-
Thu 1/28/2016• WRF tendency terms
• WRF LSM options
• Water-surface parameterization
• Representation of turbulence
Reminders/announcements:- SCM2 assignment due Tuesday- Upcoming:
WRF real-data case assignment (will post,
officially assign Tuesday)
-
cd to WRFV3/dyn_em directory. Examine solve_em.F (4558
lines):
WRF Solver
-
cd to WRFV3/dyn_em directory. Examine solve_em.F:
Solve_em.F:
Large outer Runge-Kutta loop:non-timesplit physics (for state
variables)acoustic time step (many per larger outer loop)advance
variables (moist, chemistry scalars)time-split physics (cloud
microphysics)
End large outer Runge-Kutta loop
Time filtering
Lots of MPI issues along the way. 4558 lines of code, very well
commented!
WRF Solver
-
Solve_em.F:
Large outer Runge-Kutta loop:non-timesplit physics
CALL phy_prep ( config_flags, mut, u_2, v_2, p, pb, alt, ph_2,
phb, t_2, tsk, moist, num_3d_m, mu_3d, rho, th_phy, p_phy, pi_phy,
u_phy, v_phy, p8w, t_phy, t8w, z, z_at_w, dz8w, fnm, fnp, RTHRATEN,
RTHBLTEN, RUBLTEN, RVBLTEN,
RQVBLTEN, RQCBLTEN, RQIBLTEN, RTHCUTEN, RQVCUTEN, RQCCUTEN,
RQRCUTEN, RQICUTEN, RQSCUTEN, RTHFTEN, RQVFTEN, ids, ide, jds, jde,
kds, kde, ims, ime, jms, jme, kms, kme,
grid%i_start(ij), grid%i_end(ij), grid%j_start(ij),
grid%j_end(ij), k_start, k_end)
e.g., TH = theta (potential temperature), RA = radiation,
etc.
WRF Solver
-
Physics Tendencies
In namelist.input &time_control:iofields_filename =
"myoutfields.txt"
Added variables in myoutfields.txt:
+:h:0:H_DIABATIC,RTHBLTEN,RTHRATEN
“+:h” means add these variables to the “history” (output)
file
Otherwise to add variables, would modify Registry,
re-compile
-
WRF LSM options
Sources include Jimy Dudhia WRF presentation
-
WRF LSM namelist options (V3.5.1)
4 choices
-
WRF LSM namelist options (V3.7.1)
7 choices
-
sf_surface_physics=1
5-layer thermal diffusion model from MM5
• Predict ground temp and soil temps• Thermal properties depend
on land use• No effect of water• Provides heat and moisture fluxes
for PBL• Very simple treatment, not recommended for real-
data cases
• Set num_soil_layers = 5, re-run ideal.exe or real.exe
-
sf_surface_physics=2
Noah Land Surface Model – formerly OSU LSM(Chen and Dudhia 2001,
Ek et al. 2003)
• Vegetation effects included• Predicts soil temperature and
soil moisture in 4 layers• Predicts snow cover and canopy moisture•
Handles fractional snow cover and frozen soil• Diagnoses skin temp
and uses emissivity• Provides heat and moisture fluxes for PBL
• Set num_soil_layers = 4, re-run ideal.exe or real.exe
-
Noah Land-Surface Model (LSM)
-
sf_surface_physics=3
RUC Land Surface Model (Smirnova et al. 2000)
• Vegetation effects included• Predicts soil temperature and
moisture in 6 layers• Multi-layer snow model• Provides heat and
moisture fluxes for PBL• Can also now run with 9 layers
• Set num_soil_layers = 6, re-run ideal.exe or real.exe
-
sf_surface_physics=4
Noah “MP” Land Surface Model (Niu et al. 2011)
• New “multiparameterization” options: Improved– Vegetation
canopy energy balance– Snowpack and frozen soil representation
(snow skin T)– Groundwater interactions, runoff– Evidently, dynamic
vegetation option (!)
• Requires additional set of 12 namelist settings
-
sf_surface_physics=4
-
sf_surface_physics=5
CLM4 [from NCAR Community Atmosphere Model (CAM) climate model]
LSM (Lawrence et al. 2011)
• Many additional processes included, biogeochemical model
(Carbon-Nitrogen)
• 10-layers• Urban effects• Dynamic – land-use change
capability• Aerosol deposition for snow melt, snow aging, etc.
-
sf_surface_physics=5
Lawrence et al. (2011);
Fig. 1
-
sf_surface_physics=7Pleim-Xiu LSM (Pleim and Xiu 1995)
• 2-layer model, aim towards air-quality applications
• “Force-restore”, T forcing is surface energy budget, including
radiation, SH flux, LH flux, and ground flux
• Moisture forcing includes surface precipitation, dew,
evaporation, and soil moisture flux
• Accompanies P-X PBL package (asymmetrical convective model,
ACM2); probably best to run together
-
sf_surface_physics=8SSiB (Xue et al. 1991, Sun and Xue 2001)
• 3-layer model, aim towards climate applications
• Simple Biosphere Model (SiB) was predecessor
• Now, Simplified Simple Biosphere model (SSiB)
-
WRF SCM• Greensboro, NC from 1 / 18 / 2016, 12 UTC (default)
• Changed only LSM choice
• For some, had to link additional files from ../../run dir
• Compared variation of PBLH (all with YSU PBL scheme)
-
A few SCM tests… same PBL & surface layer
Thermal Diffusion (1)
Noah (2)
RUC (3)
CLM4 (5)
PBLH max ~ 1790 m
PBLH max ~ 2100 m
PBLH max ~ 2260 m
PBHL max ~ 2050 m
-
Water Surface RepresentationGiven that 70% of Earth’s surface is
covered by water, very important specification in model
Strong influence on cyclones, climate, coastal processes,
boundary currents, planetary energy budget, etc.
Latest generation of models are coupled (AOGCMs, CFS, WRF,
MPAS)
Full coupling includes wave model (Drs. Xie, He in MEAS)
-
Water Surface RepresentationWRF features several options for
dealing with sea surface:
- Static (hold initial values constant)
- SST update (use analyses to supply evolving SST analyses)
- Ocean Mixed-Layer Model (OML): Only mixing, no currents or
upwelling
- 3-D Ocean Model (new): Based on Price (1981)
-
Water Surface RepresentationWhere do SST initial conditions come
from?
- Satellite observations crucial for SST analysis (Reynolds and
Smith 1994) – Optimum Interpolation technique to blend available
observations
- Higher resolution now available, 14-km grid or better
- NOAA AVHRR (advanced Very-High Resolution Radiometer) sensor
is used, 8 km grid, 2 IR channels. Must be cloud-free
- Satellite observations combined with in-situ ship, buoy
obs
- Supplemental data needed near ice edge (SSMI)
-
Analysis and Prediction of High-Impact Weather Events
WRF Simulations (Megan Mallard)• Advanced Research WRF,
version 3.0.1.1• 4-member ensemble with
varying parameterization schemes
– Microphysics: Morrison & WSM6– PBL: YSU (TC fluxes on)
& MYJ
• Kain-Fritsch convective scheme (on 54- & 18-km grids)
• Initial & boundary conditions:– 1 GFS - FNL & 0.5 RTG
SST
• Results shown are from 18- & 6-km nests
Δx=54 km Δx=18 km
Δx=6 km
-
Analysis and Prediction of High-Impact Weather Events
OML – default settingsNo OML
Ocean Mixed Layer Model
72-hour Lead Time
939 mb 947 mb
-
Analysis and Prediction of High-Impact Weather Events
Sea Surface Temperatures• TCs interact with ocean
– Intensify over warm water– Leave cold water in their wake
• Ocean mixed layer model used to account for cold wakes– Cold
wakes persist
• Modified source code, SST updated with observed temperatures
daily with memory of cold wakes from previous day
SST at initialization2 months later…
Cold wake
-
Micrometeorology and Turbulence Parameterization
-
Outline1.) Review of turbulence and properties
- Characteristics, worksheet
- Definitions
- Regimes and sources
- Turbulence kinetic energy (TKE)
- Flux and flux divergence
2.) Closure problem- Bulk aerodynamic
- K-theory (mixing length)
- Higher-order closures
- WRF schemes, examples
-
Planetary Boundary Layer: Processes & Parameterization
In-class worksheet to familiarize with processes and set up
parameterization problem
First, a little background on turbulence
-
Turbulence
http://www.efluids.com/efluids/gallery/gallery_pages/iso_turbulence_page.htmhttp://arstechnica.com/science/news/2008/11/is-turbulence-a-chaotic-repeller-or-attracter.ars
Chaotic, isotropicEnergy cascade, dissipation
-
Turbulence…• Contains spatial scales not resolvable by
observations
• Characterized by downscale energy cascade; viscosity
transforms turbulent to thermal energy at smallest scales
• Thus, energy source is needed to maintain turbulence
• Generally acts to mix fluid properties
• Turbulent flux divergence or convergence is a first-order term
in governing equations, must be accounted for
-
Reynolds Averaging
s
t
Rule #1:
1
1 Nn
is s
N
0
1 Ts s dtT
' at any instants s s
0 by definitions
0 etcs s Any single-prime quantity = 0
-
What are turbulent fluxes of heat, moisture, & momentum?
Measuring Turbulence
Ground
Inversion
(z)
Turbulent air motions
w’(m/s) ’ (K)0.21 1.02
- 0.42 -0.350.03 -0.120.13 1.22
- 0.23 -1.450.08 0.971.02 0.81
Sensor, measuring w’, ’
-
Mathematical form of fluxes: Correlations (covariance) between
turbulent fluctuations is what matters
In general, do we expect terms like to be positive, negative, or
zero? Why? Under what conditions?
How about the momentum flux
How about moisture flux?
Turbulent Fluxes
θw
uw
qw
These are the vertical turbulent fluxes
w’(m/s) ’ (K)0.21 1.02
- 0.42 -0.350.03 -0.120.13 1.22
- 0.23 -1.450.08 0.971.02 0.81
-
Distribution of Atmospheric Turbulence
Turbulence – generation primarily from buoyancy, shear
“Free troposphere”- generally little turbulence, but near jets
some clear-air turbulence (CAT); also turbulence in/near
convection
Stratosphere- very stable, limited turbulence, good place for
commercial aircraft to fly!
~1 km
Inversion layer, entrainment zone
CATtropopause
-
SURFACE
PBL, mixed layer, Ekman layer, or “outer layer”
surface layer(constant flux layer)
Model-defined PBL top
Lowest Model level
z0
Free atmosphere
Viscous sub-layer, or molecular boundary layer
Model PBL Components
-
PBL Overview/Review• Structure of the PBL
– Molecular boundary layer: lowest few mm of atmosphere,
molecular diffusion dominates
– Surface layer: lowest 10-30 meters over which turbulent
momentum flux may be assumed constant, logarithmic wind profile
(neutral)
– Mixed layer: From top of surface layer to inversion base
marking lower boundary of inversion layer
– “Free” atmosphere: portion of troposphere not directly
affected by surface-based turbulent fluxes, frictional effects
• Properties of PBL:– Turbulent flow– Depth ranges from a few
meters at night with light winds to several
km over intensely heated, dry terrain; very strong diurnal
signal
-
Yamada and Mellor (1975)
Wangaraexperiment.
-
TurbulencePBL regimes:
– Unstable: Free convection, buoyant turbulence production–
Neutral: Forced convection, mechanical production– Stable: Forced
convection, mechanical production
How is turbulence generated?1.) Mechanical production – forced
convection, dynamic instability
Shear – frictional near surfaceWake turbulenceShear – clear air
in free atmosphere
2.) Buoyant – free convection, static instabilityPlumes, order
100 m in scale, may merge to form thermals, order 1 km in
scaleCharacterized by static instability
-
Turbulent Kinetic Energy (TKE)
mechanical thermalTKE Advect P P TurbTrans diss
t
Equation available for prediction of TKE per unit mass:
Buoyant term can be + or – depending on static stability
22221 wvu
mTKE
-
PBL Overview/ReviewRichardson number (Ri)
– Energy required to maintain/generate turbulence related to
Ri
– Richardson # is ration of buoyant to shear turbulence
production
– Two common forms: Gradient and Bulk Richardson number
22
22
vu
zgRi
zv
zu
zg
Ri
v
vb
v
vgr
-
Some fairly recent resultsQJ papers (2007, 2008) challenge old
ideas on Ri significance
-
Some fairly recent results
-
T
Td
PBL Review Worksheet4.) Consider the vertical profile of
temperature and dew point shown below. This
hypothetical profile could represent conditions over land on a
sunny spring afternoon in the mid-latitudes.
5.) Suppose that the profile shown above remains quasi-steady
until sunset. Sketch the expected evolution of this profile for a
clear, calm evening andovernight period using a series of dashed
lines to show the temperature profile evolution (using the same
diagram). What controls the slope of these lines?
-
T What physical processesare at work in bringing about these
temperature changes?
Which model physics parameterization packages are involved?
Radiation… Land surface model… PBL scheme… maybe CP
PBL Review Worksheet
6.) Consider the temperature profile below, which is observed at
sunrise for the same general situation as the profile on the
previous page. Sketch on this diagram how the profile would evolve
after the sun has risen, into the afternoon hours.
-
Turbulence parameterization in models is crucially important
PBL/turbulence parameterization involves interactions between
several other model physics schemes
Important to gain an intuitive feel for how turbulence might
affect a given simulation; check tendencies to confirm
In the coming weeks, will spend considerable time and effort to
familiarize ourselves with turbulence parameterization
Prepare to read, discuss PBL journal papers!
Summary