Hi. I am working on seasonal forecast recently and found seasonal hindcast data on CDS. But I found that downloading seasonal forecast data on pressure levels always took dozens of hours to process. For example, here is the API request:


Originating centre:
CMCC
System:
3
Variable:
Geopotential
Pressure level:
500 hPa
Product type:
Monthly mean
Year:
1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011, 2012, 2013, 2014, 2015, 2016
Month:
January, February, March, April, May, June, July, August, September, October, November, December
Leadtime month:
1, 2, 3, 4, 5, 6
Whole available region:

Format:
GRIB


It has been "in process" for 20 hours and I even don't know when it will finish. However, it seems that variables on surface level, such as precipitation, MSLP is significantly faster to be prepared. Is there any trick to speed up downloading data on pressure levels? Do I need to split the dataset by years or months, or select more levels or variables to avoid the time of abstracting data?

Thank you!

6 Comments

  1. Hi,

    Users are encouraged to devise their requests as follows:

    • It is much more efficient to request one single month (e.g. November) for all the hindcast years (1993 to 2016) than requesting all 12 months from a single year
    • Combine in one single request all variables of interest, all levels (for pressure level fields), all product types and all leadtimes (hours or months, depending on the CDS product). However, note that there are some limitations in the size of a request. If you are using the CDS forms directly, you will be informed when you request is too big before submitting it. If you are using CDS API you can use the forms to discover whether the request you are trying to submit is affordable or not.

    Thanks

    Michela


    1. That is  very helpful to me. Thank you!

    2. Hi, 

      In your recommadations above, you said that we CDS users can combine some variables and time(like less than one month) we interest in one request.

      However, in my attempts, i've found a series of failure on doing like this. With messages like

      "Exception: the request you have submitted is not valid. Expected 1, got 137.; Request failed; Some errors reported (last error -1)."Request for ERA5 model level data t in a time.

      "Exception: the request you have submitted is not valid. Expected 1, got 720.; Request failed; Some errors reported (last error -1)." Request for ERA5 model level data lnsp in one month.

      Are these restrictions on retrieving model level data? Or what other methods i can use to retrieve model level data more efficient?

      Thanks.

      Davy.  

  2. Hi Davy,

    In this case you need to check the variable you are requesting. Some variables are only defined on 1 model level, rather than all 137. This could be the cause of the error you see.

    Hope that helps,

    Kevin

    1. Hi Kevin,

      I've figured out my mistake. Thanks

      Wish you a happy weekend.

      ^^

      Davy.

    2. Hi kevin,

      I want to konw lnsp(Logarithm of surface pressure) only defined on 1 model level?