NEW IG2 SEASONAL CLIMATOLOGIES FOR MIPAS

Size: px

Start display at page:

Download "NEW IG2 SEASONAL CLIMATOLOGIES FOR MIPAS"

Marvin Washington
6 years ago
Views:

NEW IG2 SEASONAL CLIMATOLOGIES FOR MIPAS J.J. Remedios (1), R.J. Leigh (1), H. Sembhi (1), A. M. Waterfall (2) (1) EOS, Space Research Centre, University of Leicester, U.K., Email:j.

1 NEW IG2 SEASONAL CLIMATOLOGIES FOR MIPAS J.J. Remedios (1), R.J. Leigh (1), H. Sembhi (1), A. M. Waterfall (2) (1) EOS, Space Research Centre, University of Leicester, U.K., (2) Now at Rutherford Appleton Laboratory, Chilton, Didcot, Oxfordshire, U.K., ABSTRACT Reliable reference profiles and variability estimates are a necessity for a variety of processes relating to ENVISAT including the development of key aspects of the MIPAS/ENVISAT operational processor. In relation to MIPAS, these profiles are employed as inputs to microwindow selection and occupation matrices (MW/OM). They are also used to calculate forward model tabulations and LUT s which account for variations in absorber amounts and profile shapes. Progress towards improved seasonal reference IG2 climatologies is described which particularly addresses the requirement to provide initial guess and contaminant profiles in the MIPAS level 2 processing. Enhancements from previous versions include the derivation of N 2 O profiles from CH 4 tracer correlations, calculation of annual CO 2 profiles using data from the Globalview initiative, and the inclusion of pressure/temperature data from CIRA. Comparisons of v3.1 and v4.0 IG2 climatologies are presented alongside comparisons with MIPAS data for the primary operational MIPAS species, including investigations into the variability in retrieved MIPAS data in comparison to IG2 climatology sigmas. 2. THE MIPAS INSTRUMENT The MIPAS instrument [1,2] is a Fourier transform spectrometer (FTS) providing limb sounding spectra of atmospheric infra-red emission between 685 cm -1 (14.60 µm) and 2410 cm -1 (4.15 µm). In full spectral resolution mode, the subject of this paper, observations were obtained at a spectral resolution of cm -1 (unapodized); such data were acquired between launch and March The data are recorded in five spectral channels but near-continuous coverage is available in the spectral domain albeit with narrow gaps between channels. Vertical profiles are scanned from 6 km to 68 km with a vertical spacing of 3 km in the UT and lower stratosphere commensurate with the vertical resolution of the instrument. Each profile requires approximately 75 s corresponding to an along-track distance of 500 km between vertical scans. 1. INTRODUCTION Reference atmospheres are a means of categorizing available knowledge of atmospheric states within some framework for average behaviour. Zonal mean climatologies are a particularly accessible means of delvering such data sets whilst retaining a simple representation. In the case of the operational processor for the Michelson Interferometer for Passive Atmospheric Sounding (MIPAS), this need for a simple climatology was driven by constraints imposed by the desire to constrain auxiliary data changes. This also has the beneficial impact of retaining a well-defined set of profiles for initial guess and contaminant profiles. The resulting IG2 climatology may also be used to assess the behaviour of average MIPAS data with respect to seasonal, latitudinal and altitude dependent effects. Since the IG2 climatology is computed as both mean profiles and sigmas, the intercomparison of IG2 data and observed MIPAS data is revealing about both the climatology and the satellite data. Figure 1 Comparison of MIPAS spectra for clouds of different optical depth in the field of view, tangent height 15.7 km, 5 th May [7] Calibration of the spectra to produce geo-located, radiometrically corrected spectra (level 1b) incorporates data from views of an on-board blackbody near 240 K and views of space at greater than 200 km altitude above the Earth s surface. The quality of the level 1b spectra are excellent with offsets in the calibrated spectra believed to be less than 2 nw/cm2 sr cm -1 [3]. Typical noise values for a single spectrum vary between 40 nw/cm 2 sr cm -1 at 800 cm -1 to less than half that value at 1200 cm -1. One of the aspects of the MIPAS data is that the instrument is rather sensitive to clouds/aerosols; cloud effects can appear as low resolution baseline shifts and as high resolution absorption lines (for thick clouds). Proc. Envisat Symposium 2007, Montreux, Switzerland April 2007 (ESA SP-636, July 2007)

2 These effects can be seen in Figure 1. The cloud particle effects also have implications for the retrieval of trace gases in the operational processor. By defining a cloud index as the ratio of the signals in two designated cloud microwindows (MIPAS band A), cm -1 and cm -1, it is possible to flag cloud effects [4] and remove cloud-contaminated spectra from the retrieval process. In the operational processor, a cloud index of less than 1.8 between 6 km and 45 km is used to identify clouds. Nonetheless it is also possible that optically thinner clouds (cloud index greater than 1.8) may also cause residual errors in the trace gas retrievals. [This subject is returned to in Section 4]. 3. REFERENCE ATMOSPHERES The MIPAS reference atmospheres consist of two types, standard atmospheres and the IG2 reference climatology. The standard atmospheres provide climatological profiles for five atmospheric states: tropical, mid-latitude (day/night), polar winter and polar summer (also sigma, maximum and minimum profiles). Current data versions are V3.1 for the MIPAS standard atmospheres (available for download at ml#atm). The IG2 climatology represents the seasonal average atmosphere through a four season, six latitude band set of states. The four seasons are centred on January, April, July and October. The six latitude bands are ±(90-65 ), ±(65-20 ), and ±(20-0 ). A common altitude grid of km js employed for both the five standard atmospheres and the seasonal climatologies. The current version of the IG2 seasonal climatology is V4.0 which has been delivered to ESA and will form part of the next processing update for the MIPAS operational processor. These data will become available as part of the auxiliary data for the MIPAS processing but are also available from the lead author of this paper. Both types of atmosphere cover pressure and temperature, and concentrations of thirty six species: N 2, O 2, C 2 H 2, C 2 H 6, CO 2, O 3, H 2 O, CH 4, N 2 O, F11, F12, F13, F14, F21, F22, F113, F114, F115, CH 3 Cl, CCl 4, HCN, NH 3, SF 6, HNO 3, HNO 4, NO, NO 2, SO 2, CO, HOCl, ClO, H 2 O 2, N 2 O 5, OCS, ClONO 2, COF 2. One of the major updates to the IG2 climatology relevant to IASI concerns carbon dioxide. Clearly, the concentrations of greenhouse gases in the climatologies need continuous updates due to trends in their concentrations. However, it is also the case that for carbon dioxide there is some horizontal and vertical structure to the concentrations in the troposphere and stratosphere due to the combined effects of source distributions, the seasonal cycles in CO 2 due to vegetative uptake, and the nature of transport into the stratosphere with a slow transport of air to the poles. Our new CO 2 seasonal IG2 climatology includes these effects. The process starts with ground-based and aircraft measurements from the Globalview data set, which are advected into the stratosphere according to age of air relationships as a function of N 2 O [5]; N2O values in the stratosphere have been derived using compact relationships derived from ATMOS data [6]. Upper atmosphere CO2 has been further modified to follow [7] above 80 km and a methane oxidation term has also been added in the upper stratosphere. Figure 3 shows latitude dependent profiles for April 2003 compared to a yearly average 2003 profile for CO 2. The difference plot clearly highlights the structure present in the CO 2 profiles, particularly in the troposphere. In the lower troposphere, there is a clear difference between the polluted northern hemisphere and the cleaner southern hemisphere. There is also considerable variability in the upper troposphere with potential for strong gradients into the lower stratosphere. Differing values in the stratosphere reflect age of air variations with latitude as expected. Figure 3 Differences of seasonal CO 2 profiles for April 2003 from a yearly average of CO 2 data for Similarly it will be important to re-visit the MIPAS reference atmospheres for all tropospheric species to consider in more detail the variability of key species. Currently the IG2 seasonal climatology already contains variations, for example of tropospheric ozone. Nonetheless, scrutiny of the climatologies and updates for the range of satellite data available will provide insight into the climatological representation. 4. COMPARISONS WITH MIPAS OPERATIONAL PRODUCTS The seven operational MIPAS products are profiles of pressure/temperature, H 2 O, O 3, NO 2, N 2 O, CH 4 and HNO 3. Comparisons between these products and the

reference atmospheres have been performed in order to determine the mean consistency of the data and also to establish significant discrepancies which indicate inaccuracies in the reference

3 reference atmospheres have been performed in order to determine the mean consistency of the data and also to establish significant discrepancies which indicate inaccuracies in the reference atmosphere or errors in the MIPAS retrieval. Figures 4 to 9 show example comparisons between the reference data and monthly MIPAS data from The IG2 climatologies (orange line) with the associated onesigma values (dotted orange lines) are plotted together with standard atmospheres with maximum and minimum thresholds (blue lines). These comparisons were performed on a monthly basis, with MIPAS data from Individual MIPAS data points are represented by small crosses, with blue indicating cloud-free measurements, and red crosses indicating measurements with a cloud index in band A of less than 1.8 (i.e. cloud-flagged). Also included in these comparisons are mean profiles from the MIPAS data, represented by the larger green (cloud free) and red (cloud flagged) crosses. Associated sigma values were included in this analysis, with calculated one-sigma values from the IG2 climatologies (light orange dotted lines) compared to sigma uncertainties in the standard atmospheres ((maxmin)/6: the dark orange dashed lines) and the standard deviation of MIPAS measurements for a given month for cloud-free scenes (green diamonds) and cloudflagged scenes (red asterisks). Although all 12 months of data in 2003 were analysed, January has been selected for most comparisons in this paper and is representative of general patterns. Figure 4 shows the comparison for H 2 O. The dry constrained nature of stratospheric H 2 O, with values limited to a few ppmv is evident in the figure as expected. Standard deviations are similar between MIPAS and the IG2 data in the southern hemisphere summer but are larger for MIPAS data in the northern hemisphere winter. Below 100 mb variability increases as expected, however mean MIPAS profiles are lower than either reference atmosphere, particularly near the poles. Figure 4. Data for H 2 O from January Top Panel: New IG2 climatologies (orange lines with dotted onesigma margins), standard atmospheres (blue lines with dotted maximum and minimum values), MIPAS data separated into cloud-free (small blue crosses) and cloud-flagged (small red crosses) data points, with mean profiles for cloud-free (large green crosses) and cloud-flagged (large red crosses). Bottom Panel: Sigmas for the data sets in the top panel including IG2 climatologies (light orange dotted line), standard atmospheres (dark orange dashed line), MIPAS cloudfree data standard deviation (green diamonds) and MIPAS cloud-flagged data standard deviation where applicable (dark orange stars). Figure 5 shows ozone data from January 2003 and demonstrates again the well-constrained and characterised stratospheric profile, with more variability clear in the Northern high latitudes. Tropospheric concentrations from MIPAS are lower than the reference atmosphere profiles, particularly in equatorial regions and mid-latitudes.

Figure 5. Data for O 3 for January 2003 with legend as Figure 6 shows NO 2 data comparisons and highlights some of the current issues with the operational NO 2 product.

This is particularly true of the standard deviations in all latitudes and altitudes except for the southern polar regions.

However, clear separation can be seen in equatorial regions above 1 mb, with part of the data set clearly converging near the climatological profiles, and part of the data set retrieving

These are believed to be instabilities in the operational retrieval processor within the limits of convergence allowed for computing time [6]. Figure 6.

4 Figure 5. Data for O 3 for January 2003 with legend as Figure 6 shows NO 2 data comparisons and highlights some of the current issues with the operational NO 2 product. The MIPAS data for NO 2 are retrieved in a restricted altitude range within which the mean and sigma values for NO 2 are well characterised in the reference atmospheres and match well with MIPAS data. This is particularly true of the standard deviations in all latitudes and altitudes except for the southern polar regions. In the mean profile, the comparisons are very good except at the high altitudes where the diurnal variability of NO 2 complicates interpretation of the plots. However, clear separation can be seen in equatorial regions above 1 mb, with part of the data set clearly converging near the climatological profiles, and part of the data set retrieving significantly lower concentrations. Erroneous values of are also present in the operation product, and shown here. These are believed to be instabilities in the operational retrieval processor within the limits of convergence allowed for computing time [6]. Figure 6. Data for NO 2 for January 2003 with legend as Figure 7 presents data for N 2 O for January There is generally good agreement between the datasets from 0.1 to 80 mb. Increased variability in the troposphere is evident in the MIPAS data throughout all latitudes. N 2 O displayed distinct features during August which were not evident in January, therefore an additional dataset is shown in Figure 8. Stratospheric concentrations of N 2 O in Southern Polar regions are significantly lower than either set of reference atmospheres, demonstrating a clear discrepancy with current reference atmosphere representations for this region, at this time of year. Data for methane can be seen in Figure 9. Generally good agreement can be seen between 70 and 0.3 mb in tropical regions and between 200 and 0.3 mb at high latitudes. Tropical tropopause regions show particularly low concentrations of methane in the MIPAS data, with values down to 0.1 ppmv at 100 mb. Greater variability than predicted by either of the reference atmospheres can also be seen in tropospheric MIPAS data. Stratospheric concentrations during southern polar winter are also significantly lower in the MIPAS data.

5 Figure7. Data for N 2 O for January 2003 with legend as Figure 8. Data for N 2 O for August 2003 with legend as Figure 9. Data for CH 4 for January 2003 with legend as

Figure 10 shows data for nitric acid for January 2003 and demonstrates the generally solid agreement between MIPAS data and the two sets of reference atmospheres.

significant utility of comparisons of observed data with reference climatological atmospheric states. 6.

set. 7. REFERENCES Figure 10. Data for HNO 3 for January 2003 with legend as 5. SUMMARY A new V4.0 IG2 seasonal climatology has been released and will be incorporated in the operational processor.

6 Figure 10 shows data for nitric acid for January 2003 and demonstrates the generally solid agreement between MIPAS data and the two sets of reference atmospheres. The mean profile agreement in the stratosphere is generally very good. The most significant discrepancies are seen in low MIPAS concentrations from 5 to 10 and from 100 to 500 mb. significant utility of comparisons of observed data with reference climatological atmospheric states. 6. ACKNOWLEDGEMENTS The authors would like to thank ESA for provision of MIPAS data (under AO-357, CUTLSOM), the MIPAS QWG for many valuable comments and the NOAA for provision of the Globalview data set. 7. REFERENCES Figure 10. Data for HNO 3 for January 2003 with legend as 5. SUMMARY A new V4.0 IG2 seasonal climatology has been released and will be incorporated in the operational processor. A significant update has been the development of a time varying CO 2 climatology, as well as corrections for N 2 O and pressure/temperature. The new climatology has been compared to MIPAS observations for 2003 (full spectral resolution mode) and are revealing. It is found that the MIPAS data agree very well in mean behaviour at most altitudes. Discrepancies due to clouds and noise have been noted at the upper and lower limits of the MIPAS retrieval range. There are also some regions where MIPAS data are not in good agreement with the mean climatological profiles particularly in polar regions in the lower stratosphere and troposphere. These will be investigated further. However, the results presented illustrate the 1. Fischer, H., and H. Oelhaf, Remote sensing of vertical profiles of atmospheric trace constituents with MIPAS limb-emission spectrometers, Appl. Opt., 35, , Fischer H. et al., Envisat-MIPAS An instrument for atmospheric chemistry and climate research, editors: C. Readings and R.A. Harris, ESA Publication SP- 1229, European Space Agency, Spang, R., J. J. Remedios, L. J. Kramer, L. R. Poole, M. D. Fromm, M. Müller, G. aumgarten, P. Konopka, Polar stratospheric cloud observations by MIPAS on ENVISAT: detection method, validation and analysis of the northern hemisphere winter 2002/2003, Atmospheric Chemistry and Physics, 5, , Spang R., Remedios, J.J., and Barkley M.P., Colour indices for the detection and differentiation of cloud types in infra-red limb emission spectra, Adv. Space Res., 33, 7, , Andrews, A. et al., Empirical age spectra from observations of stratospheric CO2: Mean ages, vertical ascent rates, and dispersion in the lower tropical stratosphere, J. Geophys. Res., 104, 26,581-26,595, Raspollini, P., C. Belotti, A. Burgess, B. Carli, M. Carlotti, S. Ceccherini, B.M., Dinelli, A. Dudhia, J.M. Flaud, B. Funke, M. Hoepfner, M. Lopez-Puertas, V. Payne, C. Piccolo, J.J. Remedios, M. Ridolfi and R. Spang, MIPAS level 2 operational analysis, Atmospheric Chemistry and Physics, 6, , Greenhough, J., Remedios, J. J., Sembhi, H. & Kramer L.J. (2005). Towards cloud detection and cloud frequency distributions from MIPAS infra-red observations, Adv. Space Res. 36,

CLOUD DETECTION AND DISTRIBUTIONS FROM MIPAS INFRA-RED LIMB OBSERVATIONS

CLOUD DETECTION AND DISTRIBUTIONS FROM MIPAS INFRA-RED LIMB OBSERVATIONS J. Greenhough, J. J. Remedios, and H. Sembhi EOS, Space Research Centre, Department of Physics & Astronomy, University of Leicester,