rainfall_onset2 on Ghana data about rainfall_onset2 HOT 4 CLOSED

Dear @WinnieWin ,

What type of difficulties did you encounter? Are you working with gridded data or station data? If you can share snippets of the difficulties, we can take a look at it together.

Thank you.

from rainfall_onset2.

from rainfall_onset2.

Dear @WinnieWin and @ogunjosam

Sorry for my late reply
I have already checked the code and I uploaded a new one (https://github.com/adrHuerta/rainfall_onset2/blob/master/example_gh.py) using the example of @WinnieWin

Regarding the issues for both, I found the following

• @WinnieWin as in your study area the rainfall season is within a year, is not needed to set the hydrological year
• @ogunjosam I added a new function for getting the numeric values of the onset/cessation for each year (I add some comments in the code). It begins on 1st January for each year
• Some issue that I found is that some parts of Ghana have two rainfall seasons (southern areas) - This is something that the code can not handle. So be careful with that.
  Look at this time series:
  For a grid where there is one single rainfall season
  
  For a grid where there are two rainfall seasons
  
• Another issue is that is possible to find some bad grids where the seasonality can not be well resolved by the gridded data - If you look at the onset/cessation for each year you will notice that some values present atypical values (based on its neighboring grids). This is related to the development of the gridded datasets and can be seen as a quality control process (you can compare it with other satellite-based precipitation estimation)

Finally, I would like to mention how to handle the last two points

• Remove grid areas where is possible to find two rainfall seasons
• Rather than applying the onset/cessation analysis at the grid scale, it would be better to apply it at the area scale (or meteorological station scale).

Regards
Adrian

from rainfall_onset2.

from rainfall_onset2.