Background
The ROM SAF has produced two climate reanalyses as part of its CDOP2 activities. These were recommended by the IROWG climate subgroup (see section 4.2), where is was stated that :
"Encourage NWP centers to engage in reanalysis activities based only on data types that are not bias-corrected, especially RO. ECMWF, e.g., carries out such an activity as part of the ROM SAF CDOP-2 plans."
Two reanalyses that meet this recommendation have been completed, for the period 20070101 to 20151231. The first reanalysis assimilates just reprocessed GPS-RO data provided the ROM SAF and AMSUA channel 14 to constrain the upper stratosphere. The second reanalysis also includes conventional measurements (e.g., sondes, synop, ...) that are not bias corrected as part of the variational bias correction (VAR-BC) scheme. Notably, this excludes aircraft temperature measurements that are bias corrected in VAR BC.
Access via web-api
The ROM SAF datasets are available from the ECMWF public datasets service via the web-api. This is a "programmatic approach" for data download with python scripts . We will not reproduce all the existing web-api documentation here, but we will note some key steps required to download the data and provide the main links. Firstly, you will need to download the appropriate python modules in order to run the python download scripts. For example, the python scripts will contain "from ecmwfapi import" commands, like in this simple example, retrieving a monthly mean zonal wind at 1000 hPa, in netcdf format (More complex retrieval scripts are provided here):
#!/usr/bin/env python
#
from ecmwfapi import ECMWFDataServer
# To run this example, you need an API key
# available from https://api.ecmwf.int/v1/key/
server = ECMWFDataServer(verbose=1)
server.retrieve({
'dataset' : "rom_saf_3198",
'stream' : "moda",
'levtype' : "pl",
'levelist': "1000",
'date' : "20150301",
'type' : "an",
'param' : "131.128",
'grid' : "0.5/0.5",
'format' : "netcdf",
'target' : "rom_saf_3198.nc"
})
General information related to downloading the data via web-api, is described here:
https://confluence.ecmwf.int/display/WEBAPI/Web-API+Downloads
Specifically, follow the steps here to install the latest code version:
https://pypi.org/project/ecmwf-api-client/
Briefly, install this with the pip command:
pip install ecmwf-api-client
You will now have the modules needed in the python scripts. (I actually needed this command for the ECMWF system pip install --user ecmwf-api-client because of local permissions).
Login and retrieve your "key":
Login at:
https://apps.ecmwf.int/auth/login/
Retrieve your key at:
Create and paste the "key" information in here:
$HOME/.ecmwfapirc
You will also need to accept the terms before you can access the data with web-api otherwise you will get an error when you run the script.
Need to sort out links to example scripts.