How well can we model air pollution meteorology in the Houston area? Wayne Angevine CIRES / NOAA ESRL Mark Zagar Met. Office of Slovenia Jerome Brioude, Robert Banta, Christoph Senff, HyunCheol Kim, Daewon Byun

Orientation  Surface sites to be used for temperature and wind comparisons  LaPorte wind profiler in green Galveston Bay Gulf of Mexico 55km 50km

Orientation  Satellite image on 1 September LST  Coasts low and sandy, little elevation change or terrain

Measurements and simulations  Texas Air Quality Study II (August- October 2006)  Surface meteorological and pollution monitoring sites  Mixing heights and winds from a radar wind profiler at LaPorte (on land)  WRF simulations

How can we tell if one model run is better than another?  Need metrics that clearly show improved performance  Several approaches: –Traditional bulk statistics –Case studies –Sea breeze and stagnation frequency –Plume locations

WRF simulations  75 days, 1 August – 14 October 2006  5 km inner grid spacing  Three styles: –FDDA of 3 wind profilers, reduced soil moisture, and hourly SST –FDDA of 3 wind profilers and reduced soil moisture –Reduced soil moisture only  All with ECMWF initialization every 24 hours (at 0000 UTC)  Retrospective runs, not forecasts

Impact of FDDA on wind profile  Full run, all hours  FDDA reduces random error in direction  Note this is not an independent comparison (this data was assimilated)  Red is FDDA run  Blue has FDDA, 1-h SST, and reduced soil moisture  Green has reduced soil moisture only

Impact of FDDA on surface winds  Full run, all hours  FDDA reduces random error in direction (more clearly seen if only daytime hours are considered)  ECMWF has less speed bias at C35 and C45 and less random error in speed at all sites  ECMWF has similar direction bias and random error to WRF runs over all hours, but WRF w/FDDA is better in daytime  Red is FDDA run  Blue has FDDA, 1-h SST, and reduced soil moisture  Green has reduced soil moisture only  Black is ECMWF

Impact of FDDA and soil moisture on surface winds  Episode days (17) only  Site C45, southeast of Houston very near Galveston Bay  FDDA improves random error in both speed and direction  1-h SST improves random error in the afternoon, but makes it worse at night  ECMWF has different but comparable errors, but WRF w/FDDA is better at hours 18 and 21 (and worse at hour 3)  Red is FDDA run  Blue has FDDA, 1-h SST, and reduced soil moisture  Green has reduced soil moisture only  Black is ECMWF

Impact of FDDA and soil moisture on surface temperatures  When are the errors worst?  10 days have at least one hour with temperature difference > 5K at site C35 (28 hours total) in FDDA run  All differences > 5K have model > measurement (model too warm)  All 10 days have convection or a cold front in reality  Model also has clouds and fronts but different amount, timing, or location

New metrics: Sea breeze frequency  How often does a sea breeze occur in the simulation AND measurement?  Definition: Northerly component >1 m/s between 0600 and 1200 UTC and southerly >1 m/s after 1200 UTC  FDDA or FDDA+1hSST run closer to measurement at all 7 sites (at least a little)  Results not sensitive to threshold Red is FDDA run Blue has FDDA, 1-h SST, and reduced soil moisture Green has reduced soil moisture only

New metrics: Net trajectory distance  Trajectories starting midway along the Ship Channel at 1400 UTC each day, extending for 10 hours at 190 m AGL  WRF run w/FDDA  Comparing total distance to net distance  A rough measure of recirculation  The lower left portion of the diagram is of most interest

New metrics: Net trajectory distance  Net distance was found by Banta et al. to correlate well with maximum ozone  Also holds for trajectories from WRF simulated winds, shown here  r = -0.85, r 2 = 0.72  Run with FDDA  Run with 1-h SST about the same  Total distance correlation much worse (r = -0.57)

New metrics: Vector average wind  Averaging u and v vs. averaging speed  Over 10 hours UTC  Interesting points are those below the 1:1 line since they have significant curvature  Run with FDDA and 1-h SST  Correlates well with measured wind (r > 0.9) in either run with FDDA  Non-FDDA run not as good (r < 0.85)

New metrics: Vector average wind  Good correlation with max ozone from airborne measurements  r = -0.91, r 2 = 0.83  Run with FDDA and 1-h SST  Runs without 1-h SST about the same  Without FDDA results are much worse  Scalar speed correlation slightly worse(?) (r = -0.88) but still better than net trajectory distance

Lagrangian plume comparisons  FLEXPART dispersion model with real emissions  Met fields from WRF (red) and ECMWF (blue)  SO2 measurements from NOAA aircraft (black)  WRF result has much better resolution and plume locations, even if averaged to same grid

Conclusions  ECMWF model used for initialization is already quite good, making it difficult to demonstrate improvement with high-resolution simulations  Traditional statistics (bias and std. dev.) don’t crisply display differences between runs, although they generally indicate improvement with FDDA –Different sites show different results  Looking at distribution of errors is useful –Large errors in temperature (>5K) occur when moist convection is present  New metric of sea breeze correspondence shows improvement at all 7 surface sites with FDDA  Net trajectory distance correlates better with ozone than total distance  Vector average wind correlates still better with ozone, scalar average wind speed almost as good  Average wind (vector or scalar) shows clearly that FDDA makes an important improvement under high-ozone conditions  Improvement above the surface is easy to demonstrate (eg. by comparison with wind profiler data)  Lagrangian plume model provides clear information about directly relevant performance of the model, but how to encapsulate?  Uncertainty analysis is needed  How good is good enough?  What if we know we have improved the model, but can’t show that we have improved the results?

Thanks to:  Bryan Lambeth, Texas Commission on Environmental Quality  NOAA P3 scientists  Richard Pyle and Vaisala, Inc. for funding  and many others

New metrics: Sea breeze frequency  How often does a sea breeze occur in the simulation or measurement?  Definition: Northerly component >1 m/s between 0600 and 1200 UTC and southerly >1 m/s after 1200 UTC  FDDA or FDDA+1hSST run closer to measurement at 4 of 7 sites Red is FDDA run Blue has FDDA, 1-h SST, and reduced soil moisture Green has reduced soil moisture only Black is surface site measurement

New metrics: Stagnation frequency  How often does stagnation occur in the simulation or measurement?  Definition: Wind speed < 1 m/s at any hour between 1500 and 2300 UTC  FDDA or FDDA+1hSST run closer to measurement at 3 of 7 sites Red is FDDA run Blue has FDDA, 1-h SST, and reduced soil moisture Green has reduced soil moisture only Black is surface site measurement

New metrics: Stagnation frequency  How often does stagnation occur in the simulation AND measurement?  Definition: Wind speed < 1 m/s at any hour between 1500 and 2300 UTC  No clear improvement with FDDA or FDDA+1hSST  Results not sensitive to threshold Red is FDDA run Blue has FDDA, 1-h SST, and reduced soil moisture Green has reduced soil moisture only

New metrics: Sea breeze and stagnation  Other things we can learn from these metrics: –Sea breeze correspondence is good at C45, closest to Bay and Gulf, with high frequency –Even better sea breeze correspondence at C81 with lowest frequency –C45 has the lowest stagnation frequency Red is FDDA run Blue has FDDA, 1-h SST, and reduced soil moisture Green has reduced soil moisture only