Sensitivity of satellite observations for freshly produced lightning NO x
description
Transcript of Sensitivity of satellite observations for freshly produced lightning NO x
![Page 1: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/1.jpg)
Sensitivity of satellite observations for freshly produced lightning NOx
Steffen Beirle, MPI Mainz
M. Salzmann, M. G. Lawrence,
U. Platt, and T. Wagner
![Page 2: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/2.jpg)
Lightning NOx...
... is important:• Dominating source in tropical UT• Released in the UT: long lifetime, low background
levels
→ strong impact on tropospheric O3 and OH
... is uncertain:• „Magic number“ 5 Tg [N]/yr• uncertainty still 1order of magnitude• besides total number:
– regional variations– Intracloud flashes versus Cloud to ground– ...
![Page 3: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/3.jpg)
Satellite observations...
... provide new independent measurements
... have high potential:• time series• global cover
→ analysis of temporal & spatial patterns on different scales and for different conditions
→ use them to constrain LNOx estimates!
![Page 4: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/4.jpg)
Different approaches:1. mean NO2 enhancement due to lightning activity
![Page 5: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/5.jpg)
Different approaches:1. mean NO2 enhancement due to lightning activity
Disadvantages:
Observed NO2 (~10 a.m.) is aged LNO2
Dillution Transport Lifetime
![Page 6: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/6.jpg)
1. mean NO2 enhancement due to lightning activity
2. direct observation of freshly produced LNOx
Different approaches: 2. direct obs
![Page 7: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/7.jpg)
Beirle et al., ACP 6, 2006
1.7 (0.6-4.7) Tg [N]/year
Different approaches: 2. direct obs
![Page 8: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/8.jpg)
Direct observation of fresh LNOx
Satellite observations over active thunderstorm
+ confined region of interest
+ negligible chemical loss / dilution
(within satellite pixel, 30 min.)
- highly complex observation conditions (cumulonimbus clouds)
→ what do we actually "see"?
![Page 9: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/9.jpg)
Satellite sensitivity for (L)NOx
• Measured: Slant column density S• Wanted: Vertical column density V• S = V * A A: Air Mass
Factor
• S = V * ai * pi ai: box-AMF
pi: profile (norm.)
• SNO2 = VNOx * ai * li * pi li := [NO2] / [NOx]
• SNO2 = VNOx * ei * pi = VNOx * E
Visibility ei = ai * li "box-AMF for NOx"
Sensitivity E = ai *li * pi = ei * pi = SNO2 / VNOx
![Page 10: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/10.jpg)
Satellite sensitivity for (L)NOx
SNO2 = VNOx * ei * pi = VNOx * E
Visibility ei = ai * liSensitivity E = ei * pi
E = ai * li * pi
Idea: Use cloud resolving model results for a thunderstorm simulation to investigate temporal and spatial variability/evolution of ei and E!
![Page 11: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/11.jpg)
Satellite sensitivity for (L)NOx
SNO2 = VNOx * ei * pi = VNOx * E
Visibility ei = ai * liSensitivity E = ei * pi
E = ai * li * pi
Cloud resolving model
Radiativetransfer modelMCArtim
Hydrometeors
![Page 12: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/12.jpg)
CSRMC (Marc Salzmann)(Cloud system resolving model including photochemistry)
• 3d cloud resolving model ((photo)chemistry+dynamics)
• One week of thunderstorms in the TOGA COARE/CEPEX region (Pacific)
• Output grid: dt=30 min, dx=dy=2km, dz=500m
• Output: NOx, NO2, O3, T, J ...
• Extinction simply parameterized from cloud water&ice (literature)
• Snow, graupel and rain neglected (few large particles)
McArtim: Box-AMFs for all locations at all timesteps
![Page 13: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/13.jpg)
Salzmann et al.,ACP, 2008
NOx volume mixing ratios and qtotm = 0.01g/kg contourduring the developmentof a mesoscale convective system from 24 December1992, 10:00 UTC to 24 December 1992, 12:30 UTC from the LTN1 run.
![Page 14: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/14.jpg)
LNOx Profile NOx partitioning Box-AMF Visibility ~pi li ai ei
Modelled sensitivitiesSensitivity E = ai * li * pi = ei * pi = SNO2 / VNOx
COT=
E=Two examples
![Page 15: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/15.jpg)
LNOx Profile NOx partitioning Box-AMF Visibility ~pi li ai ei
Modelled sensitivitiesSensitivity E = ai * li * pi = ei * pi = SNO2 / VNOx
COT=
E=EnsembleMeans
![Page 16: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/16.jpg)
Modelled sensitivitiesSensitivity E = ai * li * pi = ei * pi = SNO2 / VNOx
... show small dependence on COT
... show small variability
![Page 17: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/17.jpg)
Spatial patterns
• Spatial mean sensitity (relevant for applications): E=0.46 (0.37-0.55)
![Page 18: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/18.jpg)
For more details...
... see Beirle et al., Sensitivity of satellite observations for freshly produced lightning NOx, ACPD (“in print”), 2008.
In addition: – Error discussion– Impact of RTM settings (negligible!)– Impact of satellite resolution (does not change
results)
– Impact of background NOx: critical! (shielding!)
![Page 19: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/19.jpg)
Visibility & sensitivity for LNOx: results
• Satellites "look" deep inside the cloud• But at the cloud top, there is nothing to see
– Cloud top: ei~0 (due to low li)
– Ground: ei~0 (due to low ai)
– Cloud middle: ei maximum (1-2)
• We find low variability of E over one week of thunderstorm simulations!
• On average, E is lowest over anvils• Mean E=0.46 (0.37-0.55)
• Sat. meas. are sensitive for LNOx!
![Page 20: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/20.jpg)
Beirle et al., ACP 6, 2006
1.7 (0.6-4.7) Tg [N]/year
Gulf of Mexico revisited
E0.25 (inconsistent literature sources for pi, li, ai)
0.9 Tg [N]/year
E=0.46 (this study)
![Page 21: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/21.jpg)
Outlook• One model study: How representative?
– Repeat study for other thunderstorms (using other models)
• One empirical study: find more!– Systematic search for coincidences of flashes (WWLLN)
during SCIAMACHY / GOME-2 overpass: High variability!
• Possible reasons:– regional differences of LNOx production efficiency
(Huntrieser et al., 2008)
– general high variability of LNOx production/flash?
– need of higher moments / pdfs instead of means?
• Satellite observations have the potential to improve our knowledge on LNOx!
![Page 22: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/22.jpg)
Thanks for your attention
![Page 23: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/23.jpg)
![Page 24: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/24.jpg)
![Page 25: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/25.jpg)
Sensitivity results
• Single sensitivities vary (~0.1-1)
• Mean sensitivity (~0.3-0.7) always higher than in Mexico-Study (0.25) !
→ is the modelled storm representative?
• With these sensitivities, every active thunderstorm (>500 flashes per SCIAMACHY pixel 30x60km2)should be clearly visible from space (SNO2>2.5e15 molec/cm2)
→ what do we see?
![Page 26: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/26.jpg)
Observations of fresh LNOx
• GOME-event in Gulf of Mexico:– perfect spatial/temporal matching– lightning information (NLDN)
• Meanwhile:– 4 years of SCIAMACHY data– 3 years of WWLLN data (global continuous ground
based lightning counts)
• Automated search for „lightning events“ prior SCIAMACHY overpass: 1680 matches!!!
• keep in mind: 10:00 local time!
→ Coincidences almost all over oceans
![Page 27: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/27.jpg)
![Page 28: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/28.jpg)
Some examples...
Flash time CloudsNO2
Australia
Red: <30 minBlue: 90-120 minGrey: 2-24 h
Tropospheric Slant Column
SNO2
FRESCO CF&CTHBlack: cloud freeColoured: cloudyRed: high cloudBlue: low cloud
1015
![Page 29: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/29.jpg)
Some examples...
In the middle of nowhere...
Pacific
1015
Flash time CloudsNO2
![Page 30: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/30.jpg)
Some examples...
Maximum > 2e16!
USA
1015
Flash time CloudsNO2
![Page 31: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/31.jpg)
What happens here???
Korea
1015
Flash time CloudsNO2
![Page 32: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/32.jpg)
What happens here???
China
Hong Kong
1015
Flash time CloudsNO2
![Page 33: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/33.jpg)
What happens here???
Australia
1015
Flash time CloudsNO2
![Page 34: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/34.jpg)
What is different???
1015
Flash time CloudsNO2
Australia
![Page 35: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/35.jpg)
Global pattern
SNO2Regional differences:• Higher values over land • Higher values close to strong ground sources, indicating deep convection• Higher values east from Australia???
![Page 36: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/36.jpg)
Statistics: SNO2 for coinciding lightning events
Every dot: SCIA-Pixel withmore than 5 WWLLN flashesand cf>0.5
SNO2
log10(flashes)
![Page 37: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/37.jpg)
Statistics: S NO2 for coinciding lightning events
More than
0
10
100
1000
Flashes within the last hour
SNO2Every dot: SCIA-Pixel withmore than 5 WWLLN flashesand cf>0.5
CTH
![Page 38: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/38.jpg)
Some open questions
• Where is the LNOx?
– observed SNO2 enhancement for SCIA-pixels with >1000 flashes (corrected for WWLLN sensitivity):1.6*1015 molec/cm2 on average
– expected:5*1016 molec/cm2 !!!
• How to hide LNOx completely?
– in no simulation run E<0.1
• What makes the differencebetween good and no signal?
30* higher !!!
![Page 39: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/39.jpg)
E = 0.25 (0.14-0.5)
LNOx profile:
Cloud resolving modelsPickering et al. 1998
Fehr et al. 2004
fraction of total
NOx partitioning:
In-situ measurements in New Mexico for cb conditionsRidley et al. 1994, 1996
[NO2]/[NOx]
Box AMFs: (sensitivity)
RTM, cb conditionsHild et al. 2002
Box AMF
Problem: different, possibly inconsistent sources→ use one source: cloud resolving model
SNO2 = VNOx * ai * li * pi
li
piai
![Page 40: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/40.jpg)
LNOx Profile NOx partitioning Box-AMF Visibility ~pi li ai ei
Modelled sensitivitiesSensitivity E = ai * li * pi = ei * pi = SNO2 / VNOx
COT=
E=EnsembleMeans
Lowest sens.
![Page 41: Sensitivity of satellite observations for freshly produced lightning NO x](https://reader035.fdocuments.us/reader035/viewer/2022062322/56814f92550346895dbd4e78/html5/thumbnails/41.jpg)