Abstract's details
A Wet Tropospheric Correction for Global Mean Sea Level application computed from CM SAF FCDR Microwave Imager Radiances
CoAuthors
Event: 2018 Ocean Surface Topography Science Team Meeting
Session: Instrument Processing: Propagation, Wind Speed and Sea State Bias
Presentation type: Type Poster
Contribution: not provided
Abstract:
The Wet Tropospheric Correction (WTC) is the main source of uncertainty on the Global Mean Sea Level (GMSL) trend. The uncertainty on the WTC trend could be as large as 10% of GMSL (Ablain et al. 2009, Obligis et al. 2010) : the uncertainty on the WTC trend is about ±0.3 mm/yr−1 compared to the trend of 3.2 mm/yr−1 for the GMSL with an uncertainty of 0.6 mm/yr−1
WTC computed from ECMWF model is not stable enough wrt the strong requirements on the GMSL (Legeais et al. 2014). The improvements during the last 25 years on successive versions of the analysis leads to jumps at each change of versions and the modification on the observation sources also leads to artifical trends on reanalysis. Consequently, WTC computed from microwave radiometers on-board altimetry missions is the reference for GMSL trend computation But almost all these instruments suffer from instrumental drift.
Solutions exist to account for these drifts as the correction of known instrumental drifts on Jason-2 and Jason-3 based on vicarious calibration (Brown 2013), the correction of inter-mission biases and trends on the WTC provided by MWR on-board altimetry missions by comparison to the water vapour provided by SSMI/S reference missions (Fernandes et al. (2016), Legeais et al. 2018) or by optimisation of biases compared to simulated brightness temperatures from ERA-Interim dataset (Bennartz et al. 2017).
What could we learn on the uncertainty of the WTC trend from the comparison to a WTC based on a dataset of TB dedicated to climate analysis ?
On the contrary to previous work, the objective of the present work is not to provide a correction of altimetry WTC but to compare this WTC timeserie to a WTC computed from a completly independant dataset claimed to meet strong requirements on instrumental stability
EUMETSAT offers a range of climate related data products. Amongst these data products, the CM SAF FCDR MWI (Fundamental Climate Data Record of Microwave Imager Radiances, Edition 3)
comprises inter-calibrated brightness temperatures (BTRs) from the SMMR, SSM/I and SSMIS radiometers. It provides homogenised and inter-calibrated BTRs in a user friendly data format, for a variety of applications, such as analyses of the hydrological cycle and remote sensing of sea ice.
It covers the time period from October 1978 to December 2015.
The validation reports contains the results from a comparison to other FCDR (RSS Version-6 SSM/I FCDR (Semunegus 2011), Colorado State University FCDR) in terms of biases, variability and stability. The performances between CM SAF and other FCDR datasets are very similar in terms of mean bias ajd the stability of the FCDR is about 0.05 K/dec for ever channels on every platforms.
A WTC is estimated directly from FCDR brightness temperatures using the same Neural Network retrieval used for operationnal missions (AltiKa, Sentinel-3). The WTC estimated from CF SAF FCDR MWI dataset provides a consistent time series with no detectable remaining biases between missions. It is consistent with Jason-2 WTC timeseries on the overlapping period some preliminary results on the comparison of WTC FCDR to WTC computed from MWR on-board altimetry missions (User Product) seems to highlight a similar stability between the two datasets over the whole period. If this is confirmed by specific validation on GSML, the historical uncertainty of 0.3 mm/yr −1 on the WTC trend could be reduced to the stability of the FCDR dataset, 0.05 K/dec roughly converted to 0.025 mm/yr −1. This study highlights the value of such FCDR dataset based on radiances for applications to a subject not directly related to atmospheric studies.
WTC computed from ECMWF model is not stable enough wrt the strong requirements on the GMSL (Legeais et al. 2014). The improvements during the last 25 years on successive versions of the analysis leads to jumps at each change of versions and the modification on the observation sources also leads to artifical trends on reanalysis. Consequently, WTC computed from microwave radiometers on-board altimetry missions is the reference for GMSL trend computation But almost all these instruments suffer from instrumental drift.
Solutions exist to account for these drifts as the correction of known instrumental drifts on Jason-2 and Jason-3 based on vicarious calibration (Brown 2013), the correction of inter-mission biases and trends on the WTC provided by MWR on-board altimetry missions by comparison to the water vapour provided by SSMI/S reference missions (Fernandes et al. (2016), Legeais et al. 2018) or by optimisation of biases compared to simulated brightness temperatures from ERA-Interim dataset (Bennartz et al. 2017).
What could we learn on the uncertainty of the WTC trend from the comparison to a WTC based on a dataset of TB dedicated to climate analysis ?
On the contrary to previous work, the objective of the present work is not to provide a correction of altimetry WTC but to compare this WTC timeserie to a WTC computed from a completly independant dataset claimed to meet strong requirements on instrumental stability
EUMETSAT offers a range of climate related data products. Amongst these data products, the CM SAF FCDR MWI (Fundamental Climate Data Record of Microwave Imager Radiances, Edition 3)
comprises inter-calibrated brightness temperatures (BTRs) from the SMMR, SSM/I and SSMIS radiometers. It provides homogenised and inter-calibrated BTRs in a user friendly data format, for a variety of applications, such as analyses of the hydrological cycle and remote sensing of sea ice.
It covers the time period from October 1978 to December 2015.
The validation reports contains the results from a comparison to other FCDR (RSS Version-6 SSM/I FCDR (Semunegus 2011), Colorado State University FCDR) in terms of biases, variability and stability. The performances between CM SAF and other FCDR datasets are very similar in terms of mean bias ajd the stability of the FCDR is about 0.05 K/dec for ever channels on every platforms.
A WTC is estimated directly from FCDR brightness temperatures using the same Neural Network retrieval used for operationnal missions (AltiKa, Sentinel-3). The WTC estimated from CF SAF FCDR MWI dataset provides a consistent time series with no detectable remaining biases between missions. It is consistent with Jason-2 WTC timeseries on the overlapping period some preliminary results on the comparison of WTC FCDR to WTC computed from MWR on-board altimetry missions (User Product) seems to highlight a similar stability between the two datasets over the whole period. If this is confirmed by specific validation on GSML, the historical uncertainty of 0.3 mm/yr −1 on the WTC trend could be reduced to the stability of the FCDR dataset, 0.05 K/dec roughly converted to 0.025 mm/yr −1. This study highlights the value of such FCDR dataset based on radiances for applications to a subject not directly related to atmospheric studies.