search

holos-aafc / Holos

Open source version of the core algorithms used by the Holos model
Other
21 stars 21 forks source link

(ICBM) 'Growing season precipitation/evapotranspiration' output in Multiyear_Results.csv output file #178

Closed L-HEREDIA-LOGIAG closed 2 weeks ago

L-HEREDIA-LOGIAG commented 1 month ago

Contact Details

ludovic.heredia@logiag.com

Expected Behaviour

Hello Holos Teams.

We have a question regarding simulation results from the ICBM module. In the multiyear_result_test.cvs, there are two outputs: 'Growing season precipitation (mm year^-1)' and 'Growing season evapotranspiration (mm year^-1)' which gives same results for different years in our tests. For example 632.902 mm for 1985 to 1987, then 422.26 mm, then it cycles back to 632.902 mm.

However, from our understanding, Holos 4.0 extracts daily climatic data from NASA database so we are expecting different values for these outputs, for every year. Do you mind to give us some explanations why those output do not give what we are expecting?

I have attached a .csv file with this report, hope it's available for you. Thank you a lot again, Ludovic

Multiyear_Results_test.csv

Actual Behaviour

These variables can give the same values for severral years

Describe the Issue & Steps to Reproduce

  1. Simulate corn grain on one field since 1985
  2. Create a multiyear_results.csv file when analysing ICBM related results

Error Message

No error message, unexpected behaviour

Holos Version Number

4.0 (last version is automatically downloaded when launched)

Code of Conduct

holos-aafc commented 1 month ago

Could you export the farm file and send to the Holos email account or attach here if possible.

Thanks

L-HEREDIA-LOGIAG commented 1 month ago

LOGICBM REV04-ANNUAL_CROP_NO_20232024-SOIL_ADJUSTED.json

holos-aafc commented 3 weeks ago

Hi @L-HEREDIA-LOGIAG . Holos will download daily climate data from NASA for the farm's location. We have found that sometimes NASA does not return a full set of daily climate data - sometimes there are days missing from the year in question. When NASA returns a set of values representing any particular year, and there is data missing from that set, Holos will revert/fall back to using annual climate normals instead. This is what is happening on this farm and why you see repeating values for some years.

I am open to suggestions on how to improve the default response to these scenarios if you have any feedback.

The section of code that handles this is found here on line 190:

https://github.com/holos-aafc/Holos/blob/2c4b6afc63de91a266eb916e62f9ed71591c716e/H.Core/Providers/Climate/ClimateData.cs#L190

Thanks

holos-aafc commented 3 weeks ago

There is another issue with the climate data that has just been fixed. Sometimes NASA is returning data for 366 days in some years and this was not being accounted for in Holos. An update will be made available to fix these issues later this week. I will include more information on the fix when the update is ready

holos-aafc commented 2 weeks ago

Hi @L-HEREDIA-LOGIAG . This issue has been fixed with #d43e527 . The GUI will download this update when you restart Holos.

https://github.com/holos-aafc/Holos/blob/main/Changelog.md

Thank you for reporting this issue.