Origin of Reactive N in Rocky Mountain NPnadp.slh.wisc.edu/committees/minutes/spring2013/TDep... · 2013-06-21 · Origin of Reactive N in Rocky Mountain NP • Bret A. Schichtel

Origin of Reactive N in Rocky Mountain NP

• Bret A. Schichtel1, William C. Malm3, Mike Barna1, Krisiti Gebhart1, Katie Benedict2, Anthony Prenni2, Jr.2, Christian M. Carrico2 , Ezra Levin2, Derek Day3, Doris Chen2,

John Ray1, Jeffrey L. Collett2, Sonia M. Kreidenweis2

1National Park Service, CSU/CIRA, Fort Collins, Colorado 80523 2Department of Atmospheric Science, Colorado State University

3Cooperative Institute for Research in the Atmosphere (CIRA), CSU

NOx and NH3 U.S. Emissions: 2006 verse 2050 Projection

• It is projected that the U.S. will shifting from oxidized to reduced dominated RN emissions by 2050 Raluca Ellis et al., 2012

Characterize the atmospheric concentrations of sulfur and reactive nitrogen species in gaseous, particulate and aqueous phases along the east and west sides of the Continental Divide

Identify the relative contributions to atmospheric sulfur and nitrogen species in RMNP from within and outside of the state of Colorado. from more emission sources along the Colorado Front

Range versus other areas within Colorado. from mobile sources, agricultural activities, large and

small point sources within the state of Colorado.

Rocky Mountain Airborne Nitrogen and Sulfur (RoMANS) Study Objectives

Rocky Mountain NP Deposition Studies RoMANS I: 2006 – March/April (spring) and July/August (summer) RoMANS II: November 2008-November 2009 April – September 2010

• Particle composition and gases • 24 hr PM2.5 and composition • 15 minute PM2.5 ions (PILS) • 24 hr SO2, NH3 and HNO3 (URG) • Continuous NOx, NOy, NH3, O3, CO • Weekly HiVol – PM2.5

• Wet deposition • Event and sub-event/hourly • Ion chromatography • Org N = TN – inorg N

• Other measurements • Co-located w/ IMPROVE & CASTNet monitors • 2006 study had monitors throughtout RMNP and

at Colorado borders

Rocky Mountain, NP Reactive N Deposition Budget

• Wet organic N + Dry NH3 ~35% of total N deposition

• Dry organic N is still missing

0

10000

20000

30000

40000

50000

60000

Nitr

ogen

Dep

ositi

on (µ

g N

/m2)

Wet ONWet NH4Wet NO3Dry HNO3Dry NH3Dry NO3Dry NH4

Nov 2008 - Nov 2009 Annual N dep=3.4

kg/ha/yr

• Spring N deposition is primarily from northeastern CO and Front Range due to a synoptic upslope event

• Summer N deposition is from a diverse set of sources

• Local sources contribute ~25% of summer NH3 deposition

Recall Source Attribution Budgets from 2006 ROMANS I Study

Ambient Ammonia Concentrations Nov 2008 – Nov 2009

Synoptic Upslope event

April June

• Concentration gradients. • Which way is the wind coming from? • Simple back trajectories. • Frequency with which the air mass passes over

source areas before it arrives at the receptor -residence time analysis.

• Receptor models. • Trajectory receptor models. • Source Oriented Chemical transport models. • Hybrid Models.

RMNP RN Apportionment Strategy (Weight of Evidence)

Summer Diurnal Cycles

Box and whisker plots with the median, 10, 25, 75, and 90 percentiles shown

NH3 NOy

NOx

PM1

Temperature Wind direction

PILS and Winds, 2009

Spring/Summer o Daily upslope o Increased Concentrations o Generally associated with precipitation.

Fall o Fewer events but similar to spring o Important synoptic scale upslope event

90th Percentile of concentrations as a function of wind direction

Average Wind Rose, Nov 08 – Nov 09

The frequency the measured surface winds at RMNP are from a given direction and the average NH3 concentrations associated with each wind direction

Sources Winter Spring Summer Fall Annual East 29% 44% 45% 49% 44%

West 71% 56% 55% 51% 56%

The estimates of the contribution of sources east and west of RMNP to the seasonal and annual measured NH3 concentrations.

The average contribution of the NH3 concentrations associated with each wind sector to the RMNP annual average NH3.

Core Observed

wrf at Core Site wrf 12 km West wrf 12 km East

Measured 10-m Wind Rose Vs.

WRF 4-km Modeled Nov 2008 – Dec 2009

Transport Modeling

< 0.0010 001 t 0 005

ime/Equal Prob. Sfc.n endpts/grid= 1

< 0.0010 001 t 0 005

me/Equal Prob. Sfc. endpts/grid= 1

< 0.0010 001 t 0 005

me/Equal Prob. Sfc. endpts/grid= 1

< 0.0010 001 t 0 005

me/Equal Prob. Sfc. endpts/grid= 1

< 0 .0 0 10 .0 0 1 to 0 .0 0 50 .0 0 5 to 0 .0 10 .0 1 to 0 .0 20 .0 2 to 0 .0 50 .0 5 to 0 .10 .1 to 0 .1 50 .1 5 to 0 .20 .2 to 0 .2 5> 0 .2 5

R e s . T im e /E q u a l P ro b M in e n d p ts /g r id = 1

Winter

Fall Summer

Spring

Distance Weighted Residence Time Where did the air come from on an average day?

< -0.0001 & No Data

rential Probability endpts/grid= 1

< -0.0001 & No Data0 0001 t 7 5 005

rential Probabilityn endpts/grid= 1

Incremental Probability – Transport on High NH3 Days Compared Average Days

< -0.0001 & No Data

ential Probability endpts/grid= 1

< -0.0001 & No Data

ential Probability endpts/grid= 1

< -0 .0001 & N o D-0 .0001 to -7 .5e-0-7 .5e-005 to -5e-0-5e-005 to -2 .5e-0-2 .5e-005 to 00 to 2 .5e-0052 .5e-005 to 5e-0055e-005 to 7 .5e-0057 .5e-005 to 0 .0001> 0 .0001

D ifferen tial P robab ilityM in endp ts/g rid= 1

Winter

Fall Summer

Spring

Regression Techniques of Source Apportionment Assumption: On average, the concentration measured at the receptor is some linear combination of the contributions of several sources.

Concentration = a1 Source1 + a2 Source2 + … TrMB (Trajectory Mass Balance)

Use many concentrations of 1 species and counts of trajectory endpoints in source regions to predict average attributions over a long period.

Source Regions

ijijj

i PtsEnda=NH ε+∑ _#3

j - Source Regions i - observations

PtsEndaI=ContSrc jij

ij _#1_ ∑

Colorado (%)

Eastern U.S. (%)

Western U.S. (%)

URG 24-hr Conc (ug/m3)

Spring 2006 56 ± 5 8 ± 2 35 ± 4 0.14

Summer 2006 50 ± 7 21 ± 3 29 ± 5 0.35

Winter 2008/09 48 ± 14 5 ± 7 46 ± 14 0.09

Spring 2009 41 ± 10 14 ± 8 46 ± 10 0.32

Summer 2009 48 ± 11 8 ± 6 44 ± 11 0.31

Fall 2008/09 47 ± 12 10 ± 7 43 ± 12 0.23

TrMB Attributions of Ammonia Concentrations 2006 and 2009

Combining both hourly and 24-hour concentrations, 720 trials each for 2008/2009.

About half of the NH3 is from Colorado sources Results are inline with ROMANS I

NE Colorado (%)

SE Colorado (%)

Front Range (%)

W Colorado (%)

Spring 2006 7 ± 1 NA 9 ± 3 39 ± 4 Summer 2006 9 ± 5 NA 15 ± 4 24 ± 5 Winter 2008/09 10 ± 8 1 ± 2 5 ± 6 32 ± 14 Spring 2009 8 ± 4 2 ± 3 4 ± 3 26 ± 10 Summer 2009 11 ± 5 1 ± 1 6 ± 3 31 ± 12 Fall 2008/09 16 ± 7 3 ± 5 6 ± 4 22± 11

For 2009 Combed both hourly and 24-hour concentrations, 720 trials each.

TrMB Attributions of Ambient Ammonia Concentrations 2006 (final) and 2009

• Northeast CO + Front Range account for 12 - 22% of NH3

• Western CO contributes 22-32% of NH3 • Results are inline with ROMANS I • Wind fields likely underestimate contributions from east of ROMO

ikjikjj

ki Sa=C ε+∑Where

k = 1 ... m, the number of observations, i = 1 ... n, the number of sources contributing to receptor, j = 1 ... N, the number of source region vectors, Cki = concentrations of ammonia from source i for time period k, akj = time weighting functions, and Sji = source vectors. εik = error term including random and lack of fit error.

( ) ∑ Φ=j

jkjkkC α

Ck = the measured hourly aerosol concentrations Αjk = the regression coefficients Φjk = the average of modeled concentrations arriving at the receptor from sources areas grouped according to eigenvectors, v

Hybrid Receptor Model:

Source contributions were estimated by statistically relating transport from source regions to ROMO measurements

Hybrid Receptor Model • Similar to TRMB model except:

• Simulated the transport of inert NH3 emissions to the Rocky Mnt receptor site. • E.g. Emission weighted transport

• Statistically grouped collinear source regions • Did not apriori select all source regions

• Incorporated constant dry NH3 deposition

NOx

NH3

Source regions

• National scale emissions divided into ~100 source regions

• Regional scale model used to simulate transport of emissions from each region to Rocky Mountain NP

Simulated transport of “inert” NH3 emissions from western CO

Hybrid Model verse Measured NH3

Hybrid modeling ambient NH3 apportionment results compared to the tracer CAMX concentrations weighted for dry deposition.

Inert Tracer + dry deposition

Hybrid Model

Fraction of ammonia apportioned to sources in and outside Colorado

• About 55% of the ambient ammonia in all seasons is apportioned to sources inside of Colorado

• In ROMANS I spring and summer ~70% of ammonia was from Colorado sources

Ambient Ammonia Source Apportionment

Ambient Ammonia Source Apportionment • ROMANS II ROMANS I

Source regions are not exactly the same, e.g. Front Range + Northeastern CO = Denver + Northeastern CO

Both ROMANS I and II show eastern CO to be a significant contributor RMNP NH3; ROMANS II winds underestimate transport from eastern Colorado

ROMANS II shows contributions from broader array of source regions Northern Utah and Snake River Valley were not contributions in ROMANS I

Summary • About 55% of the NH3 originates from sources within Colorado

• Little of the measured ammonia at Rocky Mountain originated from the Midwest, an area with high ammonia emissions.

• Sources west of Colorado contribute ~35-45% of measured ammonia • California ~ 13% • Utah ~ 8%, • Snake River valley ~ 5-6% • Arizona ~ 5-6%.

• Rocky Mountain NP would benefit from regional reduction in NH3 emissions

Summary

• Contributions of NH3 from sources within the state of Colorado to Rocky Mountain NP • Western Colorado ~ 15%

• Central Colorado ~ 15%

• Front range including the Greely area ~ 15%.

• Northeastern Colorado ~ 6%.

• The Yampa river valley ~ 4%.

• CAMx PSAT source apportionment results will be available soon.