How Important Are Temporal Constraints and Vertical Injection of Boreal Fire Emissions?

Yang Chen1,3, Qinbin Li1,2, James Randerson3, Evan Lyons2 Ralph Kahn1,4, David Nelson5, David Diner1

1JPL 2UCLA 3UC Irvine 4GSFC 5Raytheon

The 4th GEOS-Chem Meeting, April 7, 2009, Harvard

Summer 2004 Alaska Fire Emissions: Temporal Constraints

GFED v2

Diurnal: A mean diurnal cycle imposed on the 8-day emissions

Synoptic: Synoptic variation imposed on the diurnal emissions

Temporal constraints are added cumulatively

Major fires

Significant diurnal and synoptic variability.

Summer 2004 Alaska Fire Emissions: Vertical Injection

Circles: MISR-derived smoke plume heights Vertical distributions of emissions

600+ plumes10% reached free trop.

GEOS-Chem Simulations with Different GFED Configurations

Relative Importance of Individual Temporal ConstraintsLittle effect from imposing a mean diurnal cycle.

Effects of Different Smoke Plume Heights on COUniform injection significantly decreases surface CO over source region while increases CO in the upper troposphere – Not a surprise.

Sensitivity of CO to Temporal and Vertical ConstraintsSurface CO over source region decreases while downwind CO increases

Effects on the Columns of CO and BC

1. Columns decrease over source region while increases downwind.

2. Most of that change is due to added temporal variations.

3. CO column change ~2%; BC: 10-20%

kg/km2

Effects on the Burdens of CO and BC

CO and BC burdens from N.A. biomass burning

Major fires

Effects on Downwind Vertical Distributions of CO and BC

INTEX-NA 2004, all flights

Black box: Median obs.

1. Uniform injection may overestimate upper tropospheric CO.

2. Not much difference among the temporal constraints and vertical injections.

3. Worse agreement in the lower troposphere.

Effects on the Vertical Distributions of CO and BC

Individual flights during INTEX-NA 2004

Any Improvements to CO Column Simulation?

Effects on Surface BC: Comparison with IMPROVE

µg/m3

1. Going from monthly to 8-day makes largest difference.2. Uniform injection makes things worse – underestimates surface BC.

Temporal

Vertical

Temporal

Vertical

Transport of Boreal Fire CO: The Role of Convection

(a) 300 hPa CO from N.A. biomass burning

(b) B.B. Emissions

(c) Vertical dist. of NA biomass burning CO and convective mass flux

(d) Change in CO due to MISR-derived plume injection

(e) Change in CO due to uniform plume injection

White line: HYSPLIT trajectory

Convection is possibly more important than plume injection height?

Backup slides

Emission-weighted Winds and Deep Convective Mass Fluxes

Positive values: Eastward, northward, upward winds and fluxes

Stronger vertical transport and horizontal advection with higher-temporal frequency emissions compared with the monthly mean emissions.

Any Improvement to CO Column? Temporal Comparison with MOPITT

MOPITT CO with a priori fraction < 50%

1. Some improvement from temporal constraints.2. Uniform injection makes things worse.

Any Improvement to CO Columns? Temporal Comparison with MOPITT

MOPITT CO with a priori fraction < 30%

Any Improvements to CO Column Simulation?

Effects on AOD: Comparison with AERONET AODIf anything, going from monthly to 8-day makes largest difference.