# CARRA2 - user information: x-y grid size may be different for different variables **URL:** \ **Category:** C3S - Datasets and Usage\ **Tags:** carra2\ **Created:** [16 July 2026 08:02 UTC](https://forum.ecmwf.int/t/carra2-user-information-x-y-grid-size-may-be-different-for-different-variables/15169 "2026-07-16T08:02:48Z")\ **Posts on this page:** 1\ **Page:** 1
**Author:** ![Andreas\_Walbroel](https://forum.ecmwf.int/letter_avatar_proxy/v4/letter/a/b5a626/32.png) [@Andreas\_Walbroel](https://forum.ecmwf.int/u/Andreas_Walbroel)\ **Post date:** [16 July 2026 08:02 UTC](https://forum.ecmwf.int/t/carra2-user-information-x-y-grid-size-may-be-different-for-different-variables/15169/1 "2026-07-16T08:02:49Z")
Hi, this is just an information to new users of CARRA2. When downloading sub-daily data from the Climate Data Store as netCDF files, the x-y grid of the Polar Stereographic projection may have a different size depending on the selected variable. For most variables (at least those that I have already downloaded, e.g., surface\_pressure) you will receive a file with a grid of size (x, y) = (2869, 2869). However, sea\_surface\_temperature and sea\_ice\_area\_fraction are on a grid (x, y) = (2880, 2880). Now, if you would like to merge separate data sets that use the (2869, 2869) grid, I recommend just selecting the first 2869 indices of the 2880-grid data, which ensures a perfect overlap with the other data set gridded on the 2869-grid. I verified this by comparing the latitude and longitude values of the cut 2880-grid and the regular 2869-grid. Due to download speed issues, I haven’t made the effort to try each single variable on each height grid (single level, pressure level, height level, model level). So I am not sure whether other variables might use the 2880-grid as well. I do not consider this a question but rather an information for other users. Best regards, Andreas