Surface UV Irradiance from OMI on EOS-Aura
Aapo Tanskanen(1), Anu Määttä(1), Nickolay Krotkov(2), Jay Herman(3), Jussi Kaurola(1), Tapani Koskela(1),
Alex Karpetchko(4), Vitali Fioletov (5), Germar Bernhard (6)
(1) Finnish Meteorological Institute, Helsinki, Finland
(2) GEST Center, University of Maryland, Baltimore, USA
(3) NASA Goddard Space Flight Center, Greenbelt, Maryland, USA
(4) FMI’s Arctic Research Centre, Sodankylä, Finland
(5) MSC/Environment Canada, Ontario, Canada
(6) Biospherical Instruments, San Diego, USA
ABSTRACT
The Ozone Monitoring Instrument (OMI) onboard the NASA EOS Aura spacecraft is a nadir viewing spectrometer that
measures solar reflected and backscattered light in a selected range of the ultraviolet and visible spectrum. The
instrument has a 2600 km wide viewing swath and it is capable of daily, global contiguous mapping. We have
developed and implemented a surface ultraviolet irradiance algorithm for OMI that produces noontime surface spectral
UV irradiance estimates at four wavelengths (305, 310, 324, 380 nm). Additionally, noontime erythemal dose rate and
the erythemal daily dose are estimated. The paper includes a brief overview of the OMI surface UV algorithm, current
processing status, recent validation results and plans for further development of the surface UV algorithm.
1
INTRODUCTION
OMI is a nadir-viewing spectrometer designed to monitor ozone and other atmospheric species [1]. It is a contribution
of the Netherlands's Agency for Aerospace Programs (NIVR) in collaboration with the Finnish Meteorological Institute
(FMI) to the EOS Aura mission of NASA. OMI contains two spectrometers that together cover the wavelength range
from 270 to 500 nm. The sunsynchronous polar orbit of the EOS Aura satellite and the large width of the OMI's
viewing swath provide global daily coverage of the sunlit portion of the atmosphere. OMI provides continuation to the
TOMS record for total ozone measurements. FMI’s Satellite Data Centre in Sodankylä takes care of the processing,
archiving and dissemination of the OMI surface UV product (OMUVB). The input data for the OMUVB processing is
level 2 OMI total ozone data (OMTO3) from NASA/DAAC. The OMUVB products are offline data that are primarily
intended for long-term monitoring of the surface UV irradiance. We describe the current processing and validation
status of the OMI surface UV data and discuss the plans for the future algorithm development.
2
OMI SURFACE UV IRRADIANCE ALGORITHM, PROCESSING SYSTEM AND DATA PRODUCT
2.1
Overview on the Current Operational OMI Surface UV Irradiance Algorithm
Surface UV irradiance is estimated using a radiative transfer model whose input parameters are derived from the OMI
measurements. The algorithm is similar to the TOMS UV algorithm developed by NASA [2, 3, 4, 5]. It first estimates
the clear-sky surface irradiance using measured total ozone and surface albedo climatology [6]. The clear-sky value is
then adjusted by a transmittance factor that accounts for the attenuation of UV radiation by clouds and aerosols. The
attenuation factor is derived from the ratio of measured backscatter radiances and solar irradiances at 360 nm assuming
that clouds and aerosols are non-absorbing at this wavelength. This leads to an overestimation of surface UV irradiance
when UV-absorbing aerosols such as smoke or desert dust are present. Future version of the algorithm may correct for
absorbing aerosols.
2.2
OMUVB Processing System and the Processing Status
Production of the OMUVB data is performed using the Sodankylä OMI UV Processing System (SOUPS). SOUPS
ingests Level 2 OMTO3 total ozone product and generates Level 2 OMUVB surface UV irradiance product, both of
which are in HDF5-EOS format. SOUPS communicates with the Sodankylä satellite data archive as an Oracle client.
The processing is carried out in batches of a few thousand granules, which are initiated manually by the operator. There
are additional postprocessing tools for archiving and gridding of data as well as for generation of overpass data.
Currently, some 7000 granules out of the total 8000 ECS Collection 2 OMTO3 granules available have been processed.
There are plans to reprocess all the OMI data as a completely new version of the OMTO3 data product becomes
available. We prefer not to process the most recent OMTO3 data because occasionally fresh granules are reprocessed
shortly after the initial processing. The OMUVB granule is written as an HDF-EOS5 swath file. Each product granule
corresponds to a single OMI orbit containing data for some hundred thousand observations over the sun-lit portion of
one Aura orbit. The primary contents of the OMUVB granules are local solar noon irradiances at 305, 310, 324, and
380 nm, as well as erythemally weighted irradiance. Additionally, the erythemally weighted daily surface UV dose is
included. The data are ordered in time. The information provided in these files include latitude, longitude, solar zenith
angle, and a large number of ancillary parameters that provide information to assess data quality. The nominal size of a
single product granule is 8.5 Mb. Thus, some 120 Mb of OMUVB data is produced per day.
2.3
Archiving and Dissemination of the OMUVB Data
The OMUVB data are archived locally in Sodankylä and are submitted to the Aura Validation Data Center. The satellite
data archive in Sodankylä has been developed within the EUMETSAT’s O3MSAF project. The archive is an Oracle
based relational database with web based user interface for maintanance, search, online plotting and dissemination of
data. The OMI surface UV data is used for testing of the archive, the operator and user interfaces as well as the data
dissemination process. The experience gained with the OMUVB data contributes to the development of the O3MSAF
operational services based on the data from the EUMETSAT's MetOp series of satellites.
2.4
Product Examples
In Figure 1 is shown the cloud-corrected UV Index at local solar noon on October 15, 2005. The figure shows elevation
of the UV Index over the Southern Atlantic Ocean due to ozone depletion. The figure is based on 1 by 1 degree gridded
OMUVB data and does not give full credit to the native spatial resolution of OMI. However, it's still evident that the
instrument provides much more detailed information on the atmosphere than its predecessors. In Figure 2 is shown the
erythemal daily dose in summer 2005 in Jokioinen (60.8N, 23.5E) together with daily doses derived from the groundbased Brewer measurements. Figure shows good agreement between the satellite retrieved and ground-based
measurement data.
Figure 1. OMI data based cloud-corrected UV Index at local solar noon on October 15, 2005.
Figure 2. Erythemal daily dose [kJ/m2] in summer 2005 in Jokioinen according to OMUVB and Brewer #107.
2.5
Initial Data Quality Assessment
The OMI surface UV algorithm is based on the radiative transfer model that assumes that clouds are plane parallel and
homogeneous, i.e., it doesn’t account for 3D effects of clouds. This error is the principal source of noise in comparing
satellite measurements with ground-based instruments. The OMI measurements represent the mean surface UV over a
wider region rather than at a point. OMI measurements are made once a day around 1:45 p.m. local time. No correction
is made for the change in cloudiness, ozone and aerosols between local noon and satellite overpass time, or for their
diurnal variability. Previous validation studies with TOMS data suggest that OMI UV irradiance estimates are on the
average 0-30% larger than the ground-based reference data [7, 8, 9, 10, 11]. The systematic bias can be attributed to
absorbing aerosols from natural and anthropogenic sources. Since the soot content of the urban aerosols tend to be
highly localized, these errors presumably are also localized and do not necessarily represent the error in regional
estimate of surface UV made by OMI. Snow and ice further complicate estimation of the surface UV since clouds
cannot be distinguished from them. Therefore, in regions with temporary snow or ice or highly heterogeneous surface
albedo the OMI UV irradiance estimates have much higher uncertainty.
3
VALIDATION OF THE OMI SURFACE UV DATA
3.1
Comparison with the Earth Probe TOMS Surface UV Data
The goal of the validation is to establish the OMUVB product uncertainty in different atmospheric conditions and to
gain insight for further development of the OMI surface UV irradiance algorithm. The first impression of the quality of
the OMUVB data was obtained using coexistent surface UV data from the Earth Probe satellite [12]. The noontime
erythemal surface irradiances derived from the OMI measurements were compared with those provided by NASA from
the Earth Probe TOMS measurements. Direct comparison of the global gridded data for July 15, 2005 revealed that the
surface UV irradiances from OMI were in general lower over land and higher over sea than those derived from the
Earth Probe TOMS. Moreover, the different overpass times of the Earth Probe and Aura satellites explained some of the
small scale differences, and positive bias was found in regions where the effect of absorbing aerosols was expected.
Secondly, we compared zonally and monthly averaged data. The comparison implied that the erythemal irradiances at
local solar noon from OMI were of the order of 10% higher than those from the Earth Probe TOMS.
3.2
Validation of the OMI UV Data with Ground-based Measurement Data
The ultimate reference data for validation are the ground-based spectral UV irradiance measurements. Validation relies
on spectral UV monitoring sites with high-level instrument quality assurance and control. The validation sites shall
represent various latitudes, climatic conditions, land cover types and altitudes. Some validation sites provide concurrent
measurements of aerosol optical depth and single scattering albedo that can be used to validate surface UV algorithm to
account for aerosols. The erythemal daily doses derived from the OMI measurements were compared with those
calculated from the ground-based spectral UV measurements [12]. The ground-based reference data was obtained from
six measurement sites and from seven spectroradiometers listed in Table 1.
Table 1. Ground-based reference data used for validation of the erythemal daily doses from OMI.
Instrument
Site
Geolocation
Validation Period
Brewer Mk-III #107
Jokioinen
60.81°N 23.50°E
6.9.2004 - 1.9.2005
Brewer Mk-II #037
Sodankylä
67.37°N 26.63°E
17.8.2004 - 30.9.2005
aBrewer
Mk-II #014
Toronto
43.78°N 79.47°W
6.9.2004 - 28.6.2005
aBrewer
Mk-III #145
Toronto
43.78°N 79.47°W
9.9.2004 - 9.6.2005
bSUV-100
San Diego
32.77°N 117.20°W
8.10.2004 - 1.10.2005
bSUV-100
Ushuaia
54.82°S 68.32°W
6.9.2004 - 1.10.2005
bSUV-100
Barrow
71.32°N 156.68°W
6.9.2004 - 1.10.2005
a
preliminary data: calibration errors of 3-5% are possible
b
preliminary Version 0 data: cosine corrections have not been made; Version 2 data will be higher by 4-10%
In Figures 3a and 3b are shown the comparisons of the erythemal daily doses derived from OMI and ground-based
measurements in Sodankylä and Toronto. The scatterograms show good agreement between the satellite retrieved UV
and the ground-based reference data. A library of validation programs was developed for estimation of the data quality.
This library is an essential tool for further development of the OMI surface UV algorithm. The validation includes
calculation of statistical figures of merit to quantify how well the satellite retrieved surface UV data agrees with the
reference data. These include absolute and relative Bias and Root-Mean-Square Error, which are defined as
Bias =
1 n
∑ (si −gi )
n i =1
% Bias =
1 n ⎛ si − g i ⎞
⎟ ∗100%
∑⎜
n i =1 ⎜⎝ g i ⎟⎠
RMS =
1 n
∑ (si − gi ) 2
n i =1
2
% RMS =
1 n ⎛ si − g i ⎞
⎟ ∗100%
∑⎜
n i =1 ⎜⎝ g i ⎟⎠
(1)
where n is the number of days with both data, si is the satellite retrieved daily dose, and gi is the ground-based reference
data. Additionally we calculated the correlation coefficient r between the two sets of data. The statistical validation
results are summarized in Table 2.
Figures 3a and 3b. Comparison of the erythemal daily doses [kJ/m2] derived from OMI and ground-based data in
Sodankylä and Toronto.
Table 2. OMUVB validation statistics for daily doses [J/m2]
Validation instrument
n
Bias
RMS error
r
Jokioinen Brewer Mk-III #107
292
83 (5.2%)
263 (34%)
0.98
Sodankylä Brewer Mk-II #037
175
50 (7.6%)
234 (22%)
0.97
Toronto Brewer Mk-II #014
262
1 (-3.7%)
336 (24%)
0.98
Toronto Brewer Mk-III #145
232
-92 (-9.4%)
322 (25%)
0.97
San Diego SUV-100
293
768 (31%)
974 (41%)
0.95
Ushuaia SUV-100
339
89 (2.6%)
379 (25%)
0.97
Barrow SUV-100
203
221 (19%)
492 (36%)
0.94
The validation results imply improvements in the accuracy of the satellite UV data thanks to the improved spatial
resolution of the OMI instrument and advances of the surface UV algorithm. However, there are still several sources of
uncertainty that affect the quality of the satellite estimates of the daily erythemal dose. In sites with temporary snow or
ice the agreement between the satellite and ground based data is better in summer than in winter. Because the current
surface UV algorithm does not account for absorbing aerosols, there is a systematic positive bias in satellite estimates
for sites affected by aerosols from urban or natural sources. Furthermore, the satellite estimate of the daily dose is based
on a single observation of the cloud conditions, which causes extra variance, and in some cases even a systematic bias
that depends on the satellite overpass time. Monthly averaging resulted in better agreement between the satellite and
ground based data. We plan to continue the validation effort by including more validation sites and by extending the
uncertainty analysis. In addition to validation of the daily dose data, there is a need to validate the spectral irradiances.
4
CONCLUSIONS
Global surface UV time series are continued with the OMI measurements applying an algorithm similar to the original
TOMS UV algorithm. The validation results imply improved accuracy of the surface UV estimates, and the OMUVB
data will be released publicly in spring 2006. The positive bias of the surface UV estimates produced by the operational
OMI UV algorithm can be corrected provided that the aerosol absorption optical thickness for the specific UV
wavelength is known from ground-based measurements. An alternative is to use climatological values for aerosol
absorption optical thickness. There are also plans to investigate the potential use of the OMI aerosol product data to
correct for absorbing aerosols. Furthermore, because UV radiation is involved in several photochemical processes and
can affect human health, ecosystems and materials, it has become apparent that there is a need for a wider variety of
surface UV data. Because of large number of various action spectra, it would ideal to have the full spectrum with
sufficient spectral resolution. However, due to computing and data storage limitations time-series of full spectral data
can only be produced for some selected sites. Additionally the OMI surface UV data could be extended by including
irradiance on tilted surfaces, actinic flux, and underwater irradiance.
5
REFERENCES
1. P. F. Levelt, E. Hilsenrath, G.W. Leppelmeier, G.H.J. van den Oord, P.K. Bhartia, J. Tamminen, J.F. de Haan, J.P.
Veefkind., The Ozone Monitoring Instrument, IEEE Trans Geo. Rem. Sens. Aura special issue, 2005.
2. T. F. Eck, P. K. Bhartia, J. B. Kerr, Satellite estimation of spectral UVB irradiance using TOMS derived ozone and
reflectivity, Geophys. Res. Lett., 22, 611-614, 1995.
3. N. A. Krotkov, P. K. Bhartia, J. R. Herman, V. Fioletov, and J. Kerr, Satellite estimation of spectral surface UV
irradiance in the presence of tropospheric aerosols 1: Cloud-free case, J. Geophys. Res., 103, 8779-8793, 1998.
4. N. A. Krotkov, P. K. Bhartia, J. R. Herman, Z. Ahmad, V. Fioletov, Satellite estimation of spectral surface UV
irradiance 2: Effect of horizontally homogeneous clouds and snow, J. Geophys. Res., 106, 11743-11759, 2001.
5. A. Tanskanen, N. A. Krotkov, J.R. Herman, A. Arola, Surface Ultraviolet Irradiance from OMI, IEEE Trans. Geo.
Rem. Sens. Aura Special Issue, 2005.
6. A. Tanskanen, Lambertian surface albedo climatology at 360 nm from TOMS data using moving time-window
technique, In: Proceedings of the XX Quadrennial Ozone Symposium, 1-8 June 2004, Kos, Greece, pp. 1159-1160.
7. V. E. Fioletov, M. G. Kimlin, N. A. Krotkov, L. J. B. McArthur, J. B. Kerr, D. I. Wardle, J. R. Herman, R. Meltzer,
T. W. Mathews, and J. Kaurola, UV index climatology over North America from ground-based and satellite
estimates, J. Geophys. Res., 109, 2308, 2004.
8. S. Kalliskota, J. Kaurola, P. Taalas, J. R. Herman, E. Celarier, N. Krotkov, Comparison of the daily UV doses
estimated from Nimbus7/TOMS measurements and ground-based spectroradiometric data, J. Geophys. Res., 105,
5059-5067, 2000.
9. N. Ye. Chubarova, A. Yu. Yurova, N. A. Krotkov, J. R. Herman, and P. K. Bhartia, Comparison between ground
measurements of UV irradiance 290 to 380 nm and TOMS UV estimates over Moscow for 1979-2000, Opt. Eng.
41(12), 3070-3081, 2002.
10. R. McKenzie, G. Seckmeyer, A. Bais, J. Kerr, S. Madronich, Satellite retrievals of erythemal UV dose compared
with ground-based measurements at northern and southern midlatitudes, J. Geophys. Res., 106, 24051-24062, 2001.
11. D. Meloni, A. di Sarra, J. R. Herman, F. Monteleone, S. Piacentimo, Comparison of ground-based and Total Ozone
Mapping Spectrometer erythemal UV doses at the island of Lampedusa in the period 1998–2003: Role of
tropospheric aerosols, J. Geophys. Res., 110, 2005.
12. A. Tanskanen, A. Määttä, J. Kaurola, N. Krotkov, A. Karpetchko, G. Bernhard, V. Fioletov, Validation of the OMI
surface UV data, In: Proceedings of the AGU fall meeting, 5-9 December, San Fransisco, 2005.
Acknowledgements
We are grateful to the OMI and Aura science teams for their efforts for the quality of the OMI data. The support of the
EUMETSAT's O3MSAF project greatly enhances archiving, management and dissemination of the OMUVB data. UV
measurement data for Toronto were provided by MSC/Environment Canada. UV measurement data from San Diego,
Ushuaia and Barrow was provided by the NSF UV Monitoring Network, operated by Biospherical Instruments Inc.
under a contract from the United States National Science Foundation's Office of Polar Programs via Raytheon Polar
Services Company.