
HWRF/MPIPOM-TC (blue)& HWRF/POM-TC (red)tracks are similar
HWRF/MPIPOM-TC (blue)vs. HWRF/POM-TC (red)max wind: best forecastis HWRF/MPIPOM-TC
HWRF/MPIPOM-TC (blue)vs. HWRF/POM-TC (red)pressure: best forecast isHWRF/MPIPOM-TC
HWRF/MPIPOM-TC& HWRF/POM-TCtrack standard deviations:2011083018 forecastis highlighted red
HWRF/MPIPOM-TC& HWRF/POM-TCmax wind std dev
HWRF/MPIPOM-TC& HWRF/POM-TCpressure std dev
11
Hurricane Katia 12L: 2011083018
MPIPOMTC SST + Currents POM-TC SST + Currents MPIPOMTC – POM-TC SST
Coupling east of 50°W in MPIPOM-TC: Better
intensity & cooling overall?
12
Hurricane Earl 07L: 2010082512
HWRF/MPIPOM-TC (blue)& HWRF/POM-TC (red)tracks are nearly identical
HWRF/MPIPOM-TC (blue)vs. HWRF/POM-TC (red)max wind: best forecastis HWRF/MPIPOM-TC
HWRF/MPIPOM-TC (blue)vs. HWRF/POM-TC (red)pressure: best forecast isHWRF/MPIPOM-TC
HWRF/MPIPOM-TC& HWRF/POM-TCtrack standard deviations:2010082512 forecastis highlighted red
HWRF/MPIPOM-TC& HWRF/POM-TCmax wind std dev
HWRF/MPIPOM-TC& HWRF/POM-TCpressure std dev
13
Hurricane Earl 07L: 20100825 12 & 18 Z
MPIPOMTC SST + Currents POM-TC SST + Currents MPIPOMTC – POM-TC SST
Coupling eastof 50°W in
MPIPOM-TC:better intensity?
14
Proposed Worldwide MPIPOM‐TC domains
TransatlanticEast PacificWest Pacific
North Indian
South Indian South WestPacific
South EastPacific
15
URI’s MPIPOM‐TC West Pacific Domain: Ocean Response to Supertyphoon Bolaven with
NCODA initialization: 2012082300‐2800
23/00ZSST
28/00ZSST
23/00Z77.5-m T
28/00Z77.5-m T
16
URI’s MPIPOM‐TC West Pacific Domain: Ocean Response to Supertyphoon Bolaven with
GDEMv3 + GFS SST initialization: 2012082300‐2800
23/00ZSST
28/00ZSST
23/00Z77.5-m T
28/00Z77.5-m T
17
GFDL/MPIPOM‐TC retrospective forecast of Supertyphoon Haiyan, initialized 2013110500 (NCODA)
18
Hurricane Model
Ocean Model
Air‐Sea Interface
Wind speed (Ua)Temperature (Ta)Humidity (qa)
Momentum flux (τ)Temperature fluxShortwave radiation
aaDa UUC saaHH TTUCQ saaEP
VE qqUC
CLQ
Momentum flux (τ)Sensible heat flux (QH)Latent heat flux (QE)
SST (Ts)
GFDL/N and HWRF Ocean Coupling
19
Uc Uλ SST
τdiff, τCor, α, γ, La
λ
Uref Rib, Uc
τair – atm. stress FQ – latent heat flux FS – sensible heat flux FR – radia on heat flux θ – poten al temperature Rref– atm. spec. humidity Rib – bulk Richardson num. Uref – atm. wind τc – stress on currents SST – sea surf. temperature
Ocean flu
x
e s
Atmospheric flu
x
e s
Wave flu
x
e s
Ψ
τc
τdiff τCor,τair
La λ FR FQ FS
Rref, Uref, θref
γ, α Uc, SST
τair, FQ, FS
Atmospheric Model
Ocean Model
Wave Model Uλ
U10N
Uref,FR
ASIM
Rib, τair FS, FQ
Coupler
Uc – surface current Uλ – current at depth λ – mean wavelength γ –wind/stress misalignment α – Charnock coefficient τCor‐ Coriolis‐Stokes stress τdif – wave mom. budget Ψ – wave spectrum La – Langmuir number U10N – neutral 10m wind
University of Rhode Island
Atmosphere‐Wave‐Ocean Coupled Framework for GFDL/N and HWRF models
20
GFDL/(WWIII)/POM‐TC 72‐h Forecast of Hurricane Irene (2011082612)
RHG
35-m Drag Coefficient
No wave couplingFully coupled
21
GFDL/(WWIII)/POM‐TC 72‐h Forecast of Hurricane Irene (2011082612)
RHG
35-m Wind (m/s)
No wave couplingFully coupled
22
GFDL/(WWIII)/POM‐TC 72‐h Forecast of Hurricane Irene (2011082612)
DCCM RHG
SST (oC)
No wave couplingFully coupled Fully coupled – No wave
SST (oC) ΔSST (oC)
23
Major 2014 upgrade of the GFDL/N and HWRF ocean model components: Replacing POM-TC with MPIPOM-TC.
Alternative MPIPOM-TC ocean initialization options are being evaluated and tested for future implementation.
A fully coupled atmosphere-wave-ocean framework has been developed for the GFDL/N and HWRF models to explicitly resolve effects of wind-wave-current interactions.
The next step is to investigate the influence of explicit wave coupling on hurricane prediction. Near real-time test runs are planned in 2014.
Summary