Preliminary comparisons of GOMOS and MIPAS ozone products with Odin

Page 1Validation by Model Assimilation and/or Satellite Intercomparison - ESRIN 9–13 December 2002

Preliminary comparisons of GOMOS and MIPAS ozone products with Odin

AOID 760Donal Murtagh, Ashley Jones, Patrick

Eriksson, Carlos JimenezChalmers University of Technology

Andreas JonsonStockholm University

+ Odin Teams


Aims

To provide cross-validation of Odin and ENVISAT measurements of atmospheric constituent profiles.

To assess errors and biases in the data from different instruments

To help improve data products from both platforms


Methodology

Assimilation of Envisat measurement data into the DIAMOND model during an extended period to ensure good model measurement agreement

Comparison with Odin and Sonde measurements obtained during the period.

Chosen period : 24/10/2002 to 06/11/2002 Odin data available during 8 of the chosen days


DIAMOND (Dynamical Isentropic Assimilation Model for OdiN Data)

Forecast model: Transport scheme on isentropic surfaces ECMWF winds 100 km grid spacing

Assimilation scheme: Statistical interpolation/ Optimal Estimation analysis errors, model error growth

Observations for model evaluation: UARS MLS O3 version 4 and Odin SMR data


Setup for the Envisat Comparisons

DIAMOND model modified to use MJD as the time coordinate

Vertical coordinate 400 – 1000 K potential temp in steps of 50 K ( ~2 km)

Run for MIPAS ozone and N2O profile data for the period 24/10/2002 to 06/11/2002 = MJD 52569 – 52584

Some runs have also been made for the GOMOS reference set but not reported on here due to lack of overlap with other measurements


-150 -100 -50 0 50 100 150

-80

-60

-40

-20

0

20

40

60

80

Assimilated MIPAS ozone on 550 K for 2002-11-6

1.5

2

2.5

3

3.5

4

4.5

x 10-6


-150 -100 -50 0 50 100 150

-80

-60

-40

-20

0

20

40

60

80

MIPAS N20 on 450 K surface 2002-11-6

1

1.5

2

2.5

3

x 10-7


Composite plot Initial Model values against Mipas

0 0.2 0.4 0.6 0.8 1 1.2 1.4

x 10-5

0

0.2

0.4

0.6

0.8

1

1.2x 10

-5

Measured ozone

Mod

el O

zone

Day 52571 = 2002-10-24


Composite plot Model against Mipas-O3 after 10 days assimilation

0 0.2 0.4 0.6 0.8 1 1.2 1.4

x 10-5

0

0.2

0.4

0.6

0.8

1

1.2x 10

-5

Measured ozone

Mod

el O

zone

Day 52584 = 2002-11-6


Composite plot Model against Mipas–N2O after 10 days assimilation

0 0.5 1 1.5 2 2.5 3 3.5 4

x 10-7

0

0.5

1

1.5

2

2.5

3

3.5

4x 10

-7 Day 52584 = 2002-10-24

Measured N2O

Mod

el N

2O


Comparison with Balloon sondes

Selected sondes during the assimilation period have been compared to the closest model point at the same time as the balloon ascent.


Balloon Sonde Comparisons

0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4200

300

400

500

600

700

800

900

1000Graph of MIPAS vs balloon plot for Ny Ålesund N78 E12 on 30-10-2002 0940z

EP

T (

K)

O3 concentration (PPMV)

real ozone12/213 13/214 14/215

sonde ozone12/21313/21414/215

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8200

300

400

500

600

700

800

900

1000Graph of MIPAS vs balloon plot for Payern N47 E7 on 28-10-2002 1112z

EP

T (

K)

O3 concentration (PPMV)

real ozone48/208 49/209 50/210

Sonde ozone48/20849/20950/210

Payern 28/10/2002Ny Ålesund 30/10/2002


Odin Comparisons

Most comprehensive Ozone Validation for Odin products has been carried out for the Optical Limb scattering inversions using the Flittner Method (CvS)– Agreement with independent data is better than 10% on an individual profile basis

Comparisons of the 501GHz Sub-mm data have been carried out against the co-located Flittner profiles


SMR 501 GHz Ozone vs Flittner

-2 -1.5 -1 -0.5 0 0.5 1 1.5 2

x 10-6

10

15

20

25

30

35

40

45

50Average difference SMR - Flittner

Mixing Ratio

Alti

tude

(km

)

mean (SMR-Flittner)Std (SMR-Flittner)


MIPAS / Odin Ozone comparision

0 2 4 6 8 10 12 14400

500

600

700

800

900

1000

Pot

entia

l Tem

pera

ture

(K

)

Mixing Ratio (ppmv)

Comparision Assimilated MIPAS Odin, lat 1.8 long -90.9449

MIPASOdin


MIPAS / Odin SMR 501GHz Ozone comparision

0 1 2 3 4 5 6 7 8 9400

500

600

700

800

900

1000

Pot

entia

l Tem

pera

ture

(K

)

Mixing Ratio (ppmv)

Comparision Assimilated MIPAS Odin, lat -46.8 long 90.7169

MIPASOdin


0 1 2 3 4 5 6 7 8400

500

600

700

800

900

1000

Pot

entia

l Tem

pera

ture

(K

)

Mixing Ratio (ppmv)

Comparision Assimilated MIPAS Odin, lat -57.6 long 85.662

MIPASOdin



0 1 2 3 4 5 6 7 8 9400

500

600

700

800

900

1000

Pot

entia

l Tem

pera

ture

(K

)

Mixing Ratio (ppmv)

Comparision Assimilated MIPAS Odin, lat 42.3 long -76.6598

MIPASOdin



0 0.5 1 1.5 2 2.5 3

x 10-7

400

500

600

700

800

900

1000Comparision Assimilated MIPAS Odin, lat -54 long 125.5561

MIPAS / Odin SMR 501GHz N2O comparision


MIPAS/Odin N2O comparision

0 0.5 1 1.5 2 2.5 3 3.5

x 10-7

400

500

600

700

800

900

1000Comparision Assimilated MIPAS Odin, lat 34.2 long 127.3153


MIPAS/Odin N2O comparision

0 0.5 1 1.5 2 2.5 3 3.5

x 10-7

400

500

600

700

800

900

1000Comparision Assimilated MIPAS Odin, lat 81 long 11.5064


Conclusions and Future Work

MIPAS Ozone and N2O products seem to be in good agreement with Odin measurements and balloon sondes.

More work required to put this on a quantitative basis and to look for biases.

Model will be properly tuned for N2O and the spatial characteristics of the data sets

Monitoring mode will be introduced to allow direct comparison of assimilating data sets and independent data

Preliminary comparisons of GOMOS and MIPAS ozone products with Odin

Documents