Replacement of GOES-15 with GOES-17 AMVs

Replacement of GOES-15 with GOES-17 AMVs
Date Published
EUMETSAT/ECMWF Fellowship Programme Research Report
Document Number
Katie Lean

GOES-17 is the second new generation satellite in the GOES-R series, after GOES-16, carrying the Advanced Baseline Imager (ABI) from which Atmospheric Motion Vectors (AMVs) are derived. It replaces
GOES-15 as the primary satellite in the GOES-West position at 137.2?W. Being the same generation as GOES-16, both satellites carry the same imaging instrument while also using the same wind derivation algorithm. Consequently, it would be reasonable to expect the data quality of GOES-17 to be very similar to GOES-16. However, onboard issues with the cooling system have been found to be periodically degrading the performance of the GOES-17 infrared channels which are essential to the production of AMVs. The problem leads to intervals where the imager data are unusable during local night-time with transition times where data quality is variable.
During times unaffected by the cooling issues, the combination of the different instrument and algorithm results in a large increase in the number of AMVs and changes in data characteristics compared to it’s predecessor, GOES-15. Similar changes were observed in earlier work when moving from GOES-13 to GOES-16 (Lean and Bormann, 2019). Analysis using first guess departure statistics further shows that there is good agreement in the overlap region with GOES-16 while the instrument is operating normally. However, during transitional times where onboard cooling issues are in effect there is a decline in AMV numbers and a noticeable degradation in the quality. There are large differences in the assigned pressure of the AMVs between GOES-17 and GOES-16 indicating that the degradation in the infrared channels has a significant impact in the performance of the optimal estimation height method.
Assimilation experiments tested the replacement of GOES-15 with GOES-17 using different levels of restriction on use of the data around periods affected by the cooling issues. Results show that GOES-17, even with the strictest screening option that removes all data from 9-17Z, has a positive impact. Significant reductions in the vector wind forecast error were observed in the verification against own analysis. Small positive changes were also present in the fit of independent observations to the model background such as conventional wind over the Americas and humidity sensitive channels on the Advanced Technology Microwave Sounder (ATMS) in the southern hemisphere. Changes were generally comparable or in some cases exceeding the impact from GOES-15. Active assimilation of GOES-17 AMVs started on 10 December 2019.

DOI 10.21957/adwfue2ph