Page 1 of 1

Running with Different Timesteps

Posted: Fri Apr 26, 2019 6:45 pm
by sankee82
Hello Everyone,

Is it possible to run one part of my model on daily time step and the other part of it on yearly time step?

Thanks and Regards,
Sankeerthana

Re: Running with Different Timesteps

Posted: Fri Apr 26, 2019 7:01 pm
by tomfid
No (except perhaps for some complex tricks). What are you trying to do?

Re: Running with Different Timesteps

Posted: Sat Apr 27, 2019 5:01 am
by sankee82
In the model attached, I want to run the hydrological parameters on daily timestep as I have daily data. Whereas, I want to calculate the crop yield at the end of each year. Can you please help me with it? Can the functions related to DELAY be of any help?

Re: Running with Different Timesteps

Posted: Sat Apr 27, 2019 1:17 pm
by tomfid
You don't need different time steps. Run daily, and accumulate the results for annual reporting. See viewtopic.php?f=2&t=7199

Re: Running with Different Timesteps

Posted: Sat May 11, 2019 1:48 pm
by sankee82
Hello everyone!

In the model attached, I have tried to accumulate the crop yield on annual basis after every 365 days. But from my observation the annual crop yield reduction is being calculated using every 365th value of crop yield reduction. Can you please help me in solving this? Also instead of accumulating crop yield I rather want to accumulate AET and PET annually and from there calculate the crop yield, Is that feasible? Any other suggestions to correct the model are also appreciated.

Thanks and Regards,
Sankeerthana.

Re: Running with Different Timesteps

Posted: Sat May 11, 2019 8:05 pm
by tomfid
Without units and longer variable names, it's hard to know how to interpret AET and PET here.

Re: Running with Different Timesteps

Posted: Sun May 12, 2019 5:12 am
by sankee82
Thank you fro the response. AET is Actual Evapotranspiration and PET is Potential Evapotranspiration whose units are m/day. In the attached model, I have corrected the units for all the variables except farm balance since I am confused about how to define those variables. Can you please help me with this and previously mentioned problems?

Thanks and Regards,
Sankeerthana.

Re: Running with Different Timesteps

Posted: Sun May 12, 2019 4:39 pm
by tomfid
The units don't balance on previous farm balance because the time constant (1) needs to be a dimensioned parameter.

I think the issue with the farm balance is that it should have units of Rs/ha, which means that income should be Rs/ha/day. Since income and yield are flow concepts, they should have time in the denominator.

Re: Running with Different Timesteps

Posted: Mon May 13, 2019 6:19 am
by sankee82
Thank you for the response. Can you please also help me with why the annual crop yield reduction is being calculated using every 365th value of crop yield reduction whereas I want to accumulate the crop yield on annual basis after every 365 days. Also instead of accumulating crop yield I rather want to accumulate EvapoTtanspiration and Potential EvapoTranspiration annually and from there calculate the crop yield, Is that feasible?

Regards,
Sankeerthana.

Re: Running with Different Timesteps

Posted: Mon May 13, 2019 9:24 am
by Administrator
sankee82 wrote: Mon May 13, 2019 6:19 am Thank you for the response. Can you please also help me with why the annual crop yield reduction is being calculated using every 365th value of crop yield reduction whereas I want to accumulate the crop yield on annual basis after every 365 days. Also instead of accumulating crop yield I rather want to accumulate EvapoTtanspiration and Potential EvapoTranspiration annually and from there calculate the crop yield, Is that feasible?
We cannot run the model to investigate (it's missing the source data).

I cannot find a parameter in your model called "EvapoTanspiration", can you clarify which variable you are interested in?

Re: Running with Different Timesteps

Posted: Mon May 13, 2019 12:21 pm
by sankee82
It is not EvapoTranspiration, it is Actual Evapotranspiration. I am also attaching the file that has source data.

Thank you!

Re: Running with Different Timesteps

Posted: Tue May 14, 2019 8:01 am
by Administrator
sankee82 wrote: Mon May 13, 2019 6:19 amCan you please also help me with why the annual crop yield reduction is being calculated using every 365th value of crop yield....
I'd suggest trying to keep your equations a little more simple.
Introduce a new variable (I'm not sure what it should be called, so I've named it "aux"),
aux = actual evapotranspiration/potential evapotranspiration

Now change
yield reduction = YIELD RESPONSE FACTOR*(1-(aux))/TIME STEP

Look at the values for "aux" and you will see why you are not getting the behaviour you want.
I rather want to accumulate EvapoTtanspiration and Potential EvapoTranspiration annually and from there calculate the crop yield, Is that feasible?
Use a level to accumulate these.

Re: Running with Different Timesteps

Posted: Tue May 14, 2019 1:35 pm
by tomfid
Two issues here:

1. I think the model would be clearer with more physical variable names and concepts. For example, "yield reduction" is not a thing. You want to model the physical state, which is - I think - the amount of crops in the field.

2. I'm not sure why TIME STEP is in the denominator for Yield Reduction. Generally, TIME STEP should not figure in equations. The only exception is when you want to measure an instantaneous derivative or model a discrete event. This may be the latter, but I'm not sure.

Re: Running with Different Timesteps

Posted: Tue May 14, 2019 1:38 pm
by tomfid
Combining these, I think you want something like the following:
crops.png
crops.png (14.19 KiB) Viewed 6488 times
Growth will be a continuous process, something like:

Growth = Effect of Water Availability * Crops * Normal Growth Rate * (1 - Crops/Max Crops)

Then Harvest is a discrete event:

Harvest Trigger = PULSE TRAIN( ... )
Harvest = Harvest Trigger*Crops/TIME STEP

Re: Running with Different Timesteps

Posted: Tue May 14, 2019 2:06 pm
by sankee82
Thank you so much for the response. To simulate Harvest as annual discrete event, I am using a reporting interval of 365 days. But I am running the model for 20 years. How should I account for leap years which have 366 days and give reporting interval as 366 days for those particular years?

Re: Running with Different Timesteps

Posted: Tue May 14, 2019 2:17 pm
by tomfid
My inclination would be to ignore such details - just make every year 365.25 days.

Otherwise, you could create a schedule of leap years in a lookup.

Re: Running with Different Timesteps

Posted: Tue May 14, 2019 2:20 pm
by sankee82
Thank you!