1 / 43

MM5 Meteorological Modeling for Houston-Galveston Area Air Quality Simulations

MM5 Meteorological Modeling for Houston-Galveston Area Air Quality Simulations. Daewon W. Byun Bonnie Cheng University of Houston Institute for Multidimensional Air Quality Studies (IMAQS).

duke
Download Presentation

MM5 Meteorological Modeling for Houston-Galveston Area Air Quality Simulations

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. MM5 Meteorological Modeling for Houston-Galveston Area Air Quality Simulations Daewon W. Byun Bonnie Cheng University of Houston Institute for Multidimensional Air Quality Studies (IMAQS)

  2. Objectives:Study the effects of land use and land cover modification on the urban heat island development and on the air quality in the Houston-Galveston metropolitan area. • Methods: • Improve meteorological simulations by applying better physics • (today’s presentation) • Incorporate most up-to-dated detailed land use and land cover data • (on-going)

  3. Land Surface Process • Land-sea breeze circulation is influenced by land surface processes • It controls the surface sensible/latent heat & momentum fluxes • They in turn influence PBL structure and convective processes • The land surface characteristics like topography, soil moisture and land use and land cover need to be described better.

  4. Previous Work • Meteorological simulations for the HGA provided by the Texas A&M (Dr. John Nielsen-Gammon) accurately predict maximum temperatures and reproduce the large-scale temperature patterns using slab land-surface model. • The model performance is good, but is less successful for the PBL heights in rural areas. • To better understand Houston high ozone problem, and to understand the impact of urban vegetation, we need to use LSM with better physics

  5. Force-restore 2-layer energy balance model with a constant temperature substrate (specified in INTERP the daily average surface temperature tuned to represent diurnal cycle best) (2) 5-Layer slab model Added multiple soil layers to represent heat conduction to the Force-restore model (3) Comprehensive land surface model with vegetation How to model the land-surface processes in Mesoscale models?

  6. Ground heat budget model with multiple layers (5) Improved version of a 2-layer force-restore surface energy balance model Couples surface momentum, heat, and moisture fluxes But, no treatment of evapotranspiration No moisture diffusion in the soil layer Current TAMU Base Simulation Uses Simple “5-layer” slab model Predict temperature at 5-soil layers. 1, 2, 4, 8, 16 cm thick

  7. Modeling Land Surface Process • Current LU/LC data used in MM5 simulation: USGS ~1-km resolution. • Higher resolution of LU/LC data now becomes available • Dominant LU/LC algorithm used in MM5 does not allow mosaic LU/LC land-surface processes (unless the code is modified) • The question is how to make best MM5 land-surface simulations if our LU/LC data is limited by the current data

  8. Use comprehensive land-surface model NOAH LSM (N:National Center for Environmental Prediction; O: Oregon State University; A: Air Force; H: Hydrological Research Lab.) (Ek et al., 2001). 4-layers (10, 30, 60, and 100 cm thick) Predicts soil temperature, soil water, canopy water, and snow/ice Central difference of the NOAH from slab model is stomatal resistance and transpiration formulation How to better model the land-surface processes?

  9. dominant landuse data used in MM5 Urban : 10 sites Rural : 18 sites

  10. Stephen Stetsen, GEM Pete Smith, Texas Forest Service

  11. Stephen Stetsen, GEM Pete Smith, Texas Forest Service

  12. Design of Meteorological Simulation: S1(SLAB) • First, (Base case):MM5 simulation coupled with the slab soil model (case S1). Soil moisture was increased in the urban area to make it wetter; decreased in the rural area make it drier. (Soil moisture value (%),Dr. Nielsen-Gammon, J. W., 2002) • S1 simulation is provided by TAMU.

  13. Configuration of MM5 simulations D4 D3 D2 • Analysis nudging for d1,d2,d3; observation nudging(wind vector)for d4 • d1, d2 2way nesting; d3,d4 continuous one-way nesting • MRF PBL Parameterization • Dudhia explicit moisture scheme • RRTM radiation scheme • Slab land-surface model (LSM) D1 D4

  14. Design of Meteorological Simulation: S2(NOAH) • Experiment #2: Use the recently developed NOAH Land Surface Model (NOAH LSM) (EK, 2001) with identical inputs and model configurations as in S1 case except using different land-surface parameterizations (S2).

  15. Scattered Diagram of 2-m Temp (a) (b) Scattered diagram of 2-m temperature with (a) S1; (b) S2 simulations.

  16. 10m WS,WD from metstat program (Emery, 2001) Averaged over the CAMS sites CST

  17. Summary of problems with the MM5/NOAH simulation with default parameters. • Simulated daytime temperature too high, and nighttime temperature too low at urban sites. • The urban area was treated as if totally covered with impervious surfaces. Therefore, we have large diurnal variations in temp and very low latent/sensible heat flux ratio. • At rural sites, we have no daytime temperature bias, but serious nighttime temperature bias. • Serious delays in the development of diurnal wind speed build up – related to #3? • Initially, we thought it were soil moisture problem, but …?

  18. Vegetation Parameters in LSM Vegetation Parameters USGS 27,1, 'ALBEDO Z0 SHDFAC NROOT RS RGL HS SNUP LAI MAXALB' 1, .15, 1.00, .10, 1, 200., 999., 999.0, 0.04, 4.0, 40., 'Urban' 2, .19, .07, .80, 3, 40., 100., 36.25, 0.04, 4.0, 64., Dryland Cropland 3, .15, .07, .80, 3, 40., 100., 36.25, 0.04, 4.0, 64., 'Irrigated Cropland 4, .17, .07, .80, 3, 40., 100., 36.25, 0.04, 4.0, 64., Dryland/Irrigated' 5, .19, .07, .80, 3, 40., 100., 36.25, 0.04, 4.0, 64., 'Cropland/Grassland 6, .19, .15, .80, 3, 70., 65., 44.14, 0.04, 4.0, 60., 'Cropland/Woodland' 7, .19, .08, .80, 3, 40., 100., 36.35, 0.04, 4.0, 64., 'Grassland' 8, .25, .03, .70, 3, 300., 100., 42.00, 0.03, 4.0, 69., 'Shrubland' 9, .23, .05, .70, 3, 170., 100., 39.18, 0.035, 4.0, 67., 'Shrubland/Grass' 10, .20, .86, .50, 3, 70., 65., 54.53, 0.04, 4.0, 45., 'Savanna' 11, .12, .80, .80, 4, 100., 30., 54.53, 0.08, 4.0, 58., 'Broadleaf Forest' 12, .11, .85, .70, 4, 150., 30., 47.35, 0.08, 4.0, 54.,'Needleleaf Forest' 13, .11, 2.65, .95, 4, 150., 30., 41.69, 0.08, 4.0, 32., 'Broadleaf Forest' 14, .10, 1.09, .70, 4, 125., 30., 47.35, 0.08, 4.0, 52.,'Needleleaf Forest' 15, .12, .80, .80, 4, 125., 30., 51.93, 0.08, 4.0, 53., 'Mixed Forest‘ : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : Parameters that determine evapotranspiration

  19. Question: How to treat the evapotranspiration from 20% tree/vegetation coverage when the current dominant category of USGS 25-LU/LC is used? • Experiment S3: We added new anthropogenic canopy water source in the urban area in the NOAH LSM to reflect Houston’s 20% urban vegetation and thus reduce the daytime temperature bias. Design of Meteorological Simulation

  20. Addition of Canopy Moisture is intercepted canopy water content is green vegetation fraction is input total precipitation canopy evaporation If exceeds S (maximum canopy capacity: 0.5 mm), the excess precipitation or drip, reaches the ground. is the anthropogenic contribution to the canopy water content. A reasonable value 3x10-6 (meter of available water per second) was picked for the simulation (* need to be justified).

  21. Surface (2-m) Temp Analysis • Now, temperature simulations at urban areas improved, not much change at rural sites. • Minimum temperature problem still exists both urban and rural sites Urban Rural CST

  22. Scattered Diagram of 2-m Temp (a) (b) Scattered diagram of 2-m temperature with (a) S2; (b) S3 simulations.

  23. Wind Profiler Sites (TexAQS-2000)

  24. PBL Height Analysis from S1, S2 and S3 simulations on Aug. 25 Profiler PBL ht. from NOAA/FSL CST

  25. PBL Height Analysis from S1, S2 and S3 simulations on Aug. 27 CST

  26. PBL Height Analysis from S1, S2 and S3 simulations on Aug. 30 CST

  27. PBL Height Analysis from S1, S2 and S3 simulations on Aug. 31 CST

  28. Name Category Emissivity Dryland 2  0.98 Cropland 5 0.95 Grassland 7  0.95 Evergreen forest 14  0.95 MM5 sensitivity test continued tocorrect min. temperature problem • Test emissivity value for different LU categories • (original emissivity in NOAH =1.0)

  29. To improve minimum temperature bias --- test Test case with modified emissivity: at a few sites we see imbalance between the energy budget and PBL growth, which results in abnormally high surface temperatures simulated

  30. MM5 sensitivity test continued tocorrect min. temperature problem Fei Chen modification for urban LU parameters • Change heat capacity from 3.0E06 to 2.0E06 (J/m^3/K) • Thermal conductivity equals 3.24 (W/m/K) • Change RSMIN from 200 to 300 (s/m) But, max. temperature problem for urban LU comes back at 4-km simulation even with the added canopy water. We consider changes in items “a” and “c” may cause this problem

  31. With modified land use parameters to reflect urban vegetation better ( in particular, heat capacity change is the key in fixing min. temperature) (12-km result, Fei-Chen urban mod)

  32. 2-mtemperature Fei Chen urban LU modification Lower bias problem was resolved in the urban area but maximum temp is still too high. Rural min. temp problem still existing

  33. 2-m temperature Fei Chen urban LU Mod. + Canopy moisture + Emissivity change Some improvement, but Still some problem.

  34. 2-m temperature result (R: rural areas; U: urban areas) More sensitivity test to fix high max. temperature bias in Urban area at 4-km resolution

  35. 2-m temperature Fei Chen urban LU mod (thermal conductivity only) + Canopy moisture + Emissivity change Improved over others Urban CAMS site average Rural CAMS site average

  36. 2-m temperature; 4-km domain Fei Chen urban LU Modification (heat capacity only), + Canopy water + rural Emissivity change

  37. Summary of MM5/NOAH simulation with added urban canopy moisture. • Original NOAH/USGS data for HGA: Urban area was treated as if it were totally covered with impervious surface. Therefore, we have large diurnal variations in temp and very low latent/sensible heat flux ratio in urban areas. • Bias in daytime temperature mostly fixed with added canopy water for the urban LU. • Min. temperature bias at urban sites was fixed with the heat capacity modification and by emissivity at rural sites

  38. Additional analysis & work needed • Need to look at development of daytime wind speed build up. • Moisture advection needed to be looked at. • Compare with other “sensitivity” studies • -- satellite surface temperature nudging (UAH & TAMU) • -- modified PBL algorithm, etc (ATMET) • -- Kiran’s GEM?

  39. Future MM5 work at IMAQS • Continue to test MM5/NOAH • need to continue to improve meteorological simulations especially for August 30 and 31 for wind development and moisture advection • Use improve land use and land cover data • - compare w/ TCEQ LU/LC used for biogenic emissions • - develop methods to incorporate fractional LU/LC effects • Future study will focus on the urban canopy parameterization (UCP) in meteorological modeling. By implementing the urban canopy parameterization into MM5, the meteorological simulation will be expected to have more accurate results on urban area simulation.

  40. 12-km CMAQ results

  41. dominant landuse data used in MM5 Urban : 10 sites Rural : 18 sites

  42. 12-km CMAQ results

More Related