Development of a Lightning NOx Algorithm for WRF-Chem

Development of a Lightning NOx

Algorithm for WRF-Chem

Amanda Hopkins HansenDepartment of Meteorology

Florida State [email protected]

Henry E. FuelbergFlorida State University

Kenneth E. PickeringNASA Goddard Space Flight Center

N2

O2

N2

N2

O2

O2

O2

N2N2

N2

N2

N2

N2

N2 + O2 NO + NOAir is ~78%N2 and ~21% O2

Ozone Formation in the Troposphere

N2 + O2 NO + NO

NO + O3 NO2 + O2

NO2 + hν NO + O

O + O2 + M O3 + M

h = 290nm < <410nm

•These reactions set up the photostationary state.•O3 depends on the rates of formation and photolysis of NO2

•In order for ozone to increase, NO2 formation must happen without destroying O3

Air is ~78%N2 and ~21% O2

Oxidation of methaneOxidation of carbon monoxide

Rationale - why is LNOx important1. Lightning is an important source of NOx in the

relatively clean upper troposphere

Zonal Mean Lightning NOx production (10-2kg/s)From the NASA GISS GCM

Pre

ssu

re (

hP

a)

June, July, and August December, January, and February

•is the third most important greenhouse gas

•impacts the Earth’s radiation budget and can cause changes in atmospheric circulation patterns.

•is toxic to humans, plants and animals.

TROPOSPHERIC OZONETROPOSPHERIC OZONE:

Photo courtesy of University of California at Berkley

NONOxx [NO+NO [NO+NO22]]:•is a primary pollutant found in photochemical smog•is a precursor for tropospheric ozone formation

Rationale - why is LNOx important

2. LNOx indirectly affects our local air quality and global climate

Has a strong influence on Ozone (O3) and hydroxyl radical (OH) concentration

Global NOx Budget

fossil fuels52%

lightning16%

soils15%

biomass burn15%

aircraft2%

Figures from Global Emissions Inventory Activity data setsNote: lightning figures are calculated by model internally

3. LNOx is difficult to realistically model, but is Important in the global budget of NOx

LNOx algorithm for WRF-Chem with Parameterized Convection

what are the necessary ingredients?

1)Flash rate parameterization

2) Average NO production per flash

3) Method of specifying the vertical distribution of LNOx emissions

Schemes often used for Flash Rate Production

1. Cloud Top Height [Price and Rind, 1992; Michalon et al., 1999]2. Convective Precipitation Rate [Meijer et al., 2001; Allen and Pickering, 2002]

Separate parameterizations are needed to simulate land and oceanic lightning

3. Upward Cloud Mass Flux/Updraft Velocity [Grewe et al., 2001; Allen andPickering, 2002]

This scheme cannot be developed based on observations. Model output must be used. Therefore, the relationship between these variables and flashrate will be model-dependent.

None of these schemes yield global flash distributions that compare wellwith satellite flash observations. What is needed is a moremicrophysically-based scheme. Possibilities include the schemes byDeierling et al. (2005) and Futyan and Del Genio (2007).

Deierling et al., 2005

Flux Hypothesis

They propose that lightning frequency, f, is proportional to the product of the precipitation rate, p, and the mass upflux of ice crystals, Fi .

The charging mechanism used in the this study involves rebounding collisions between heavy graupel pellets and the lighter ice crystals (Reynolds et al., 1957; Takahashi, 1978; Jayaratne et al., 1983; Baker et al, 1987). This process creates a vertical electric dipole by gravitational separation of oppositely charged graupel and ice crystals.

iJpFf

This scheme would work very well on the cloud-scaleWe need a Regional scale relationship!!!

Convective Radar Storm HeightFutyan and Del Genio, 2007

fifth order power law for radar top height (above surface)

Cloud top height may be several kilometers higher than the height of significant radar signal

second order for radar top height above 0 degree isotherm.

WSR-88 Doppler Radar and the Lightning Detection and Ranging (LDAR) data

3-D Lightning Mapping locations to be incorporated

Kennedy Space Center, Florida** - LDAR/VaisalaHuntsville, Alabama – NM TechDallas, Texas – LDAR/VaisalaWashington D.C. – NM Tech

**We will start with KSC

Futyan and Del Genio (2007) used TRMM data for their research:

we will use Doppler radar and LDAR data to form a relationship between flash rate and radar storm height.

WSR-88 Doppler Radar and the Lightning Detection and Ranging (LDAR ) data from Kennedy Space Center (KSC)

KSC LDAR. Sensor 0 is the central LDAR receiver. (After Poehler and Lennon 1979 and Vollmer 2002)

Doppler RadarMelbourne, Florida

The location of the Melbourne, Florida National Weather Service Office WSR-88D radar in relation to the Kennedy Space Center. Range rings are provided at 10 km intervals. (After Nelson 2002).

WSR-88 Doppler Radar and the Lightning Detection and Ranging (LDAR ) data from Kennedy Space Center (KSC)

WDSS: Warning Decision Support System

This will provide a way to visualize lightning flashes with radar reflectivity

The image depicts individual Lightning Detection and Ranging (LDAR) sparks as measured by the Kennedy Space Center LDAR network, the cloud-to groundstrikes measured by the National Lightning Detection Network (NLDN), and quality controlled reflectivity data from the Melbourne NWS radar. From the LDAR sparkswe can calculate flash rate.

WDSS: Warning Decision Support System

Cross section view of cell 49 in previous slide

LNOx algorithm for WRF-Chem what are the necessary ingredients?

•Flash rate (F) will be calculated based on the following relationship: •F=AHn

•Relationship between Radar storm height above the freezing level and LDAR data is needed to determine A and n above.

•Radar Reflectivity is calculated within WRFChem using hydrometeorsfrom microphysical scheme (WSM6) coupled with the Kain-Fritschcumulus parameterization.

•Convective storm height (H) is found using the 20dbz contour and the freezing level is obtained from the WRF temperature field.

Lightning Flash Rate Parameterization

•Production rate of NO from both IC and CG lightning: -500 moles per flash (Ott et al.,in progress)

•Vertical Distribution of NO: Pickering et al., 1998: -Wind fields from Goddard Cumulus

Ensemble (GCE) model were used to redistribute LNOx throughoutthe duration of the storm. Profiles were constructed for mid-latitudecontinental, tropical continental, and tropical marine regimesbased on profiles computed for individual storms in each regime

LNOx algorithm for WRF-Chem what are the necessary ingredients?

LNOx parameterization

VerificationSimulations for Summer 2004 INTEX-NA period

over eastern US and comparisons with aircraft data

DC-8 flight track showingthe path through the Huntsville, AL storms

We have the same data forKennedy Space Center

Plot courtesy of Mike Porter

SummarySummary

Flash rate parameterization being developed for Flash rate parameterization being developed for

WRF-Chem usingWRF-Chem using observed radar and 3-D lightning observed radar and 3-D lightning mapping array data.mapping array data.

WRF-Chem will first be tested with the Futyan and WRF-Chem will first be tested with the Futyan and Del Genio (2007) relationship and then with the Del Genio (2007) relationship and then with the newly developed scheme.newly developed scheme.

Existing NO production per flash and vertical Existing NO production per flash and vertical distribution information will be used.distribution information will be used.

Testing of WRF-Chem with lightning will be Testing of WRF-Chem with lightning will be conducted using aircraft NOx observations from the conducted using aircraft NOx observations from the ICARTT (NASA and NOAA data) experiment from ICARTT (NASA and NOAA data) experiment from Summer 2004.Summer 2004.