# ERA5-Land data grid mismatch on recently downloaded data vs older data **URL:** \ **Category:** C3S - Datasets and Usage\ **Tags:** era5-land\ **Created:** [30 January 2025 09:28 UTC](https://forum.ecmwf.int/t/era5-land-data-grid-mismatch-on-recently-downloaded-data-vs-older-data/11080 "2025-01-30T09:28:19Z")\ **Posts on this page:** 8\ **Page:** 1
**Author:** ![Luke\_Parsons](https://forum.ecmwf.int/letter_avatar_proxy/v4/letter/l/67e7ee/32.png) [@Luke\_Parsons](https://forum.ecmwf.int/u/Luke_Parsons)\ **Post date:** [30 January 2025 09:28 UTC](https://forum.ecmwf.int/t/era5-land-data-grid-mismatch-on-recently-downloaded-data-vs-older-data/11080/1 "2025-01-30T09:28:19Z")
Over the last ~month I re-downloaded ERA5-Land t2m data (hourly 1950-2023). The data have the same number of lat/lon grid points as the data I downloaded a 1-1.5 years ago, but the data grids don’t match any more. For example, the ERA5-Land land-sea mask no longer matches, nor do any of the other nc files I regridded to match the old ERA5-Land data I downloaded in 2023. Has there been a change to the lat/lon grid structure or formatting in the ERA5-Land data? ---
**Author:** ![Imanol\_Uriarte\_Latorre](https://forum.ecmwf.int/letter_avatar_proxy/v4/letter/i/898d66/32.png) [@Imanol\_Uriarte\_Latorre](https://forum.ecmwf.int/u/Imanol_Uriarte_Latorre)\ **Post date:** [30 January 2025 09:58 UTC](https://forum.ecmwf.int/t/era5-land-data-grid-mismatch-on-recently-downloaded-data-vs-older-data/11080/2 "2025-01-30T09:58:19Z")
Can you give the numbers of the displacement? What’s the origin? I’d like to check on my side too ---
**Author:** ![Matt](https://forum.ecmwf.int/letter_avatar_proxy/v4/letter/m/a6a055/32.png) [@Matt](https://forum.ecmwf.int/u/Matt)\ **Post date:** [30 January 2025 12:40 UTC](https://forum.ecmwf.int/t/era5-land-data-grid-mismatch-on-recently-downloaded-data-vs-older-data/11080/3 "2025-01-30T12:40:06Z")
Are the files netcdfs? I think this might have to do with the [change in the grib/netcdf converter from grib\_to\_netcdf to cfgrib](https://prod.ecmwf-forum-prod.compute.cci2.ecmwf.int/t/changes-to-grib-to-netcdf-converter-on-cds-beta-ads-beta/4322/35), it seems the lat/lon coordinates are not well preserved (-89.7999999999983 instead of -89.8 for instance) ---
**Author:** ![Tristan\_Schuler](https://forum.ecmwf.int/user_avatar/forum.ecmwf.int/tristan_schuler/32/872_2.png) [@Tristan\_Schuler](https://forum.ecmwf.int/u/Tristan_Schuler)\ **Post date:** [30 January 2025 15:25 UTC](https://forum.ecmwf.int/t/era5-land-data-grid-mismatch-on-recently-downloaded-data-vs-older-data/11080/4 "2025-01-30T15:25:35Z")
Yes both files are netcdf. Where are you seeing the discrepancy in lat/lon? I have exact 0.25 intervals in both datasets . ---
**Author:** ![Matt](https://forum.ecmwf.int/letter_avatar_proxy/v4/letter/m/a6a055/32.png) [@Matt](https://forum.ecmwf.int/u/Matt)\ **Post date:** [31 January 2025 09:43 UTC](https://forum.ecmwf.int/t/era5-land-data-grid-mismatch-on-recently-downloaded-data-vs-older-data/11080/5 "2025-01-31T09:43:49Z")
I was looking at the 0.1º data, so it might be a different issue. Anyway, with the change in converter, discrepancies in the resulting netcdfs are to be expected: > It is important to understand that given the intrinsic differences between GRIB and netCDF there is not a one size fits all approach when converting GRIB to netCDF. We aim to provide users with some sensible default options, but given the code used to perform the conversion is open source and documented it is possible for users to fine tune the conversion to their requirements if needed. [https://confluence.ecmwf.int/display/CKB/GRIB+to+netCDF+conversion+on+new+CDS+and+ADS+systems#GRIBtonetCDFconversiononnewCDSandADSsystems-Whywearechanging](https://confluence.ecmwf.int/display/CKB/GRIB+to+netCDF+conversion+on+new+CDS+and+ADS+systems#GRIBtonetCDFconversiononnewCDSandADSsystems-Whywearechanging) ---
**Author:** ![Luke\_Parsons](https://forum.ecmwf.int/letter_avatar_proxy/v4/letter/l/67e7ee/32.png) [@Luke\_Parsons](https://forum.ecmwf.int/u/Luke_Parsons)\ **Post date:** [31 January 2025 16:30 UTC](https://forum.ecmwf.int/t/era5-land-data-grid-mismatch-on-recently-downloaded-data-vs-older-data/11080/6 "2025-01-31T16:30:01Z")
Hi all, thanks for the replies! I am also looking at the 0.1 degree data as Matt indicated. Yes, the files are netcdf (nc) files that I downloaded both times. If the GRIB to nc converter varies this would make sense. I output the latitude values in Xarray for the older nc t2m data from 2023: [90. , 89.9, 89.8, …, -89.8, -89.9, -90.] And for the newer t2m zipped nc data I downloaded in the last few months: [90. , 89.9, 89.8, …, -89.8, -89.9, -90.] So I don’t see an immediate difference, but if I calculate the difference in latitude values and print that, I do see tiny differences: [ 0.00000000e+00, -1.52587890e-06, -3.05175780e-06, …, 3.05175948e-06, 1.52588058e-06, 0.00000000e+00] ---
**Author:** ![Manuarii\_Vazquez\_Ballesta](https://forum.ecmwf.int/letter_avatar_proxy/v4/letter/m/a8b319/32.png) [@Manuarii\_Vazquez\_Ballesta](https://forum.ecmwf.int/u/Manuarii_Vazquez_Ballesta)\ **Post date:** [23 July 2026 09:30 UTC](https://forum.ecmwf.int/t/era5-land-data-grid-mismatch-on-recently-downloaded-data-vs-older-data/11080/7 "2026-07-23T09:30:07Z")
Hi, I recently encounter the same issues with ERA5 at 0.25°, where grid is not more align wit my previous dowload dataset. I’ve dowload my first dataset in 2024-02 using : `grib_to_netcdf-2.25.1: /opt/ecmwf/mars-client/bin/grib_to_netcdf.bin -S param -o /cache/data4/adaptor.mars.internal-1707106744.416648-11385-9-0fb20271-f29c-4fad-a10e-05ffa037302c.nc /cache/tmp/0fb20271-f29c-4fad-a10e-05ffa037302c-adaptor.mars.internal-1707087681.3455684-11385-13-tmp.grib ;` `}` I have this figure : ![beofre_ERA5](https://forum.ecmwf.int/uploads/default/original/2X/c/c32e9e34b7a423b1953aac2ff77a074d9906393a.png) And now (2026-07) when I download data I get a mismatch of grid compare to my previous dataset, which use: `GRIB to CDM+CF via cfgrib-0.9.15.1/ecCodes-2.42.0 with {"source": "tmptzd_v3i3/data.grib", "filter_by_keys": {"stream": ["oper"], "stepType": ["instant"]}, "encode_cf": ["parameter", "time", "geography", "vertical"]} ;` `}` Now I have this figure : ![now_ERA5](https://forum.ecmwf.int/uploads/default/original/2X/e/e18f40d6a9c82f4de2ee6ea14f5173d0290036b9.png) Is it link to the grib to netcdf converter ? Thanks in advance ! ---
**Author:** ![Michela](https://forum.ecmwf.int/letter_avatar_proxy/v4/letter/m/898d66/32.png) [@Michela](https://forum.ecmwf.int/u/Michela)\ **Post date:** [27 July 2026 11:17 UTC](https://forum.ecmwf.int/t/era5-land-data-grid-mismatch-on-recently-downloaded-data-vs-older-data/11080/8 "2026-07-27T11:17:04Z")
Hi, if you used the area keyword please look at the following link: [Software upgrade for data extraction of a geographical area from selected ERA5 and Seasonal forecast datasets](https://forum.ecmwf.int/t/software-upgrade-for-data-extraction-of-a-geographical-area-from-selected-era5-and-seasonal-forecast-datasets/14583) Thanks