14
Limb tracegases: verification with scientific processors K-U- Eichmann, C. von Savigny, H. Bovensmann, A. Kokhanovsky, (presented by M. Weber) Bremen 6.12.2010

Bremen 6.12.2010

Embed Size (px)

DESCRIPTION

Limb tracegases : verification with scientific processors K-U- Eichmann, C. von Savigny , H. Bovensmann , A. Kokhanovsky , (presented by M. Weber). Bremen 6.12.2010. IUP & DLR limb ozone retrieval:. IUP V2.1 (L1 V6/V7): - PowerPoint PPT Presentation

Citation preview

Page 1: Bremen 6.12.2010

Limb tracegases: verification with scientific processors

K-U- Eichmann, C. von Savigny, H. Bovensmann, A. Kokhanovsky,

(presented by M. Weber)

Bremen

6.12.2010

Page 2: Bremen 6.12.2010

IUP & DLR limb ozone retrieval:

• IUP V2.1 (L1 V6/V7):

– central wavelength of the Chappuis triplet is shifted from 601 to 589 nm

– Tikhonov parameter linearly decreases with altitude from 9 at 72 km to 3 at 7 km instead of constant value of 6

• IUP V2.2:

– Foward model is changed from SCIATRAN 2.2 to SCIATRAN 3.0

– Spectral region: 2 nm spectral intervals centered around: 264 nm, 267.5, 273, 283, 286, 288, 290.5, and 305 (UV) as well as 525, 589, and 675 nm (Chappuis band)

– Surface albedo: geographical database [Matthews, 1983]

• IUP V2.3:

– Use of a latitude dependent aerosol climatology ECSTRA [Bingen, 2000]

• OP 5.1 (DLR operational retrieval on L1 V7 )

– DOAS retrieval in Chappuis band

– altitude coverage: 15-40 km (IUP retrieval: 10-70 km)

Page 3: Bremen 6.12.2010

Differences between IUP 2.1/2.3 and OP5

• The differences between the IUP 2.1 and 2.3 are found mainly below 33 km down to the tropopause.

• These are mainly due to the use of the ECSTRA aerosol climatology in IUP2.3. Another reason could be different p/T inputs, IUP uses ECWMF data.

Page 4: Bremen 6.12.2010

Limb Ozone Verification: OP5.01 vs. ST2.3

201001-03200301-04• Verification of OP5 with SCIATRAN 2.3: 4 months of data in 2003 and 2010 were used.

• OP5 ozone is on average about 8% higher for all tangent heights. Results are similar for the years 2003 and 2010.

Page 5: Bremen 6.12.2010

O3: Latitude dependency (200301-04)

N50-90 N25-50

-25-+25

S50-90 S25-50

N50-90: +10%N25-50: +6%-25/25 : +10%S25-50: +8%S50-90: +6%

Page 6: Bremen 6.12.2010

O3: Sun zenith angle dependency (200301-04)

30-50 50-70 70-90• The difference between both models is generally lower than 10% for all

scenarios. • Highest differences are found in the lowest stratosphere below 20km. For

higher SZAs we find differences (>10%) in the upper stratosphere.

Page 7: Bremen 6.12.2010

Validation results: O3• Validation with ozone sondes (GAW-contributing ground-based networks, see D.

Hubert, SCIAVALIG validation meeting 09/2010): On average the SGP 5.01 processor showed ~10% higher O3 density values than SGP 3.01. O3 profiles retrieved with SGP 5.01 are of equivalent or better quality than those retrieved with SGP 3.01. E.G. Ascension 8°S(+10%), Uccle 50°N(~0%), Belgrano 78°S (+5%)

• Comparison with lidar, sonde and microwave 2002-2004 (VALID, see talk J.A.E. van Gijsel, SCIAVALIG validation meeting 09/2010): positive bias in the tropics (~12%), midlatitudes (~7%), negative bias in the polar region(~-8%).

Page 8: Bremen 6.12.2010

Summary O3 OP5.1 vs IUP 2.3

• The IUP ozone retrieval in V2.3 was finetuned to reduce the high bias found in the tropical regions. This leads to less ozone found in the midlatitudes and polar regions.

• Comparisons with the OP5.1 shows on average 8% lower values of the IUP2.3. This differs slightly for different latitude and sun zenith angle regions between 6 and 10%.

• Highest differences above 10% were found in the upper stratosphere at high SZA and near the troposphere.

• Comparing these results with the validation datasets:– Tropics: OP5.1 ~10% higher -> IUP2.3 more accurate as expected– Midlatitudes: “inconsistent” validation data: OP5.1 ~7% higher ->

IUP2.3 possibly nearer to validation results– Polars: “inconsistent” validation data: OP5.1 +/-6% -> low bias of

IUP2.3 can be expexted

Page 9: Bremen 6.12.2010

Limb NO2: OP3/4 (from PM4) and OP5.1

OP5.01: comparison with IUP V3.1 (200301-04)

• The OP5.1 NO2 gives basically the same results as OP4, which is as expected.

Page 10: Bremen 6.12.2010

NO2: Summary

• Differences between two retrievals of the OP5.1 and the IUP 3.1 are generally below 10% for the height range of 21 and 38 km.

• Highest differences above 20% were found for high SZAs in the upper stratosphere and lowest stratosphere.

Page 11: Bremen 6.12.2010

Limb BrO: all data 2003 (01—04)

•On average the differences between the BrO OP5.1 and IUP3.2 are below 10% between 20 and 30 km, but the scatter is 50% and more. These results are as expected.

Page 12: Bremen 6.12.2010

BrO: time dependent differences

201004 200304• Comparing data from 2003 and 2010:

• The differences between both models are slightly increasing.• The scatter drastically increases.

Page 13: Bremen 6.12.2010

BrO: Summary of results

• On average both models OP5.1 and IUP3.2 give similar results.• The differences and the scatter found are higher in 2010 than in 2003.• The differences also depend on altitude, latitude, and SZA:

– Polar NH: OP5.1 -15% (01/03) -> -5% (04/03), similar in 2010 – Polar SH: OP5.1 +50% (01/03) -> +5% (04/03)– Tropics: OP5.1 -10% (30km) / +40%(20km), no change over time

Page 14: Bremen 6.12.2010

Limb cloud top height - Verification with MIPAS

• Cloud top height between 9.0-25.0 km• Max. Difference in distance=300 km and seconds=990

sec.

• SCIAMACHY - MIPAS [km]: 200902– 3382 Diff= -0.9 (1.9)– 30 N >50 Mean=12.0 Diff=1.4 (2.0)– 542 N 15/50 Mean=12.4 Diff:-0.7 (1.7)– 2047 T Mean=14.9 Diff=-1.30 (1.7)– 762 S -15/-50 Mean=14.1 Diff=-0.0 (2.0)

• SCIAMACHY - MIPAS [km]: 200906– 3722 Diff=-1.2 (1.5)– 57 N >50 Mean=12.3 Diff=0.4 (2.0)– 1147 N 15/50 Mean=13.6 Diff=-1.0 (1.6)– 2051 T Mean=14.4 Diff=-1.4 (1.4)– 459 S -15/-50 Mean=12.6 Diff=-0.7 (1.3)– 8 S <-50 Mean=11.8 Diff=-0.8 (0.8)

• The comparison with MIPAS done for CTH above 9 km gives good results. The mean height is between 12 and 15km. The mean difference is on the order of -1 km.• Below about 9 km MIPAS has very low sensitivity.