# Keep averaging kernel variables

**URL:** https://forum.atmospherictoolbox.org/t/keep-averaging-kernel-variables/447
**Category:** HARP
**Tags:** sentinel-5p
**Created:** [December 22, 2021, 5:01am UTC](https://forum.atmospherictoolbox.org/t/keep-averaging-kernel-variables/447 "2021-12-22T05:01:24Z")
**Posts on this page:** 2
**Page:** 1

<div class="post-metadata">

### Author: ![atmosphericmodeling](https://forum.atmospherictoolbox.org/user_avatar/forum.atmospherictoolbox.org/atmosphericmodeling/32/214_2.png) [@atmosphericmodeling](https://forum.atmospherictoolbox.org/u/atmosphericmodeling)
#### Post date: [December 22, 2021, 5:01am UTC](https://forum.atmospherictoolbox.org/t/keep-averaging-kernel-variables/447/1 "2021-12-22T05:01:24Z")

</div>

Hello,  
I would like to compare S5P HCHO with model simulated results.  
For comparing simulated and observed results, I need averaging kernel.  
So I tried converting L2 tropospheric\_HCHO\_column\_number\_density\_avk variable into L3 grid by following code.

 ![image](https://forum.atmospherictoolbox.org/uploads/default/original/1X/bf3fe3c36e72bde0260f6977363c34179b8b507d.png)  
However, running this code raised error.  
 ![image](https://forum.atmospherictoolbox.org/uploads/default/original/1X/06d845eaaa8d8b3a15936432b597b23173873247.png)  
I also tried running this code without specifying variables in keep() parameter, but resulting L3 files do not contain tropospheric\_HCHO\_column\_number\_density\_avk variable.  
Is there any method to convert tropospheric\_HCHO\_column\_number\_density\_avk variable into L3 grid?  
Thanks in advance.

---

<div class="post-metadata">

### Author: ![sander.niemeijer](https://forum.atmospherictoolbox.org/user_avatar/forum.atmospherictoolbox.org/sander.niemeijer/32/5_2.png) [@sander.niemeijer](https://forum.atmospherictoolbox.org/u/sander.niemeijer)
#### Post date: [December 22, 2021, 7:38am UTC](https://forum.atmospherictoolbox.org/t/keep-averaging-kernel-variables/447/2 "2021-12-22T07:38:24Z")

</div>

You should never regrid averaging kernels. See also [this topic](https://forum.atmospherictoolbox.org/t/using-harp-and-cams-aq-forecast-profile-as-an-apriori-to-derive-tropomi-no2-tropospheric-column/240/2).