Maarten Krol, IMAU,

1
How to locate sources of atmospheric
constituents?
Maarten Krol, IMAU, Utrecht University,
Netherlands Atmospheric Chemistry Applications
Workshop Estec, 20 21 Jan. 2004 Sander
Houweling (SRON/IMAU) Peter Bergamaschi (JRC/
Ispra Italy) Wouter Peters (NOAA/ Boulder USA)
2
METEOR Cruise 2002
courtesy Valerie Gros Jonathan Williams
3
Backplume Concept

• Sensitivity of measurement at time t and location
  x for constant emissions at location y
• dCx,t/dEy
• Synoptical time scale typical 10 days back in
  time
• Calculated with adjoint TM5 model

4
TM5 zoom model
http//www.phys.uu.nl/tm5
5
(No Transcript)
6
MINOS 2001 back-plume calculated every 6 hours
7
Region of influence MINOS campaign (1/8/2001 -
19/8/2001)
8
(No Transcript)
9
(No Transcript)
10
MINOS methyl chloroform measurements (binned
6-hourly)
11
30 singular vectors (of 58)
12
(No Transcript)
13

• Formal Inversion
• ill-posed
• quantifying emission difficult
• Need more measurements/prior information
• Back-Plume concept
• Valuable information present
• Qualitative source localization
• Possible role satellite data?

14
CH4 columns 15-28 Feb. 2003
15
Role of satellite data

• Good points
• many independent measurements
• Good alignment with models
• Weak Points
• Accuracy of measured columns
• Biases due to clouds, aerosols

16
The end
17
Case 2 noise 10 Singular Vectors
18
Case 2, 55 Singular vectors
19
(No Transcript)
20
Do not use all degrees of freedom reason you
adjust emissions to explain noise