Urban PBL Seminar Series Part 2 of 2: Numerical simulation of urban climate, weather, and air quality

1
Urban PBL Seminar SeriesPart 2 of 2 Numerical
simulation of urban climate, weather, and air
quality
Prof. Bob Bornstein Dept. of Meteorology San Jose
State University San Jose, CA USA pblmodel_at_hotmail
.com presented at Ben Gurion University 18 May
2005
2
OVERVIEW

• Formulation
• Basic equations
• Parameterizations
• Applications
• Urban climate
• Severe Wx
• Air quality
• ER
• Conclusions

3
Starting Points

• Newtons a F/m (all are vector eqs)
• For atm ?V/?t -adv F/m g FL
• In rotating (x, y, z) Cartesian coordinates
• ?V/?t -adv F/m g FL Co Ce
• Where Co has 4 non-zero components
• Assume
• g g Ce
• FL is ignored
• Tangent plane coordinates (for now)?
• Earth curvature ignored
• Result ?V/?t -adv F/m g Co

4
Reynolds (R.) Averaged Equations

• If mesoscale-gap (diurnal-scale eddies) exists
  (?) in energy-spectrum b/t
• Large eddies (synoptic scale waves)
• Small eddies (turbulence)
• Then can R.-decompose all variables
• A A A (instantaneous mean turbulent)
• A is freq written ()
• () is average over a ?t ?Vol
• Thus R.-average each eq (mV, heat,) ?
• ?V/?t -adv F/m g Co VV all ()
• where vv is effect of turbulence on mean flow
  (a frictional drag) from adv term

5

• Resulting R.-averaged Navier-Stokes Eqs
• 1) Horizontal Momentum
•  
• 2) Vertical Momentum

6
Eqs. (cont.)

• 3) Temperature (error DT/Dt? ?T/?t)
• 4) Pressure
• But model uses -transformed Eqs., where
  z-coor-dinate horiz-derivatives defined,
  respectively, by

7
Schematic representation of s-coordinates in
MM5
8
Map Projection Required on Spherical Earth

• Drop tangent-plane assumption
• Isometric vs Conformal (MM5s choice)
• Map scale factor defined
• Polar stereographic vs Mercator Cylindrical vs
  Lambert Conical, with
• The constants and make the projection
  true at
• so that

9
Map Projection (cont.)where map scale factor m
and constant K are given by
10

• Transformed-projected eqs.
• Horizontal momentum

11
Transformed-projected eqs. (cont.)

• 4) Temperature
• where,

12
MM5 PBL TKE Eqs.

• f(z)-formulations (e.g., MRF, Blackadar)
• Assumed z-profiles depend on
• local z/L fails in stable unstable conditions
• SBL z/L uses an assumed profile
• Dont allow for advective effects
• 3-D level 1.5-closure (e.g., G-S, B-T)
• prognostic TKE-eq. diagnostic l-eq.
• l-eq. is arbitrary wrong at PBL top

13

• Freedman (Stanford, NCEP, SJSU) has
• 3-D level-2 closure
• prognostic TKE- e-eqs. ?
• correct l (z) K (z) profiles to PBL top
• TKE- e- Ks have correct mag relative to Km

14
PROG APPROACH FOR LENGTH SCALE Freedman
Jacobson (2002 2003, BLM) Freedman at SJSU
2 prog Eqs. TKE DISSIPATION RATE e
__
__

• Where l ceE3/2/e
• Values of se sE are reversed in Mellor
  Yamada ?
• reversed in all atm models ?
• K TKE in upper PBL were wrong!

15
CALIBRATION TO NEUTRAL ABL l vs. z
Lines various values of ? ce2se/sE

• x COLEMAN (99) DNS
• New (R-panel) best-fit
• ? 1.3 (dashed line),
• w/ better results (l?) in
• upper PBL
• Standard approach
• (left panel) best fit
• with ? 2.5, w/ poor
• results in upper PBL

new
old
16
Same, but for K (z)
x COLEMAN (99) DNS New (R-panel) best-fit
? 1.3 (dashed line), w/ better results in
lower PBL K? aloft Standard approach (left
panel) best fit with ? 2.5, w/ poor results
in lower PBL
new
old
17
Internal BC at SfcUrban Parameterizations for

• Surface energy- and moisture- balance upsets
• Building radiation multiple reflections
• Building heat-storage release
• Anthropogenic heat- moisture-fluxes
• Porous-flow effects
• 3-d building data
• zo and do distributions
• Urban canyon profiles ff, dd, T, q, TKE
• Urban effects in all PBL prognostic eqs.

18
Sfc Energy Models SiB has new vegy-paramters
19
Within Gayno-Seaman PBL/TKE scheme
From EPA uMM5 Mason Martilli (by Dupont)
20
_________
______
3 new terms in each prog equation
? Advanced urbanization scheme from Masson (2000)
21
New GIS/RS inputs for uMM5 as f (x, y, z)

• land use (38 categories)
• roughness elements
• anthropogenic heat as f (t)
• vegetation and building heights
• paved-surface fractions
• drag-force coefficients for buildings
  vegetation
• building height-to-width, wall-plan,
  impervious-
• area ratios
• building frontal, building plan, and rooftop
  area-
• densities
• wall and roof e, c?, a, etc.
• vegetation canopies, root zones, stomatal
  resistances

22
Urbanized meso-met model results
urban effect
_________________
23
Martilli/EPFL results
Urbanization ? day nite on same line ?
stability effects not important
Non-urban
urban
24
uMM5 performance by CPU
? With 1 CPU MM5 is 10x faster than uMM5
With 96 CPU MM5 is still gaining, but MM5 has
ceased to gain at 48 CPU then it starts to
loose
? With 96 CPU MM5 is only 3x faster than uMM5
(lt 12 CPU not shown)
25
Performance by physics
sound waves PBL schemes take most CPU in both
urban/PBL scheme in uMM5 takes almost 50 of all
time
26
Wmax VS. NO. OF CPU DIFFERENCES AT 16 17 HR
COULD BE DUE TO CHANGES IN INTEGRATION TIME-STEP
27
KEY IDEA IDEAL MESO-MET ATM-MODEL CAPTURES ALL
BC FORCINGS IN CORRECT ORDER

• O3 EPISODES OCCUR ON A GIVEN DAY
• NOT B/C TOPO, EMISSIONS, OR SFC MESO-FORCING
  (EXCEPT FOR FOG) CHANGES
• BUT DUE TO CHANGES IN UPPER-LEVEL SYNOPTIC WX
  PATTERNS, WHICH
• COME FROM AN EXTERNAL MODEL WHICH
• ALTER MESO SFC-FORCINGS (i.e., TOPO, LAND/SEA,
  URBAN) VIA MESO-TEMP AND THUS WIND
• MUST THUS EVALUATE ABOVE FACTORS
• UPPER LEVEL SYN WX Patterns pressure then
  winds
• TOPOGRAPHY (via grid spacing) channeling of wind
• MESO SFC temperature then winds

28
NYC (a) z0 vs UHI effect in z-section (upper L)
(b) Sea breeze obs (upper R), (b) no
barrier (lower L), with barrier (lower R)
29
ATLANTA UHI-INITIATED DAYTIME STORM (0BS) T V
(UPPER L) CONV (UPPER R) PRECIP (LOWER L)
CLOUDS (LOWER R)
30
MM5 simulation of previous storm UHI (upper L),
CONV (UPPER R) w (lower L), PRECIP (LOWER R)
31
SFBA Inter-basin O3-Transport

• Small changes in meso Ls and/or Hs
• (not seen on NWS maps) ?
• open or closed inter-basin transport paths
• Downwind basins O3 problem thus cannot be solved
  by their own emission reduc-tions (which costs
  lots of )

32
NYC inter-state O3
OBJECTIVE WINDS DO NOT SHOW SEA BREEZE FLOW, AS
MODEL DOES
BOTH WIND FIELDS PRODUCE CORRECT O3 MAX VALUE,
BUT ONLY MODEL-WINDS CORRECTLY GET (R-FIG) ITS
LOCA- TION RIGHT
33
Simulations of ozone over Israel, West Bank, and
JordanE. Weinroth, M. Luria, A. Ben-Nun, C.
Emery, J. Kaplan, M. Pelegand Y. MahrerSeagram
Center for Soil and Water Sciences Faculty of
AgricultureThe Hebrew University Rehovot 76100
Israelweinroth_at_agri.huji.ac.ilS. Kasakseh,
Applied Research Institute JerusalemBethlehem,
West BankJ. Safi, Environmental Protection
Research InstituteGaza City, Gaza R.
Bornstein, Dept. of Meteorology, San Jose State
University, San Jose, CA, USAAtmospheric
Sciences and Air Quality Conference2729 April
2005, San Francisco, California
34

• Specific project objectives
• Install environmental monitoring sites
• Prepare environmental databases
• Prepare regional climatology
• Conduct field campaigns during periods conducive
  to poor regional AQ
• (5) Apply RAMS MM5 to CAMx to increase
  understanding of current future air quality
  problems

35
Emission Inventory 1997- 8

• 15 Large Stationary (point) sources (58 fuel
  consumption)
• 400 Medium Stationary (point) sources (7)
• Small Stationary (area) sources (12)
• Solvents (area) sources
• Biogenic Stationary (area) sources (isoprene and
  monoterpene)
• Mobile (area) sources, both ground based and
  aerial (22)

36
(No Transcript)
37
(No Transcript)
38
(No Transcript)
39
Weather Conditions

• Pre-episode
• weak Persian trough ?
• slow V low mixing depth ? high NOx ?
• O3 titration (costal)
• Episode
• trough strengthens?
• increased surface HPG ? augmented westerly sea
  breeze front ?
• high inland O3 concentration

Med Sea
40
MM5
1.8.97 700
41
RAMS/CAMx Results (left) vs Airborne
Measurements (right)for Mid-East USAID project
showsmax impact of Tel Aviv sources is in Jordan
Flight Path
Jerusalem
1.8.97 1500
42
RAMS/Camx O3 Results vs Measurements
43

1. All emission sources
2. All Industry sources
3. Main (large) Industry sources
4. Medium and small (low) industry
5. Without Industry only Vehicles, Solvents
   Vegetation
6. Vehicles only
7. Without vehicles All Industry, Solvents
   Vegetation
8. Without emissions (initial and boundary
   conditions)

8 Emission Input Scenarios
All emission sources
All industry sources
Without industry sources
1.8.97 1500
1.8.97 1500
1.8.97 1500
1.8.97 1500
44
Ozone Peaks for all Scenarios 1 Aug 97
Source O3 Peak (ppb) Comparison to All Sources Peak in (discounting initial 45 ppb)
All sources 116 100
Without emissions 56 15
Industry low 58 18
All industry 98 75
Industry large 97 73
Without Industry 82 51
Without vehicles 103 81
Vehicles 80 49
45
SJV episode Fresno eddy moved N H moves inland
(both better defined than in D-1) flow around
eddy blocks SFBA flow to SAC, but forces it S
into SJV
L
H
46
Sfc obs at 0700 PST 31 July (LIV episode
morning) Note confluent flow into LIV
47
Episode afternoon (1400 PDT) W flow thru GGG
strong con into E-Liv
48
LIV-episode late-afternoon (1800 PDT) flow to
SAC from SFBA blocked
49
SAC-episode late-afternoon (1800 PDT) flow to
SAC from SFBA not blocked
50
LA Basin O3-Episode Example(Boucouvula et al.,
2003)

• Episode was due to synoptic change
• onshore movement of 700 mb coastal H ?
• Reduced Marine BL depth
• Subsidence warming ?
• strengthened subsidence inversion layer
• Upper-level easterly flow
• Easterly-flow at inland surface-sites
• Sea-breeze surface convergence zone
• Max surface ozone (180 ppb) at inland sites in
  afternoon on 5 Aug within convergence zone

51
SCOS Temp (full domain)
RUN 1
RUN 5
03-Aug-1997
04-Aug-1997
05-Aug-1997
06-Aug-1997
52
uMM5 for Houston O3 SIP

• 106 CPU cluster
• GIS/RS gridded urban sfc parameters
• uMM5 reforestation ?
• reduced daytime max UHI ?
• CMAQ O3 model uMM5 output ?
• reduced emissions photolysis rates ?
• lower O3 ? emission-reduction credits ?
• big savings of

53
From S. Stetson Houston zo data
54
Coastal Cold-Core L on episode day at 3 PM for
Domains 1-3
L
55
Domain 3 (12 km) 4 PM cold-core L (from
SST-eddy??)
L
56
(No Transcript)
57
Domain 4 (3 PM) Note cold-core L off of Houston
on O3 day (25th)
? Episode day
L
L
58
Urbanized Domain 5 near-sfc 3 PM winds on 4
successive days

• Episode
• day

59
Domain 5 end of daytime UHI (8 PM 21 Aug)

• Upper L MM5
• Upper R uMM5
• Lower L uMM5-MM5
• uMM5? 1.5 K warmer (stronger UHI)
• Blob is LU/LC error

60
Base-case (current) vegetation cover (urban min)
Modeled increases in vegetation cover (urban
max) values are 0.1 of those above
61
Run 12 (urban-max reforestation) minus Run 10
(base case) near-sfc ?T at 4 PM shows
thatreforested central urban-area cools
surrounding deforested rural-areas warm
62
CMAQ ozone modeling for Houston SIP 6
tree-planting scenarios ? reduced UHIs (right)
in urban-box 1 (left) for run 17? lower
max-ozone ? emission-reduction credits from EPA
Max impact
63
ER applications
64
EMERGENCY RESPONSEINFO REQUIRED (IN 20-30 MIN)

• WHAT IT IS
• WHAT IS ITS CONCENTRATION
• WHERE IS ITS SOURCE
• WHERE WILL IT GO
• WHAT ARE ITS FUTURE
• CONCENTRATION PATTERNS
• INFO TO 1ST RESPONDERS USEFUL TO THEM
• HOLD IN PLACE VS EVACUATE
• STAY INDOORS THEN EXIT

65
REAL-TIME FORECASTS REQUIRE

• GOOD LARGE-SCALE WX-MODEL RESULTS
• GOOD URBAN-CANYON MODULES IN MESO-MODELS
• GOOD MESOSCALE (SFC UPPER AIR) OBS NETWORKS
• REAL-TIME COMMUNICATIONS
• GOOD SIMPLE MODELS FOR EMERGENCY RESPONSE

66
(No Transcript)
67
(No Transcript)
68
(No Transcript)
69
(No Transcript)
70
QUIC Simulation dd 215 deg
MSG
wind vectors at 5 m height
71
LBNL Group
72
LINKING MESO-MET ER MODELS CHALLENGES

• DIFFERENCES IN SCALE
• LOWER-LIMIT HORIZ GRID-RESOLUTION OF MESO-MET
  MODEL B/F SPECTRAL-GAP VIOLATED?
• HOW LARGE AREA CAN CFD CODES COVER?
• SCIENCE QUESTIONS
• DOUBLE COUNTING TURBULENCE?
• MISMATCH B/T TKE SCHEMES?
• MESO-SCALE WIND VARIATIONS IN TIME SPACE?
• ONE- VS TWO-WAY LINKAGES?

73
The EndAny questions??