View Source

What is the objective of this page?

To help users to improve S2S CMA MARS requests performance via the WebAPI.

A good understanding of the MARS efficiency issues is essential especially for users that are interested in downloading large amounts of data.

How is the S2S reforecast data organised in MARS?

In general it is organised, as a huge tree, with the indentation below, showing different levels down that tree:

centre (CMA, ECMWF, NCEP, JMA, ...)
- realtime or reforecast
  - type of data (control forecast or perturbed forecast)
    - type of level (single level or pressure level or potential temperature)
      - model version date (2014-05-01 or ...)
        hindcast dates (2014-01-01 or 2014-01-02 or 2014-01-03, ...)
        time-steps
        members (for perturbed forecast)
        levels (for pl or pt)
        parameters

What would be the natural way to group requests?

The idea is to request as much data as possible from the same tape file. The natural way to group requests would be:
all parameters, all levels, all members, all time-steps for 1 hindcast date for a type of level for a type

Note the following:

'all' means 'all' that the user wants. It doesn't have to be all parameters.
If a user is interested only on z500, he may request more hindcast dates in one go, since the overall request will not be so big.

Best practise to iterate over all hindcastDates of several hindcastYears for CMA

The best approach is to iterate over the hindcastYears. For each hindcastYear iterate over all the available hindcastMonths and for each hindcastMonth iterate over all the available hindcastDays.

At this point you may wish to check CMA availability and to view a CMA request

for hindcastYear in hindcastYears
    for hindcastMonth in hindcastMonths        
        for hindcastDay in hindcastDays             
            hindcastDate = HindcastYear-hindcastMonth-hindcastDay               
                S2S-request(hindcastDate)

Web-API examples:

A CMA reforecast request for one hindcastDate

The request below is for all members of the perturbed forecast, 10 meter U and V wind components, for all time-steps and for model version 2014-05-01

#!/usr/bin/env python
from ecmwfapi import ECMWFDataServer
modelVersionDate = "2014-05-01"  # This is the first model version for CMA (babj)
hindcastDate = "2014-04-01"      # The selected hindcast date
server = ECMWFDataServer()
server.retrieve({
    "class": "s2",
    "dataset": "s2s",
    "date": modelVersionDate,
    "expver": "prod",
    "hdate": hindcastDate, 
    "levtype": "sfc",
    "origin": "babj",
    "param": "165/166",
    "step": "24/to/1440/by/24",
    "stream": "enfh",
    "target": "data.pf.sfc",
    "time": "00",
    "number": "1/2/3",
    "type": "pf",
})

A CMA reforecast request for all the available hindcastDates

The objective of this example is to demonstrate how you can write a request efficiently by iterating properly over several hindcastYears, hindcastMonths and hindcastDays
It can be used as a starting point however you need to adapt it to your needs eg to set the keyword values according to your requirements (eg "param", "levtype", "step" etc).
In this way you can extend this request to download the whole S2S CMA reforecast. Don't forget to check CMA availability

Please note:

use the variable "target" to write each hindcastDate on a separate file .
taking under consideration your request's size (eg nr of fields and volume) you can merge several hindcastDates on the same "target"

Useful links