Supplement of Atmos. Meas. Tech., 9, 2497–2534, 2016 http://www.atmos-meas-tech.net/9/2497/2016/ doi:10.5194/amt-9-2497-2016-supplement © Author(s) 2016. CC Attribution 3.0 License.
Supplement of Ground-based assessment of the bias and long-term stability of 14 limb and occultation ozone profile data records Daan Hubert et al. Correspondence to: D. Hubert (
[email protected])
The copyright of individual parts of the supplement might differ from the CC-BY 3.0 licence.
D. Hubert et al.: Supplement to Ground-based assessment of 14 limb and occultation ozone profile data records Contents This document contains following supplementary figures which could not be shown in the main manuscript due to space limitations. – Fig. S1: latitude-time cross section of the co-location samples used in the analysis. – Fig. S2: vertical profile of drift uncertainty adjustment factor κ for ozonesonde and lidar networks, and all satellite records. – Fig. S3: sensitivity of the network-averaged drift results to several regression analysis parameters. – Fig. S4: sensitivity of the network-averaged drift results to selection of ground sites. – Fig. S5–S18: bias, comparison spread and networkaveraged drift results in four ozone profile representations. One figure per satellite record; results are split up per type of correlative instrument and latitude band to allow a more quantitative reading of the results. – Fig. S19: bias relative to ozonesonde and lidar of sunrise and sunset observations by solar occultation instruments. – Fig. S20: month-dependence of the bias and comparison spread of seven sounders relative to Arctic ozonesondes.
1
2
D. Hubert et al.: Supplement to Ground-based assessment of 14 limb and occultation ozone profile data records
Figure S1. Latitude-time cross section of the co-locations of NDACC/GAW/SHADOZ ozonesonde (blue) and NDACC stratospheric ozone lidar (red) with fourteen limb/occultation ozone profile records. Latitude and time correspond to that of satellite profile. Total number of co-locating pairs for sonde and lidar are mentioned separately between brackets.
D. Hubert et al.: Supplement to Ground-based assessment of 14 limb and occultation ozone profile data records
3
90N–90S 45
10
30
20
25 20
50
15 lidar
100
Altitude [km]
HALOE v19 POAM II v6 POAM III v4 SAGE III v4.0
GOMOS IPF6.01
10
30
20
25 20
50
15
100
10
200
5
500
1
UARS MLS v5
OSIRIS v5.07
ozonesonde
0
SAGE II v7.0
SMR v2.1
2
Pressure [hPa]
Altitude [km]
5
35
Pressure [hPa]
2
40
MIPAS ML2PP6.0 SCIAMACHY SGP5.02 ACE−FTS v3.0 MAESTRO v1.2 Aura MLS v3.3
3
χ²-adjustment factor κ Figure S2. Adjustment factor κ to incorporate unaccounted for sources of uncertainty (sampling, ground-network homogeneity, ...) in the estimation of network-averaged drift of limb/occultation ozone profile data relative to lidar (top) and ozonesonde (bottom) measurements. See Sect. 4.1.2 for more details.
4
D. Hubert et al.: Supplement to Ground-based assessment of 14 limb and occultation ozone profile data records
45 40
SAGE II v7.0
OSIRIS v5.07
SCIAMACHY SGP5.02
10/1984 - 8/2005
10/2001 - 5/2013
8/2002 - 4/2012
Aura MLS v3.3
HALOE v19
8/2004 - 5/2013
10/1991 - 11/2005
Altitude [km]
35 30 25 20 15 10 5
2
-2.5
-1.5
-1 -0.5
0
0.5
∆α [%/dec]
1
1.5
2
2.5
Sensitivity network-averaged drift ∆(alternative set-up – baseline set-up)
5
Pressure [hPa]
-2
10
lidar
ozonesonde
20
bootstrap-derived uncertainty (vs analytic expression)
50
monthly-averaged time series (vs daily-averaged)
100
linear & annual regression term (vs only linear term)
200 500 -2.5
45 40
-2
-1.5
-1 -0.5
0
0.5
∆α [%/dec]
1
1.5
2
2.5
-2.5
-2
-1.5
-1 -0.5
0
0.5
∆α [%/dec]
1
1.5
2
2.5
SAGE II v7.0
OSIRIS v5.07
SCIAMACHY SGP5.02
10/1984 - 8/2005
10/2001 - 5/2013
8/2002 - 4/2012
Aura MLS v3.3
HALOE v19
8/2004 - 5/2013
10/1991 - 11/2005
Altitude [km]
35 30 25 20 15 10 5
2
-2.5
-1.5
-1 -0.5
0
0.5
∆σ [%/dec]
1
1.5
2
2.5
Sensitivity network-averaged drift uncertainty ∆(alternative set-up – baseline set-up)
5
Pressure [hPa]
-2
10
ozonesonde
20
lidar
bootstrap-derived uncertainty (vs analytic expression)
50
monthly-averaged time series (vs daily-averaged)
100 200
linear & annual regression term (vs only linear term)
500 -2.5
-2
-1.5
-1 -0.5
0
0.5
∆σ [%/dec]
1
1.5
2
2.5
-2.5
-2
-1.5
-1 -0.5
0
0.5
∆σ [%/dec]
1
1.5
2
2.5
∗ Figure S3. Sensitivity of ozonesonde and lidar network-averaged drift α ¯ (top part) and one sigma uncertainty σα ¯ (bottom part) to several regression analysis parameters (cfr. Sec. 4.1.3). Shown is the difference of an alternative minus the baseline regression set-up for five satellite records. Three alternatives are depicted: (a) uncertainties derived from bootstrap (red) instead of analytic expression, (b) regression of monthly-averaged comparison timeseries (green) instead of daily-averaged, and, (c) regression model with linear and annual terms (yellow) instead of only linear term. Larger differences appear for the latter set-up since only timeseries with a significant annual cycle are regressed, hence leading to a quite different sample of timeseries included in the network-averages.
D. Hubert et al.: Supplement to Ground-based assessment of 14 limb and occultation ozone profile data records
Sensitivity network-averaged drift results:
∆(all sites – selected sites)
45
10
30
20
25
50
20
100
lidar ozonesonde
10
Altitude [km]
30
20
25 20
50
15
100
10
200
5
500
-2
-1
0
1
∆α [%/dec]
2
-2
-1
0
1
2
∆σ [%/dec]
-2
-1
0
1
Pressure [hPa]
Altitude [km]
5
35
Pressure [hPa]
2
40
15
5
2
∆κ
SAGE II v7.0
SAGE III v4.0
SCIAMACHY SGP5.02
UARS MLS v5
SMR v2.1
ACE−FTS v3.0
HALOE v19
OSIRIS v5.07
MAESTRO v1.2
POAM II v6
GOMOS IPF6.01
Aura MLS v3.3
POAM III v4
MIPAS ML2PP6.0
∗ Figure S4. Sensitivity of ozonesonde (bottom) and lidar (top) network-averaged drift α, ¯ one sigma uncertainty σα ¯ and κ-adjustment factor to the selection of ground sites (cfr. Sec. 4.2, Tables 1 and 2). Shown is the difference of the result obtained for the entire ground network minus the result for the selected sites, for the fourteen satellite records.
D. Hubert et al.: Supplement to Ground-based assessment of 14 limb and occultation ozone profile data records 60N–90N
30N–60N
30N–30S
30S–60S
60S–90S
45
5
30
10 20
25
50
20
Altitude [km]
ozonesonde 30
10
25
20
20
50
15
100
10
200
5
500 -15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
Bias b [%]
Bias b [%]
Bias b [%]
Bias b [%]
Bias b [%]
60N–90N
30N–60N
30N–30S
30S–60S
60S–90S
10
45
Altitude [km]
35
5
30
10 20
25
50
20
100
lidar
15
Altitude [km]
ozonesonde 30
10
25
20
20
50
15
100
10
200
5
500 0
10
20
30
0
10
20
30
Comp. spread s [%]
Comp. spread s [%]
90N–90S
90N–90S
45
10
20
30
0
Comp. spread s [%]
10
30
20
25
50
20
Pressure [hPa]
5
35
10
20
30
0
Comp. spread s [%]
10
20
30
Comp. spread s [%]
ozonesonde
(altitude, VMR)
10
30
20
25 20
50
15
100
10
200
5
SAGE II v7.0 (Oct 1984/Sep 1991 – Aug 2005) (altitude, number density)
100
lidar
Pressure [hPa]
Altitude [km]
0
2
40
Altitude [km]
15
2
40
15
Pressure [hPa]
100
lidar
15
Pressure [hPa]
Altitude [km]
35
Pressure [hPa]
2
40
Pressure [hPa]
6
(pressure, number density) (pressure, VMR)
500 -10
-5
0
5
Drift α [%/dec]
10
0
5
10
2σ uncertainty [%/dec]
Figure S5. SAGE II v7.0 ozone data quality in different profile representations (see legend): bias (top part), comparison spread (middle) and drift (bottom) from comparisons to NDACC/GAW/SHADOZ ozonesonde and NDACC stratospheric lidar instruments. Only latitude bands with at least 10 comparison pairs are shown. The nominal representation of SAGE II is ozone number density on fixed altitude levels. Satellite and ozonesonde profiles are converted using auxiliary information included in the ozone data files, lidar profiles are converted using ERAInterim fields. More information in Sect. 6. The SAGE II versus lidar results shown here cover 1991–2005, since we had no ERA-Interim data at our disposal to cover the earlier years. The SAGE II versus ozonesonde results cover 1984–2005.
D. Hubert et al.: Supplement to Ground-based assessment of 14 limb and occultation ozone profile data records
60N–90N
30N–60N
30N–30S
30S–60S
60S–90S
45
5
30
10 20
25
50
20
30
10
25
20
20
50
15
100
10
200
5
500 -15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
Bias b [%]
Bias b [%]
Bias b [%]
Bias b [%]
Bias b [%]
60N–90N
30N–60N
30N–30S
30S–60S
60S–90S
10
45
2
Altitude [km]
40 35
5
30
10 20
25
50
20
100
lidar
15
Altitude [km]
ozonesonde 30
10
25
20
20
50
15
100
10
200
5
500 0
10
20
30
0
10
20
30
Comp. spread s [%]
Comp. spread s [%]
90N–90S
90N–90S
45
0
10
20
30
0
Comp. spread s [%]
10
20
30
0
Comp. spread s [%]
10
20
30
Comp. spread s [%]
5
35
10
30
20
25 20
50
15
100
lidar ozonesonde
Pressure [hPa]
2
40
(altitude, VMR)
10
30
20
25 20
50
15
100
10
200
SAGE III v4.0 (Mar 2002 – Nov 2005) (altitude, number density)
Pressure [hPa]
Altitude [km]
15
(pressure, number density) (pressure, VMR)
500
5 -10
-5
0
5
Drift α [%/dec]
10
Pressure [hPa]
Altitude [km]
ozonesonde
Pressure [hPa]
100
lidar
15
0
5
10
2σ uncertainty [%/dec]
Figure S6. Same as Fig. S5, but for SAGE III v4.0 (nominal representation: ozone number density on fixed altitude levels).
Pressure [hPa]
Altitude [km]
35
Pressure [hPa]
2
40
Altitude [km]
7
D. Hubert et al.: Supplement to Ground-based assessment of 14 limb and occultation ozone profile data records
60N–90N
30N–60N
30N–30S
30S–60S
60S–90S
45
5
30
10 20
25
50
20
Altitude [km]
ozonesonde 30
10
25
20
20
50
15
100
10
200
5
500 -15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
Bias b [%]
Bias b [%]
Bias b [%]
Bias b [%]
Bias b [%]
60N–90N
30N–60N
30N–30S
30S–60S
60S–90S
10
45
Altitude [km]
35
5
30
10 20
25
50
20
100
lidar
15
Altitude [km]
ozonesonde 30
10
25
20
20
50
15
100
10
200
5
500 0
10
20
30
0
10
20
30
Comp. spread s [%]
Comp. spread s [%]
90N–90S
90N–90S
45
0
10
20
30
0
Comp. spread s [%]
10
20
30
0
Comp. spread s [%]
10
20
10
30
20
25 20
50
15
100
lidar ozonesonde
Pressure [hPa]
5
35
(altitude, VMR)
20
50
15
100
10
200
Pressure [hPa]
20
25
HALOE v19 (Oct 1991 – Nov 2005) (altitude, number density)
10
30
(pressure, number density) (pressure, VMR)
500
5 -10
-5
0
5
Drift α [%/dec]
10
0
5
30
Comp. spread s [%]
2
40
Altitude [km]
15
2
40
Altitude [km]
Pressure [hPa]
100
lidar
15
Pressure [hPa]
Altitude [km]
35
Pressure [hPa]
2
40
10
2σ uncertainty [%/dec]
Figure S7. Same as Fig. S5, but for HALOE v19 (nominal representation: ozone VMR on fixed pressure levels).
Pressure [hPa]
8
D. Hubert et al.: Supplement to Ground-based assessment of 14 limb and occultation ozone profile data records
60N–90N
30N–60N
30N–30S
30S–60S
60S–90S
45
5
30
10 20
25
50
20
30
10
25
20
20
50
15
100
10
200
5
500 -15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
Bias b [%]
Bias b [%]
Bias b [%]
Bias b [%]
Bias b [%]
60N–90N
30N–60N
30N–30S
30S–60S
60S–90S
10
45
2
Altitude [km]
40 35
5
30
10 20
25
50
20
100
lidar
15
Altitude [km]
ozonesonde 30
10
25
20
20
50
15
100
10
200
5
500 0
10
20
30
0
10
20
30
Comp. spread s [%]
Comp. spread s [%]
90N–90S
90N–90S
45
0
10
20
30
0
Comp. spread s [%]
10
20
30
0
Comp. spread s [%]
10
20
30
Comp. spread s [%]
5
35
10
30
20
25 20
50
15
100
lidar ozonesonde
Pressure [hPa]
2
40
(altitude, VMR)
10
30
20
25 20
50
15
100
10
200
UARS MLS v5 (Sep 1991 – Jun 1997) (altitude, number density)
Pressure [hPa]
Altitude [km]
15
(pressure, number density) (pressure, VMR)
500
5 -10
-5
0
5
Drift α [%/dec]
10
Pressure [hPa]
Altitude [km]
ozonesonde
Pressure [hPa]
100
lidar
15
0
5
10
2σ uncertainty [%/dec]
Figure S8. Same as Fig. S5, but for UARS MLS v5 (nominal representation: ozone VMR on fixed pressure levels).
Pressure [hPa]
Altitude [km]
35
Pressure [hPa]
2
40
Altitude [km]
9
D. Hubert et al.: Supplement to Ground-based assessment of 14 limb and occultation ozone profile data records
60N–90N
30N–60N
30N–30S
30S–60S
60S–90S
45
5
30
10 20
25
50
20
Altitude [km]
ozonesonde 30
10
25
20
20
50
15
100
10
200
5
500 -15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
Bias b [%]
Bias b [%]
Bias b [%]
Bias b [%]
Bias b [%]
60N–90N
30N–60N
30N–30S
30S–60S
60S–90S
10
45
Altitude [km]
35
5
30
10 20
25
50
20
100
lidar
15
Altitude [km]
ozonesonde 30
10
25
20
20
50
15
100
10
200
5
500 0
10
20
30
0
10
20
30
Comp. spread s [%]
Comp. spread s [%]
90N–90S
90N–90S
45
0
10
20
30
0
Comp. spread s [%]
10
20
30
0
Comp. spread s [%]
10
20
30
Comp. spread s [%]
5
35
10
30
20
25 20
50
15
100
lidar ozonesonde
Pressure [hPa]
2
40
(altitude, VMR)
10
30
20
25 20
50
15
100
10
200
Aura MLS v3.3 (Aug 2004 – May 2013) (altitude, number density)
Pressure [hPa]
Altitude [km]
15
2
40
Altitude [km]
Pressure [hPa]
100
lidar
15
(pressure, number density) (pressure, VMR)
500
5 -10
-5
0
5
Drift α [%/dec]
10
Pressure [hPa]
Altitude [km]
35
Pressure [hPa]
2
40
0
5
10
2σ uncertainty [%/dec]
Figure S9. Same as Fig. S5, but for Aura MLS v3.3 (nominal representation: ozone VMR on fixed pressure levels).
Pressure [hPa]
10
D. Hubert et al.: Supplement to Ground-based assessment of 14 limb and occultation ozone profile data records
60N–90N
30N–60N
30N–30S
30S–60S
60S–90S
45
5
30
10 20
25
50
20
30
10
25
20
20
50
15
100
10
200
5
500 -15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
Bias b [%]
Bias b [%]
Bias b [%]
Bias b [%]
Bias b [%]
60N–90N
30N–60N
30N–30S
30S–60S
60S–90S
10
45
2
Altitude [km]
40 35
5
30
10 20
25
50
20
100
lidar
15
Altitude [km]
ozonesonde 30
10
25
20
20
50
15
100
10
200
5
500 0
10
20
30
0
10
20
30
Comp. spread s [%]
Comp. spread s [%]
90N–90S
90N–90S
45
0
10
20
30
0
Comp. spread s [%]
10
20
30
0
Comp. spread s [%]
10
20
10
30
20
25 20
50
15
100
lidar ozonesonde
Pressure [hPa]
5
35
(altitude, VMR)
20
50
15
100
10
200
Pressure [hPa]
20
25
POAM II v6 (Nov 1993 – Nov 1996) (altitude, number density)
10
30
(pressure, number density) (pressure, VMR)
500
5 -10
-5
0
5
Drift α [%/dec]
10
0
5
30
Comp. spread s [%]
2
40
Altitude [km]
15
Pressure [hPa]
Altitude [km]
ozonesonde
Pressure [hPa]
100
lidar
15
10
2σ uncertainty [%/dec]
Figure S10. Same as Fig. S5, but for POAM II v6 (nominal representation: ozone number density on fixed altitude levels).
Pressure [hPa]
Altitude [km]
35
Pressure [hPa]
2
40
Altitude [km]
11
D. Hubert et al.: Supplement to Ground-based assessment of 14 limb and occultation ozone profile data records
60N–90N
30N–60N
30N–30S
30S–60S
60S–90S
45
5
30
10 20
25
50
20
Altitude [km]
ozonesonde 30
10
25
20
20
50
15
100
10
200
5
500 -15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
Bias b [%]
Bias b [%]
Bias b [%]
Bias b [%]
Bias b [%]
60N–90N
30N–60N
30N–30S
30S–60S
60S–90S
10
45
Altitude [km]
35
5
30
10 20
25
50
20
100
lidar
15
Altitude [km]
ozonesonde 30
10
25
20
20
50
15
100
10
200
5
500 0
10
20
30
0
10
20
30
Comp. spread s [%]
Comp. spread s [%]
90N–90S
90N–90S
45
0
10
20
30
0
Comp. spread s [%]
10
20
30
0
Comp. spread s [%]
10
20
30
Comp. spread s [%]
5
35
10
30
20
25 20
50
15
100
lidar ozonesonde
Pressure [hPa]
2
40
(altitude, VMR)
10
30
20
25 20
50
15
100
10
200
POAM III v4 (Apr 1998 – Dec 2005) (altitude, number density)
Pressure [hPa]
Altitude [km]
15
2
40
Altitude [km]
Pressure [hPa]
100
lidar
15
(pressure, number density) (pressure, VMR)
500
5 -10
-5
0
5
Drift α [%/dec]
10
Pressure [hPa]
Altitude [km]
35
Pressure [hPa]
2
40
0
5
10
2σ uncertainty [%/dec]
Figure S11. Same as Fig. S5, but for POAM III v4 (nominal representation: ozone number density on fixed altitude levels).
Pressure [hPa]
12
D. Hubert et al.: Supplement to Ground-based assessment of 14 limb and occultation ozone profile data records
60N–90N
30N–60N
30N–30S
30S–60S
60S–90S
45
5
30
10 20
25
50
20
30
10
25
20
20
50
15
100
10
200
5
500 -15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
Bias b [%]
Bias b [%]
Bias b [%]
Bias b [%]
Bias b [%]
60N–90N
30N–60N
30N–30S
30S–60S
60S–90S
10
45
2
Altitude [km]
40 35
5
30
10 20
25
50
20
100
lidar
15
Altitude [km]
ozonesonde 30
10
25
20
20
50
15
100
10
200
5
500 0
10
20
30
0
10
20
30
Comp. spread s [%]
Comp. spread s [%]
90N–90S
90N–90S
45
0
10
20
30
0
Comp. spread s [%]
10
20
30
0
Comp. spread s [%]
10
20
30
Comp. spread s [%]
5
35
10
30
20
25 20
50
15
100
lidar ozonesonde
Pressure [hPa]
2
40
(altitude, VMR)
10
30
20
25 20
50
15
100
10
200
OSIRIS v5.07 (Oct 2001 – May 2013) (altitude, number density)
Pressure [hPa]
Altitude [km]
15
(pressure, number density) (pressure, VMR)
500
5 -10
-5
0
5
Drift α [%/dec]
10
Pressure [hPa]
Altitude [km]
ozonesonde
Pressure [hPa]
100
lidar
15
0
5
10
2σ uncertainty [%/dec]
Figure S12. Same as Fig. S5, but for OSIRIS v5.07 (nominal representation: ozone number density on fixed altitude levels).
Pressure [hPa]
Altitude [km]
35
Pressure [hPa]
2
40
Altitude [km]
13
D. Hubert et al.: Supplement to Ground-based assessment of 14 limb and occultation ozone profile data records
60N–90N
30N–60N
30N–30S
30S–60S
60S–90S
45
5
30
10 20
25
50
20
Altitude [km]
ozonesonde 30
10
25
20
20
50
15
100
10
200
5
500 -15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
Bias b [%]
Bias b [%]
Bias b [%]
Bias b [%]
Bias b [%]
60N–90N
30N–60N
30N–30S
30S–60S
60S–90S
10
45
Altitude [km]
35
5
30
10 20
25
50
20
100
lidar
15
Altitude [km]
ozonesonde 30
10
25
20
20
50
15
100
10
200
5
500 0
10
20
30
0
10
20
30
Comp. spread s [%]
Comp. spread s [%]
90N–90S
90N–90S
45
0
10
20
30
0
Comp. spread s [%]
10
20
30
0
Comp. spread s [%]
10
20
10
30
20
25 20
50
15
100
lidar ozonesonde
Pressure [hPa]
5
35
(altitude, VMR)
20
50
15
100
10
200
Pressure [hPa]
20
25
SMR v2.1 (Jun 2001 – May 2013) (altitude, number density)
10
30
(pressure, number density) (pressure, VMR)
500
5 -10
-5
0
5
Drift α [%/dec]
10
0
5
30
Comp. spread s [%]
2
40
Altitude [km]
15
2
40
Altitude [km]
Pressure [hPa]
100
lidar
15
Pressure [hPa]
Altitude [km]
35
Pressure [hPa]
2
40
10
2σ uncertainty [%/dec]
Figure S13. Same as Fig. S5, but for SMR v2.1 (nominal representation: ozone VMR on variable altitude levels).
Pressure [hPa]
14
D. Hubert et al.: Supplement to Ground-based assessment of 14 limb and occultation ozone profile data records
60N–90N
30N–60N
30N–30S
30S–60S
60S–90S
45
5
30
10 20
25
50
20
30
10
25
20
20
50
15
100
10
200
5
500 -15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
Bias b [%]
Bias b [%]
Bias b [%]
Bias b [%]
Bias b [%]
60N–90N
30N–60N
30N–30S
30S–60S
60S–90S
10
45
2
Altitude [km]
40 35
5
30
10 20
25
50
20
100
lidar
15
Altitude [km]
ozonesonde 30
10
25
20
20
50
15
100
10
200
5
500 0
10
20
30
0
10
20
30
Comp. spread s [%]
Comp. spread s [%]
90N–90S
90N–90S
45
0
10
20
30
0
Comp. spread s [%]
10
20
30
0
Comp. spread s [%]
10
20
30
Comp. spread s [%]
5
35
10
30
20
25 20
50
15
100
lidar ozonesonde
Pressure [hPa]
2
40
(altitude, VMR)
10
30
20
25 20
50
15
100
10
200
GOMOS IPF 6.01 (Jul 2002 – Apr 2012) (altitude, number density)
Pressure [hPa]
Altitude [km]
15
(pressure, number density) (pressure, VMR)
500
5 -10
-5
0
5
Drift α [%/dec]
10
Pressure [hPa]
Altitude [km]
ozonesonde
Pressure [hPa]
100
lidar
15
0
5
10
2σ uncertainty [%/dec]
Figure S14. Same as Fig. S5, but for GOMOS IPF6.01 (nominal representation: ozone number density on variable altitude levels).
Pressure [hPa]
Altitude [km]
35
Pressure [hPa]
2
40
Altitude [km]
15
D. Hubert et al.: Supplement to Ground-based assessment of 14 limb and occultation ozone profile data records 60N–90N
30N–60N
30N–30S
30S–60S
60S–90S
45
5
30
10 20
25
50
20
Altitude [km]
ozonesonde 30
10
25
20
20
50
15
100
10
200
5
500 -15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
Bias b [%]
Bias b [%]
Bias b [%]
Bias b [%]
Bias b [%]
60N–90N
30N–60N
30N–30S
30S–60S
60S–90S
10
15
45
2
Altitude [km]
40 35
5
30
10 20
25
50
20
100
lidar
15
Altitude [km]
ozonesonde 30
10
25
20
20
50
15
100
10
200
5
500 0
10
20
30
0
10
20
30
Comp. spread s [%]
Comp. spread s [%]
90N–90S
90N–90S
45
10
20
30
0
Comp. spread s [%]
10
30
20
25 20
50
15
100
lidar ozonesonde
Pressure [hPa]
5
35
10
20
30
0
Comp. spread s [%]
10
20
30
Comp. spread s [%]
(altitude, VMR)
10
30
20
25 20
50
15
100
10
200
MIPAS ML2PP 6.0 (Jan 2005 – Apr 2012) (altitude, number density)
Pressure [hPa]
Altitude [km]
0
2
40
Altitude [km]
Pressure [hPa]
100
lidar
15
Pressure [hPa]
Altitude [km]
35
Pressure [hPa]
2
40
Pressure [hPa]
16
(pressure, number density) (pressure, VMR)
500
5 -10
-5
0
5
Drift α [%/dec]
10
0
5
10
2σ uncertainty [%/dec]
Figure S15. Same as Fig. S5, but for MIPAS ML2PP6.0 2005–2012 nominal observation data (nominal representation: ozone VMR on variable pressure levels). The large increase in the spread of tropical lidar comparisons is due to outliers in the averaging kernels. It is not seen when a triangular smoothing function is used.
D. Hubert et al.: Supplement to Ground-based assessment of 14 limb and occultation ozone profile data records
60N–90N
30N–60N
30N–30S
30S–60S
60S–90S
45
5
30
10 20
25
50
20
30
10
25
20
20
50
15
100
10
200
5
500 -15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
Bias b [%]
Bias b [%]
Bias b [%]
Bias b [%]
Bias b [%]
60N–90N
30N–60N
30N–30S
30S–60S
60S–90S
10
45
2
Altitude [km]
40 35
5
30
10 20
25
50
20
100
lidar
15
Altitude [km]
ozonesonde 30
10
25
20
20
50
15
100
10
200
5
500 0
10
20
30
0
10
20
30
Comp. spread s [%]
Comp. spread s [%]
90N–90S
90N–90S
45
0
10
20
30
0
Comp. spread s [%]
10
20
30
0
Comp. spread s [%]
10
20
30
Comp. spread s [%]
5
35
10
30
20
25 20
50
15
100
lidar ozonesonde
Pressure [hPa]
2
40
(altitude, VMR)
10
30
20
25 20
50
15
100
10
200
SCIAMACHY SGP 5.02 (Aug 2002 – Apr 2012) (altitude, number density)
Pressure [hPa]
Altitude [km]
15
Pressure [hPa]
Altitude [km]
ozonesonde
Pressure [hPa]
100
lidar
15
Pressure [hPa]
Altitude [km]
35
Pressure [hPa]
2
40
Altitude [km]
17
(pressure, number density) (pressure, VMR)
500
5 -10
-5
0
5
Drift α [%/dec]
10
0
5
10
2σ uncertainty [%/dec]
Figure S16. Same as Fig. S5, but for SCIAMACHY SGP5.02 (nominal representation: ozone number density on fixed altitude levels).
D. Hubert et al.: Supplement to Ground-based assessment of 14 limb and occultation ozone profile data records
60N–90N
30N–60N
30N–30S
30S–60S
60S–90S
45
5
30
10 20
25
50
20
Altitude [km]
ozonesonde 30
10
25
20
20
50
15
100
10
200
5
500 -15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
Bias b [%]
Bias b [%]
Bias b [%]
Bias b [%]
Bias b [%]
60N–90N
30N–60N
30N–30S
30S–60S
60S–90S
10
45
Altitude [km]
35
5
30
10 20
25
50
20
100
lidar
15
Altitude [km]
ozonesonde 30
10
25
20
20
50
15
100
10
200
5
500 0
10
20
30
0
10
20
30
Comp. spread s [%]
Comp. spread s [%]
90N–90S
90N–90S
45
0
10
20
30
0
Comp. spread s [%]
10
20
30
0
Comp. spread s [%]
10
20
30
Comp. spread s [%]
5
35
10
30
20
25 20
50
15
100
lidar ozonesonde
Pressure [hPa]
2
40
(altitude, VMR)
10
30
20
25 20
50
15
100
10
200
ACE-FTS v3.0 (Feb 2004 – Sep 2010) (altitude, number density)
Pressure [hPa]
Altitude [km]
15
2
40
Altitude [km]
Pressure [hPa]
100
lidar
15
(pressure, number density) (pressure, VMR)
500
5 -10
-5
0
5
Drift α [%/dec]
10
Pressure [hPa]
Altitude [km]
35
Pressure [hPa]
2
40
0
5
10
2σ uncertainty [%/dec]
Figure S17. Same as Fig. S5, but for ACE-FTS v3.0 (nominal representation: ozone VMR on fixed altitude levels).
Pressure [hPa]
18
D. Hubert et al.: Supplement to Ground-based assessment of 14 limb and occultation ozone profile data records 60N–90N
30N–60N
30N–30S
30S–60S
60S–90S
45
5
30
10 20
25
50
20
30
10
25
20
20
50
15
100
10
200
5
500 -15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
10
15
-15 -10 -5
0
5
Bias b [%]
Bias b [%]
Bias b [%]
Bias b [%]
Bias b [%]
60N–90N
30N–60N
30N–30S
30S–60S
60S–90S
10
45
2
Altitude [km]
40 35
5
30
10 20
25
50
20
100
lidar
15
Altitude [km]
ozonesonde 30
10
25
20
20
50
15
100
10
200
5
500 0
10
20
30
0
10
20
30
Comp. spread s [%]
Comp. spread s [%]
90N–90S
90N–90S
45
0
10
20
30
0
Comp. spread s [%]
10
20
30
0
Comp. spread s [%]
10
20
30
Comp. spread s [%]
5
35
10
30
20
25 20
50
15
100
lidar ozonesonde
Pressure [hPa]
2
40
(altitude, VMR)
10
30
20
25 20
50
15
100
10
200
MAESTRO v1.2 VIS (Feb 2004 – Sep 2010) (altitude, number density)
Pressure [hPa]
Altitude [km]
15
Pressure [hPa]
Altitude [km]
ozonesonde
Pressure [hPa]
100
lidar
15
Pressure [hPa]
Altitude [km]
35
Pressure [hPa]
2
40
Altitude [km]
19
(pressure, number density) (pressure, VMR)
500
5 -10
-5
0
5
Drift α [%/dec]
10
0
5
10
2σ uncertainty [%/dec]
Figure S18. Same as Fig. S5, but for the nominal representation of MAESTRO v1.2 (ozone VMR on variable altitude levels). Non-native representations were not verified here since the required auxiliary data are identical to those of ACE-FTS v3.0. So we expect a similar representation-dependence as in Fig. S17.
20
D. Hubert et al.: Supplement to Ground-based assessment of 14 limb and occultation ozone profile data records
45
SAGE II v7.0
90N–90S
SAGE III v4.0
HALOE v19 sunrise sunset
Altitude [km]
40 35 30 25 20 15 10 5 45
POAM II v6
POAM III v4
ACE-FTS v3.0
MAESTRO v1.2
Altitude [km]
40 35 30 25 20 15 10 5 -15 -10
-5
0
5
Bias b [%]
10
15
-15 -10
-5
0
5
Bias b [%]
10
15
-15 -10
-5
0
5
Bias b [%]
10
15
-15 -10
-5
0
5
10
15
Bias b [%]
Figure S19. Bias relative to ozonesonde (dashed) and lidar (solid) of the sunrise and sunset measurements by seven solar occultation instruments. The bias is calculated over the entire ground-based network. The analysis is performed in the native profile representation of each satellite record.
D. Hubert et al.: Supplement to Ground-based assessment of 14 limb and occultation ozone profile data records
SAGE II v7.0
HALOE v19
Aura MLS v3.3
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Altitude [km]
30 25 20 15 10 5
OSIRIS v5.07
SMR v2.1
MIPAS ML2PP6.0
21
Ozonesonde 60°N─90°N
SCIAMACHY SGP5.02
Altitude [km]
30 25 20 15 10 5 -15 -10
-5
0
5
10
15
-15 -10
Bias b [%]
-5
0
5
10
15
-15 -10
Bias b [%]
SAGE II v7.0
-5
0
5
10
15
-15 -10
HALOE v19
Aura MLS v3.3
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Altitude [km]
25 20 15 10 5
SMR v2.1
0
5
10
15
Bias b [%]
30
OSIRIS v5.07
-5
Bias b [%]
MIPAS ML2PP6.0
Ozonesonde 60°N─90°N
SCIAMACHY SGP5.02
Altitude [km]
30 25 20 15 10 5 0
10
20
30
Comp. spread s [%]
0
10
20
30
Comp. spread s [%]
0
10
20
30
Comp. spread s [%]
0
10
20
30
Comp. spread s [%]
Figure S20. Dependence on month of the bias (top part) and spread (bottom) for comparisons of seven satellite records to Arctic ozonesonde data. The analysis is performed in the native profile representation of each satellite record.
