News

GSICS Quarterly

From the GSICS Quarterly, Vol. 3, No. 3, 2009

Importance of satellite inter-calibration for CM-SAF data sets

The EUMETSAT Satellite Application Facility on Climate Monitoring (CM- SAF, www.cmsaf.eu) aims at the provision of satellite-derived geophysical parameter data sets suitable for climate monitoring (Schulz et al., 2009). CM-SAF focuses on the atmospheric part of the Essential Climate Variables (ECV), as defined within the Global Climate Observing System framework in support of the United Nations Framework Convention on Climate Change (UNFCCC). Satellite observations are vital for climate monitoring due to global coverage in combination with high spatial resolution. In addition, satellite data sets also need to cover a long time period in order to be useful for climate monitoring. The demands on accuracy increase with increasing length of the time series. At the seasonal to inter-annual scale the detection of small changes in an observed parameter requires already stringent accuracy levels. In order to detect trends in data sets covering decennial or centennial time scales, their accuracy must be one order of magnitude higher (Ohring et al., 2005) relative to the needs of detecting inter-annual variability. Thus, a central goal of CM-SAF is to further improve all existing CM-SAF data products to a quality level that allows for studies of inter-annual variability and beyond.

Two data set categories are, and will be, provided by CM-SAF. In near- real time, so called Environmental Data Records (EDR) are obtained by converting satellite sensor data into geophysical variables using nominal calibration. Within an operational environment such EDRs are integrated over time to obtain daily and monthly averages. In its current implementation these operational products are suitable for analyses on diurnal to seasonal time scales. The products support routine climate monitoring applications at the national meteorological services of the EUMETSAT member states. A second class of products is generated by improving the calibration, by homogenisation of the time series of data from different satellites, and by utilisation of only one single, state-of-the-art retrieval scheme for the whole series. Such Thematic Climate Data Records (TCDRs) are produced by a reprocessing of the whole data record of a satellite instrument and might extend the suitable analysis capability to inter-annual scales. In these records any obvious error, e.g., jumps created by instrument changes on successive satellites, are already resolved. Other subtle changes caused by changes in the spectral response function or orbital drift of satellites need further attention to achieve quality that is suitable for studies on decadal variability and trend detection.

Currently the following EDR products are available from CM-SAF:

• Cloud parameters (based on SEVIRI and AVHRR): cloud fractional cover, cloud type, cloud top properties, cloud phase, cloud optical thickness, and cloud water path;
• Radiation budget products (AVHRR, SEVIRI, CERES, DIARAD, VIRGO and GERB): Incoming shortwave, net shortwave, net longwave, downward and outgoing longwave radiation, surface radiation budget, various thermal radiative fluxes, and surface albedo; and
• Humidity products (ATOVS): Total column-integrated water vapour, layer- integrated water vapour, mean temperature and relative humidity for 5 layers, specific humidity and temperature at the six layer boundaries.

Recently, CM-SAF released an approximately 20-year TCDR of total column- integrated water vapour from SSM/I. The data record of SSM/I radiances was homogenised by matching radiance probability distribution functions of overlap periods between satellites. An exemplary anomaly analysis is presented in Figure 1.

CM-SAF will generate a long time series (~30 years) of free tropospheric humidity from MVIRI and SEVIRI in the near future. While the MVIRI time series is homogenised, its temporal extension with SEVIRI observations is an open issue. Also, a long time series of spectrally-resolved solar irradiances from MVIRI is currently under development at CM-SAF. For this, a self-calibration in count space was developed and implemented.

As described above, the main objective of CM-SAF is to establish TCDRs suitable for the detection of climate variability and for trend detection. Homogenisation and inter-calibration of the various satellite observations is a major and ambitious effort that needs international collaboration. CM-SAF can largely benefit from inter-calibration efforts carried out within GSICS. Partly, GSICS results can be used to confirm or improve currently applied homogenisations at CM-SAF. In particular for SEVIRI-dependent TCDRs, CM-SAF relies on GSICS (inter-)calibration coefficients or radiance records reprocessed by space agencies utilising GSICS results. Such efforts will strongly accelerate the production of TCDRs at CM-SAF. In addition, the (inter-)calibration coefficients might be useful for operational processing, if provided in near-real time.

CM-SAF aims at supporting GSICS by evaluating GSICS-type intercalibrated satellite radiances, employing a fast radiative transfer model (RTTOV) applied to observations from ground-based instruments such as lidars, microwave radiometers and research quality radiosondes from GCOS Reference Upper-Air Network (GRUAN) sites. These measurements can be operationally used to infer residual biases after applying the GSICS inter-calibration by comparing simulated and observed radiances. Additionally, a line-by-line radiative transfer model in the infrared will be utilised in an off-line mode to estimate uncertainties of RTTOV, and to extend the evaluation to the high spectral resolution observations of IASI on a regular basis. In this way the efforts of the (inter-)calibration of satellites, in particular of SEVIRI, carried out within the GSICS framework, are tested independently.

References

[Drs. M. Schröder and J. Schulz, (CM-SAF, DWD)]