Abstract's details
Inspecting Jason-3 and Sentinel-3 WPD over their first 3 years of mission
CoAuthors
Event: 2019 Ocean Surface Topography Science Team Meeting
Session: Instrument Processing: Propagation, Wind Speed and Sea State Bias
Presentation type: Type Oral
Contribution: PDF file
Abstract:
Since 1992, satellite altimetry observations from two complementary missions have been playing a crucial role in studies such as climate change. Launched in January 2016, Jason3 (J3) is the fourth of the series of the so-called reference missions, flying in 10-day repeat orbits. Launched in February 2016, Sentinel-3A (S3A) is the first of the Sentinel-3 constellation, in a new 27-day repeat orbit, followed by Sentinel-3B (S3B) in April 2018, flying on the same orbit, 140º apart. The 27-day orbit replaced the 35-day configuration of the ERS/Envisat and SARAL series.
The importance of having range and geophysical corrections with the same level of accuracy of the altimeter range retrievals has long been recognised. Amongst these corrections, the wet tropospheric correction (WTC), accounting for the path delay induced by the presence of water vapour and cloud liquid water in the atmosphere, has particular relevance due to its high 4D variation. The most accurate method to retrieve the WTC is by means of collocated observations made by microwave radiometers (MWR) placed aboard the main altimetric missions. However, these instruments and respective retrieval algorithms have been designed for open ocean. Moreover, they are prone to instabilities such as drifts or external interferences.
The aim of this study is to inspect the MWR on board Jason-3 and Sentinel-3 and corresponding wet path delays (WPD) derived from these instruments, in terms of their accuracy and stability. The absolute and relative calibrations of these MWR are crucial tasks to ensure the retrieval of sea level climate data records from the respective altimeters.
An in-depth study of J3, S3A and S3B MWR performance has been carried out by means of i) validation against other MWR such as the Special Sensor Microwave Imager Sounders (SSMIS) aboard F16, F17 and F18 and the Global Precipitation Measurement (GPM) Microwave Imager (GMI), with GNSS observations, with a GPD+ WTC using only third-party data and with the ERA5 reanalysis model; ii) inter-comparison of the 3 instruments, namely S3A/S3B during the S3 mission tandem phase. In what concerns the analysed parameters, both the brightness temperature (TB) of the various MWR channels and the WTC were analysed. The two retrieval algorithms present on S3A products (from 3 and 5 parameters) have also been assessed.
Results reveal an RMS difference between SSM/IS and J3 of 1.2 and 0.9 cm before and after adjustment of the two datasets, solving for two parameters: offset (0.95 cm) and scale factor (0.99). On average, J3 measures dryer than SSM/IS by 1 cm.
After the last reprocessing, the differences between S3A and S3B are very small (∆TB23 = -0.2 ± 0.4 K; ∆TB36 = -0.1 ± 0.8 K; ∆WTC = 0.2 ± 0.2 cm), evidencing the good agreements between the two sensors.
Concerning S3A data for the first 3 years of mission, the MWR-derived WTC agrees well with that from the GMI and J3 MWR, with RMS differences of 0.9 and 1.1 cm, respectively. Scale factors and offsets of S3A with respect to other sensors are 1.0 and 0.2 cm (w.r.t. GMI) and 1.0 and 0.1 cm (w.r.t J3, previously adjusted to SSM/IS), respectively. This illustrates that, once J3 is aligned with the SSM/IS sensors, both J3, S3A and S3B sensors seem to be aligned.
Strong ice and land contamination can be observed in both S3A and S3B instruments, the latter one being mainly up to 25 km, as shown by comparisons with GNSS. The corresponding results for J3, reveal a much reduced land effect, still up to about 25 km from the coast. No significant difference has been found between the two WPD retrieval algorithms adopted in Sentinel-3.
By comparison with SSM/IS, GMI, GNSS and the GPD+ WTC solely derived with third-party data, the long-term evolution of J3 and S3A radiometers for the first 3 years of their missions has been analysed. Considering this time span, too short to infer any possible drift in any of these instruments, the respective WPD seem to be relatively stable.
In spite of the good overall performance of J3 and S3A/B radiometers over open-ocean, in view to get accurate WTC, valid everywhere, in particular in coastal and polar regions, updated GPD+ WTC have been derived and the gain obtained by these datasets in these regions will also be presented.
The importance of having range and geophysical corrections with the same level of accuracy of the altimeter range retrievals has long been recognised. Amongst these corrections, the wet tropospheric correction (WTC), accounting for the path delay induced by the presence of water vapour and cloud liquid water in the atmosphere, has particular relevance due to its high 4D variation. The most accurate method to retrieve the WTC is by means of collocated observations made by microwave radiometers (MWR) placed aboard the main altimetric missions. However, these instruments and respective retrieval algorithms have been designed for open ocean. Moreover, they are prone to instabilities such as drifts or external interferences.
The aim of this study is to inspect the MWR on board Jason-3 and Sentinel-3 and corresponding wet path delays (WPD) derived from these instruments, in terms of their accuracy and stability. The absolute and relative calibrations of these MWR are crucial tasks to ensure the retrieval of sea level climate data records from the respective altimeters.
An in-depth study of J3, S3A and S3B MWR performance has been carried out by means of i) validation against other MWR such as the Special Sensor Microwave Imager Sounders (SSMIS) aboard F16, F17 and F18 and the Global Precipitation Measurement (GPM) Microwave Imager (GMI), with GNSS observations, with a GPD+ WTC using only third-party data and with the ERA5 reanalysis model; ii) inter-comparison of the 3 instruments, namely S3A/S3B during the S3 mission tandem phase. In what concerns the analysed parameters, both the brightness temperature (TB) of the various MWR channels and the WTC were analysed. The two retrieval algorithms present on S3A products (from 3 and 5 parameters) have also been assessed.
Results reveal an RMS difference between SSM/IS and J3 of 1.2 and 0.9 cm before and after adjustment of the two datasets, solving for two parameters: offset (0.95 cm) and scale factor (0.99). On average, J3 measures dryer than SSM/IS by 1 cm.
After the last reprocessing, the differences between S3A and S3B are very small (∆TB23 = -0.2 ± 0.4 K; ∆TB36 = -0.1 ± 0.8 K; ∆WTC = 0.2 ± 0.2 cm), evidencing the good agreements between the two sensors.
Concerning S3A data for the first 3 years of mission, the MWR-derived WTC agrees well with that from the GMI and J3 MWR, with RMS differences of 0.9 and 1.1 cm, respectively. Scale factors and offsets of S3A with respect to other sensors are 1.0 and 0.2 cm (w.r.t. GMI) and 1.0 and 0.1 cm (w.r.t J3, previously adjusted to SSM/IS), respectively. This illustrates that, once J3 is aligned with the SSM/IS sensors, both J3, S3A and S3B sensors seem to be aligned.
Strong ice and land contamination can be observed in both S3A and S3B instruments, the latter one being mainly up to 25 km, as shown by comparisons with GNSS. The corresponding results for J3, reveal a much reduced land effect, still up to about 25 km from the coast. No significant difference has been found between the two WPD retrieval algorithms adopted in Sentinel-3.
By comparison with SSM/IS, GMI, GNSS and the GPD+ WTC solely derived with third-party data, the long-term evolution of J3 and S3A radiometers for the first 3 years of their missions has been analysed. Considering this time span, too short to infer any possible drift in any of these instruments, the respective WPD seem to be relatively stable.
In spite of the good overall performance of J3 and S3A/B radiometers over open-ocean, in view to get accurate WTC, valid everywhere, in particular in coastal and polar regions, updated GPD+ WTC have been derived and the gain obtained by these datasets in these regions will also be presented.