Air Pollution Monitoring / REF : TEM/SQAR2/001
ISSUE : 1.7
DATE : 05.12.06
PAGE : 33/42
DOCUMENT TYPE: Service Implementation Document
TITLE:
Service Quality Assessment Report
Air Pollution Monitoring
Authors:
Ronald van der A, Nadege Blond, Folkert Boersma, Jean-Christopher Lambert, Pieter Valks, Andrea Weiss, Walter Di Nicolantonio, Rodolfo Guzzi, Jos van Geffen
DOCUMENT STATUS SHEET
1.0 / 15.10.03 / First Version, including FRESCO validation (chapter 2)
1.1 / 04.05.04 / Validation of tropical tropospheric ozone added
1.2 / 15.11.04 / Validation of tropospheric NO2 and global tropospheric O3 added
1.3 / 26.11.04 / Validation of regional NO2 added
1.4 / 10.02.05 / Aerosol (GOME/SCIAMACHY) added
1.5 / 01.03.05 / Aerosol (ATSR-2/AATSR) added
1.6 / 13.09.06 / User reponse chapters for regional NO2 and tropospheric ozone added
1.7 / 05.12.06 / User feedback on SO2 added
TABLE OF CONTENTS
1. Introduction 5
1.1 Purpose and scope 5
1.2 Document overview 5
1.3 Definitions, acronyms and abbreviations 5
1.4 Applicable Documents 6
2. FRESCO 7
2.1 Validation Approach 7
2.1.1 MODIS 7
2.1.2 GOME 7
2.1.3 References 7
2.2 Intercomparison with MODIS cloud fraction 8
2.3 Intercomparison with GOME cloud fraction and cloud-top pressure 9
3. GloBAL NO2 14
3.1 Validation approach and results 14
3.1.1 Preliminary results of geophysical validation 14
3.1.2 Model comparison 15
3.1.3 References 17
3.2 User response 17
4. Regional NO2 19
4.1 Validation approach 19
4.1.1 Case study comparison with ground-based in-situ measurements 19
4.1.2 Intercomparison with ground-based DOAS 20
4.2 Results 22
4.2.1 Intercomparison with ground-based DOAS 22
4.3 User response 23
4.3.1 ARPA Emilia-Romagna 23
4.3.2 BUWAL and NABEL/Empa 24
5. Tropical Ozone 25
5.1 Validation approach 25
5.2 Results 25
5.3 References 28
5.4 User response 28
6. Global Tropospheric Ozone 29
6.1 Quality estimate 29
6.2 User response 31
7. SO2 32
7.1 Validation approach 32
7.2 Results 32
7.3 User response 33
8. Aerosol from ATSR-2 and AATSR 34
8.1 Validation approach 34
8.2 Results 35
8.3 References 37
8.4 User response 37
9. Aerosol from GOME and Sciamachy 38
9.1 Validation approach 38
9.2 Results 39
9.2.1 GOME and AERONET 39
9.2.2 Comparison between GOME and other satellites 41
9.2.3 SCIAMACHY 45
9.3 User response 48
1. Introduction
1.1 Purpose and scope
The Data User Programme (DUP) is an optional programme of ESA which aims at supporting Industry, Research Laboratories, User Communities as well as European and National Decision Makers to bridge the gap that exists between research at the level of pilot projects and the operational and sustainable provision of Earth Observation products at information level.
TEMIS is a project (started September 2001) in response to an Invitation To Tender from ESA in the context of ESA's Data User Programme. The aim of the project is the delivery of tropospheric trace gas concentrations, and aerosol and UV products, derived from observations of the nadir-viewing satellite instruments GOME and SCIAMACHY.
This document contains the validation approach and results of the products for TEMIS. The current version is part of the final deliverables of the implementation phase of TEMIS.
The data products, images and reading routines can be found on the web-site www.temis.nl. A description of the products and their retrieval is presented in the Service Report.
1.2 Document overview
Section 1 contains the introduction and applicable documents. Sections 2 to 9 give detailed descriptions of the product validation.
1.3 Definitions, acronyms and abbreviations
AMF / Air-Mass FactorAOD / Aerosol Optical Depth
ATSR / Along Track Scanning Radiometer
AATSR / Advanced Along Track Scanning Radiometer
ASCAR / Algorithm Survey and Critical Analysis Report
BIRA-IASB / Belgian Institute for Space Aeronomy
BrO / Bromine Oxide
CCD / Convective Cloud Differential
CH2O / Formaldehyde
DLR / German Aerospace Centre
DOAS / Differential Optical Absorption Spectrometry
DUP / Data User Programme
ENVISAT / Environmental Satellite
ERS / European Remote Sensing Satellite
ESA / European Space Agency
ESRIN / European Space Research Institute
EUMETSAT / European Organisation for the Exploitation of Meteorological Satellites
FRESCO / Fast Retrieval Scheme for Cloud Observables
GOFAP / GOME Ozone Fast Delivery and value-Added Products
GOME / Global Ozone Monitoring Instrument
ISAC / Institute of Atmospheric and Climate Sciences ( formerly ISAO )
ISCCP / International Satellite Cloud Climatology Project
KNMI / Royal Netherlands Meteorological Institute
LIDORT / Linearized Discrete Ordinate RTM
LUT / Look-Up Table
METEOSAT / Meteorological Satellite
NDSC / Network for the Detection of Stratospheric Change
NO2 / Nitrogen Dioxide
NOx / Nitrogen Oxides (NO+NO2)
NOAA / National Oceanic and Atmospheric Administration
NRT / Near-Real Time
PSD / Product Specification Document
RIVM / National Institute of Public Health and the Environment
RTM / Radiative Transfer Model
SCIAMACHY / SCanning Imaging Absorption spectroMeter for Atmospheric CartograpHY
SDD / Service Definition Document
SDP / Service Development Plan
SO2 / Sulphur Dioxide
TBC / To Be Confirmed
TBD / To Be Defined
TEMIS / Tropospheric Emission Monitoring Internet Service
TOA / Top Of Atmosphere
TOMS / Total Ozone Mapping Spectrometer
USD / User Specification Document
URD / User Requirements Document
1.4 Applicable Documents
AD-1 / Data User Programme II period 1st call For Proposal ref:EEM-AEP/DUP/CFP2001AD-2 / User Specifcation Document, v1.4, TEM/USD/005, May 2002
AD-3 / User Requirement Document, v2.0, TEM/URD/006, October 2002
AD-4 / Algorithm Survey and Critical Analysis Report, v1.2, TEM/ASCAR/003, May 2002
2. FRESCO
2.1 Validation Approach
2.1.1 MODIS
MODIS (Moderate Resolution Imaging Spectroradiometer) is an instrument aboard the TERRA (EOS AM) and AQUA (EOS PM) satellites. TERRA is flying in sun-synchronous polar orbit with an descending node at the equator of 10.30 AM local time and AQUA has an ascending node with a local time of 13.30 PM. The MODIS instrument is viewing the entire Earth's surface every 1 to 2 days, acquiring data in 36 spectral bands ranging in wavelength from 0.4 µm to 14.4 µm. Since the TERRA orbit shows the most similarity with the SCIAMACHY orbit, only data from TERRA is considered.
For our intercomparison we use the global, level-3 MODIS product, the so-called MOD08_D3 files. These contain a day of 1 x 1 degree grid average values of atmospheric parameters related to atmospheric aerosol particle properties, total ozone burden, atmospheric water vapor, cloud optical and physical properties, and atmospheric stability indices. From these files we use only the observations on the day-side of an orbit.
2.1.2 GOME
The FRESCO method was originally developed for near-real-time ozone column retrieval from GOME. The SCIAMACHY FRESCO algorithm has been based on this algorithm.
The FRESCO method for GOME is validated using ATSR-2 data [Koelemeijer et al., 2001], and a comparison is made with cloud top pressures and effective cloud fractions of ISCCP on a monthly average basis [Koelemeijer et al., 2002]. Recent validation of several cloud retrieval methods [Grzegorski et al., 2003] presented on the EGS in 2003 showed that FRESCO performs very well except for the Sahara region.
GOME and SCIAMACHY are very similar instruments, which make them ideal for validation purposes. Also the orbit is the same with a time differences of half an hour. The GOME data is gridded to 1 by 1 degree before comparison.
2.1.3 References
§ Koelemeijer, R. B. A., P. Stammes, J. W. Hovenier, and J. F. de Haan, A fast method for retrieval of cloud parameters using oxygen A band measurements from GOME, J. Geophys. Res., 106, 3475-3490, 2001.
§ Koelemeijer, R. B. A., P. Stammes, J. W. Hovenier, and J. F. de Haan, Global distributions of effective cloud fraction and cloud top pressure derived from oxygen A band spectra measured by the Global Ozone Monitoring Experiment: comparison to ISCCP data, J. Geophys. Res., 107, 2002.
§ Grzegorski, M. et al., A new cloud algorithm for GOME: Heidelberg Iterative Cloud Retrieval Utilities (HICRU), presented at the EGS, Nice 2003.
2.2 Intercomparison with MODIS cloud fraction
An intercomparison has been made between the cloud fraction and top-pressure of FRESCO-SCIAMACHY and MODIS for the day of April 14, 2003. Both data has been bin on 1 by 1 degree grid cells before comparison. Grid cells with 5 or less observations of SCIAMACHY are rejected. The comparison for the cloud-top pressure has been done only if the cloud fraction was higher than 10 %.
The different resolution of the original observation of MODIS and SCIAMACHY result in a high variability of cloud fraction values above land. Therefore, a land-sea mask is used to perform the comparison only for pixels above sea surface. This avoids also difficult snow-covered regions and the known FRESCO problem above the Sahara region.
The results for cloud fraction are plotted in Figure 2.A and for cloud-top pressure in Figure 2.B
Figure 2.A Cloud fraction of MODIS against the cloud fraction of SCIAMACHY for the grid cells of a global field on April 14, 2003.
Figure 2.B Cloud fraction of MODIS against the cloud fraction of SCIAMACHY for the grid cells of a global field on April 14, 2003.
It is obvious that especially the cloud fraction is in complete disagreement with each other. The MODIS cloud fraction is often 100 % where SCIAMACHY sees much lower values. This can be explained by the fact that we are dealing with very different instruments. In general, in the infrared used by MODIS more clouds, especially cirrus, are observed than in the visible spectrum used by FRESCO.
In addition to the instrumental differences an error has been reported in the cirrus detection in the SWIR channel of TERRA: “most of the land areas are flagged as 100% cirrus cover”.
From this comparison it can only be concluded that MODIS is not suitable for validation of the FRESCO algorithm.
2.3 Intercomparison with GOME cloud fraction and cloud-top pressure
An intercomparison has been made between the cloud fraction and cloud-top height of FRESCO for SCIAMACHY and GOME for 1, 2 and 3 March 2003. Both data sets have been binned on 1 by 1 degree grid cells before comparison. Grid cells with 5 or less observations of SCIAMACHY are rejected. The comparison for the cloud-top height has been done only if the cloud fraction was higher than 10 %.
The different resolution of the original observation of GOME and SCIAMACHY result in a high variability of cloud fraction values above land. Therefore, a land-sea mask is used to perform the comparison only for pixels above sea surface.
Since there are several software versions of FRESCO in use for GOME and SCIAMACHY, these versions are compared with each other. From this it was concluded that there were only large differences between the FRESCO used on SCIAMACHY data and FRESCO used on GOME data.
Figure 2.C FRESCO cloud fraction of GOME against the cloud fraction of SCIAMACHY for the grid cells of a global field on March 1, 2003.
The comparison between FRESCO data from GOME and SCIAMACHY has been showed for the cloud fraction in Figure 2.C and for the cloud-top height in Figure 2.E. A linear dependence (red line) has been fitted through the data. The fitted linear dependence did not change much as function of latitude, cloud fraction or surface type. Considering the cloud fraction calculation in FRESCO Figure 2 C suggests a fixed radiometric mismatch in the SCIAMACHY level 1 data of about 20 %. After testing several radiometric correction factors, it turned out that using a factor of 1.25 gave the best results. An example of the cloud fraction from corrected SCIAMACHY data versus cloud fraction from is given in Figure 2.D
Figure 2.D FRESCO cloud fraction of GOME against the cloud fraction of corrected SCIAMACHY spectra for the grid cells of a global field on March 1, 2003.
Figure 2.E FRESCO cloud-top height of GOME against the cloud fraction of SCIAMACHY for the grid cells of a global field on March 1, 2003.
The cloud-top height calculation is less sensitive for the absolute radiometric calculation but depends more on the spectral signature. Therefore, the radiometric error on the spectrum can strongly affect the fit of the cloud-top height. The radiometric error as given in the level 1c data is shown in Figure 2F, where you can see that the this error for one wavelength has a very large dynamic range and shows unrealistic patterns. The error is also given in the unit BU, which makes it already useless for the retrieval. The error has been checked for the wavelengths 327, 329 and 334 nm.
Since the radiometric error given in the level 1 data is not very helpful, initially an error of 0.5 % has been used. After testing several fractional error values, it was found that a 5 % error gave the most realistic results for the cloud-top height. The results for the cloud-top height from corrected SCIAMACHY spectra versus GOME cloud-top height is shown in Figure 2G.
Figure 2F The radiance error at a wavelength of 327 nm as reported in the level 1c data of March 2, 2003.
Figure 2.G FRESCO cloud-top height of GOME against the cloud fraction of corrected SCIAMACHY spectra for the grid cells of a global field on March 1, 2003.
3. GloBAL NO2
3.1 Validation approach and results
3.1.1 Preliminary results of geophysical validation
A few SCIAMACHY NO2 products, where among the TEMIS NO2 product, have been validated by Lambert et al [2004]. The preliminary results of this validation have been presented at the ACVE meeting in 2004. A citation of Lambert’s results regarding the TEMIS NO2 product follows.
Figure 3.1 – Meridian variation of the mean absolute difference between near real time total nitrogen dioxide in July/November 2002 as reported by NDSC ground-based spectrometers and by the SCIAMACHY NO2 slant columns retrieved from raw level-1 data by the research processor developed at BIRA-IASB, converted into vertical columns by KNMI. (source: Lambert et al.)
“The pole-to-pole comparison between NDSC NO2 data and the SCIAMACHY data set generated at KNMI/BIRA is depicted in Figure 3.1. The main difference is observed at polluted sites of the Northern middle latitudes where the discrepancy between SCIAMACHY and NDSC data reaches values as high as 3.5 1015 molec.cm-2. According to modelling results and (unfortunately non correlated) airborne measurements, such high values would be better estimates of the tropospheric NO2 column. Such a difference between KNMI/BIRA retrievals and retrievals of other institutes is supposed to come mostly from the use of tropospheric AMF. GOME and SCIAMACHY total NO2 retrievals using a pure stratospheric AMF are known to underestimate the vertical column in the presence of tropospheric NO2. Usually limited to 10% to 30%, this underestimation can reach a factor of two under extreme conditions. Another likely source of discrepancy, explaining the difference in behaviour at the stratospheric reference stations of the southern middle latitudes, comes from the use of a correction factor to account for the temperature dependence of the NO2 absorption cross-sections. DAK RTM-computed AMFs approach geometrical AMF values in the clean Southern mid-latitude case. However, the total AMF used in the retrieval is the product of this quasi-geometrical AMF by the temperature correction based on ECMWF analyses. Studies indicate that stratospheric temperatures are typically about 15K lower than the temperature of the cross section used in the DOAS fit (243K, SCIA PFM). The typical NO2 correction applied for such a temperature difference is about -5%, which corresponds to an increase in the reported AMF of 5%. Further quantitative interpretation of the results depicted in Figure 3.1 should take into account also how the temperature effect is taken into account in the retrieval of ground-based NDSC data.”