Timestamp setting for future climate scenarios

[lucia]

hi，everyone!
I would like to ask you a question. I am currently conducting a simulation of carbon dioxide flux for the entire future climate scenario in China. The input data I am using is cmip6, but the input data I downloaded is on a daily scale and does not have 6-hour scale climate data. could I use daily scale climate data as my input atmospheric forcing data? If so, how should I set my timestamp？

 

[oleson]

The coarsest temporal resolution we usually use to force CLM is 6-hourly. Although I think the datm would accept daily data, I'm not sure how well daily would work in terms of adequately forcing CLM. For example, for a day with a total amount of precipitation, the datm would distribute this evenly over each model time step in the day. Or how to impose a diurnal cycle on the forcing temperature. I'm not sure what would happen with solar radiation either, although there is a solar zenith angle parameterization in the datm associated with that. I don't know what the best time stamp setting would be, you'd have to experiment, unless there is someone else on this Forum who has experience with this.
You might also post in the ocean forum. I wonder if any of their datasets use daily forcing and would be a template for moving forward.

 

[lucia]

Hello,Oleson. Thank you for your answer. I am currently using CLM to run daily scale atmospheric forcing data. I have set the time variable of the atmospheric forcing data from the original 0.0625:0125:1.9375 to 0.5:1:1.5 (assuming only two days of data), but I encountered an error while running the CLM model. I don't know where the problem is, I don't know if there are any variables related to timestamps that need to be changed besides changing the time variable of atmospheric forcing data.
Here are my settings. I feel that I may need to change dlimit or dtime, but I am not sure what these two things are used for. I don't know how to change them. May I ask what I need to change in order for CLM to recognize that my data is on a daily scale and process it




This is my error message：






I have two questions: first, is the value of the time variable in my atmospheric forcing data set correctly？
Secondly, in addition to setting atmospheric forcing data, which parameters of CLM do I need to set so that CLM can recognize that my time is on a daily scale rather than a 3-hour scale, and then enable it to operate successfully

 

[slevis]

I don't know, either, but I see that you set dt_limit = 1.5 in the atm namelist. The error seems to suggest an inconsistency between 1 and 1.5. So the next thing that I would try in the atmosphere's user_nl file is dt_limit = 1, in case it works.

 

[wvsi3w]

The coarsest temporal resolution we usually use to force CLM is 6-hourly. Although I think the datm would accept daily data, I'm not sure how well daily would work in terms of adequately forcing CLM. For example, for a day with a total amount of precipitation, the datm would distribute this evenly over each model time step in the day. Or how to impose a diurnal cycle on the forcing temperature. I'm not sure what would happen with solar radiation either, although there is a solar zenith angle parameterization in the datm associated with that. I don't know what the best time stamp setting would be, you'd have to experiment, unless there is someone else on this Forum who has experience with this.
You might also post in the ocean forum. I wonder if any of their datasets use daily forcing and would be a template for moving forward.

Dear Oleson,
I was reading these links regarding the temporal resolution of forcing data for CLM5 (CESM2.1.3):
link one
link two
link three
link four
and as far as I understood the highest temporal resolution is half an hour and the lowest is 6 hours (with some exceptions regarding daily resolutions).

For the purpose of my project, I should use modified LME (last millennial) as forcing and in the transient run it would be running for 1000 years, so I don't think the institution that is going to give us the data has this high resolution (or even 3h) data available, thus I would end up with a monthly resolution.

Do you think it is possible to run the model with monthly temporal resolution? or if there is a way in the model to adjust it with that?
Thank you in advance.

 

[slevis]

The coarsest temporal resolution we usually use to force CLM is 6-hourly. Although I think the datm would accept daily data, I'm not sure how well daily would work in terms of adequately forcing CLM. For example, for a day with a total amount of precipitation, the datm would distribute this evenly over each model time step in the day. Or how to impose a diurnal cycle on the forcing temperature. I'm not sure what would happen with solar radiation either, although there is a solar zenith angle parameterization in the datm associated with that. I don't know what the best time stamp setting would be, you'd have to experiment, unless there is someone else on this Forum who has experience with this.
You might also post in the ocean forum. I wonder if any of their datasets use daily forcing and would be a template for moving forward.

@wvsi3w I interpret @oleson's response to mean that you can experiment with data of any temporal resolution to see how the model will behave.

 

[wvsi3w]

@wvsi3w I interpret @oleson's response to mean that you can experiment with data of any temporal resolution to see how the model will behave.

Thanks a lot for your response.
I will keep that in mind.
I think there must be a standard way of trying to use monthly data to create 3h (or 6h) files using some standard approach to account for the monthly and daily cycle before putting it in the forcing path for the model.

 

[slevis]

That's possible. I expect you will need to search through the literature to find what you are looking for, unless others have different suggestions.

 

[wvsi3w]

That's possible. I expect you will need to search through the literature to find what you are looking for, unless others have different suggestions.

I appreciate your responses. Yes, of course, I didn't mean to ask about those methods here as those are my job as a student to do.