Is it possible to have acquisition time variable for geostationary images like goes and himawari?

[akasom89]

Hi.

As it takes almost 10 minutes (at best case) for a geostationary full disk image to be taken, it might be important for some applications to know the exact time acquisition for the region of interest.

Wanted to know if it is possible to have such layer in Satpy.

Thanks in advance

[akasom89]

I found that time information for some scanlines is available in the metadata of the HRIT format (Himawari data), but I could not find anything similar for GOES ABI.

One simple idea is to interpolate the time linearly, but that does not seem to be accurate since ABI is not a full-disk image over the entire 10 minutes.

I think another idea would be to implement scan mode timeline if possible.

I would appreciate any suggestions.

[simonrp84]

For GOES it is not really possible to attain an accurate scanning timeline in the images, as the meso sectors and cooling sectors mean - as you suggest - there's no possibility to integrate linearly.
The same is true for Himawari, unfortunately.

For MTG, it is more suitable to linearly integrate as it has no meso or continental scan modes.

[gerritholl]

MTG FCI L1C files contain data from which the measurement time per pixel can be calculated with precision to the second. I don't know if ABI or AHI contain this information.

[akasom89]

Thanks.

I found that @ghiggi covered this under his exciting goes_api project.

https://github.com/ghiggi/goes_api/blob/main/tutorials/Others/ABI_pixel_scan_time_with_satpy.py

in the abi_pixel_time.py module, it was mentioned that :

The pixel scan time maximum error is 40 s (at ABI scan swath edges)

[ghiggi]

Hi @akasom89. I took it from the https://www.star.nesdis.noaa.gov/GOESCal/goes_tools.php ABI-Time_Model_LUTS tar file in case ;)

[mraspaud]

@akasom89 it would be great to have times in satpy too…

[pdebuyl]

For Himawari, the "JMA HRIT Mission Specific Implementation" document states:

Image observation time of both the first line and the last line in each image segment will be set and
that of other lines will be additionally set if necessary. Image observation time for all other lines
within each segment can be calculated by linear interpolation from the given values.
Value is expressed with Modified Julian Day (MJD) and is described in ASCII character form for a
real number represented to the sixth decimal place.

(emphasis mine).

So, is it not correct to interpolate within a HRIT segment there?

[gerritholl]

In #3161 / #3168 I am working on an implementation to track time per-pixel for SEVIRI, FCI, ABI, and AHI, with a unified interface, and on adding support for writers to write a representative time for the image. Work is only just starting. Contributions are welcome!