Hi,
I wonder if HARP has implemented a de-stripe correction for CH4 as it has for CO (Usecase_6_CO_European_wildfires_2022).
I have read some recent papers where they apply a de-stripe correction based on the Borsdorff et al. (2018) for XCH4 L2 data.
Thank you in advance for your help.
Best,
Diego
Note that HARP does not have any destripe correction built in. The destriping for CO is actually a destriping performed by the processor itself. The CO product contains both ‘normal’ and destriped column values.
For CH4, please check this post.
Thank you Sander for your quick reply and thank you for pointing me to that other post. I’m already using the “ch4=bias_corrected” option but still observe stripes patterns. Sorry for my misunderstanding about thinking that HARP could do destripe correction. I wish it could.
Best,
Destriping product access via HARP now appears to be documented for CH4 at S5P_L2_CH4 — HARP 1.23 documentation. However, when I use HARP 1.23 and the “ch4=corrected” option during ingestion, I get an error even when processing recent data from processor version >= 02.07.00: “invalid value ‘corrected’ for option ‘ch4’ of ingestion module ‘S5P_L2_CH4’”.
The addition seems to be indicated as “S5P L2 CH4” at harp/CHANGES at master · stcorp/harp · GitHub.
Do you have suggestions for further debugging?
Thanks!
The changes were already implemented in the source code repository of HARP, but not included in a new release. This new release was still pending, but we will create the new release today.
Thank you Sander! Your work and responsiveness are greatly appreciated.
Hello Sander! just saw this updated post and quickly tested the new option for destriped ch4: “ch4=corrected”. I get this error, though:
*** harp._harppy.NoDataError: product contains no variables, or variables without data
I confirm I’m using harp 1.24:
harp 1.24 py313r44hdda3348_0 conda-forge
and that the option “ch4=bias_corrected” works just fine.
Thanks!
What product are you trying this on? Note that it will only work on products produced with processor version 2.7.0 or higher.
good point. I have checked the L2 file I downloaded TODAY using copernicus dataspace OData API. It is data from 2019 and I see “:ProcessorVersion = “1.3.2”;”. So, I guess I can confirm that the destripe has not worked on this previous processor version. And I guess that the methane database has not been fully reprocessed and served via this API. Thanks Sander!
