# Minor difference between MARS archive data and 

**URL:** <https://forum.ecmwf.int/t/minor-difference-between-mars-archive-data-and/14253>\
**Category:** Open data\
**Tags:** mars\
**Created:** [14 October 2025 09:26 UTC](https://forum.ecmwf.int/t/minor-difference-between-mars-archive-data-and/14253 "2025-10-14T09:26:10Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![Toshiyuki\_Bandai](https://forum.ecmwf.int/letter_avatar_proxy/v4/letter/t/958977/32.png) [@Toshiyuki\_Bandai](https://forum.ecmwf.int/u/Toshiyuki_Bandai)\
**Post date:** [14 October 2025 09:26 UTC](https://forum.ecmwf.int/t/minor-difference-between-mars-archive-data-and/14253/1 "2025-10-14T09:26:10Z")

</div>

I’m seeing small differences between ECMWF Open Data and MARS. For example, when I compare the temperature data at p = 1000 hPa for step=24h at 0.25 degrees resolution between ECMWF Open Data and MARS, there is a tiny difference like below. The difference seems a noise rather than a systematic difference. I’d like to confirm whether this is expected (e.g., due to grid/interpolation/packing) or if I’m comparing the products incorrectly. Note that ECMWF Open Data is given with 0.25 degrees resolution while MARS data is regridded to 0.25 degrees when downloading.

 ![image](https://forum.ecmwf.int/uploads/default/original/2X/9/9fd0b6eb563226199279792283cd78c3ee2ef301.jpeg)