Hello there,
I would like to run a simulation using a weather file that contains weather data from five consecutive years.
If we assume for example that I need three years to achieve dynamic equilibrium using a single year weather file, does this mean that I will ned to run the simulation for 5*3=15 years in order to achieve dynamic equilibrium?
Thanks in advance,
Regards,
Kostas
Multiple-year weather file
Re: Multiple-year weather file
Hi Kostas,
dynamic equilibrium will be achieved after three years or so, but you will have to run the simulation for 3*5 years to prove it.
Dynamic equilibrium has been achieved when the simulated component has "forgotten" the initial state. That is, when the current state is only determined by the recent boundary conditions, and all influence from the initial state has died away.
If the component is exposed to an infinitely repeated one-year climate, its hygrothermal state repeats year after year once dynamic equilibrium has been reached. This allows you to determine that dynamic equilibrium has indeed been reached.
If the component is exposed to a repeating 5-year climate sequence, its "memory" will be the same and it will have forgotten the initial state after about the same time as before. However, because the five years are now different among them, the hygrothermal state can only repeat every five years (even if dynamic equilibrium has been reached much earlier).
So if you want to prove that dynamic equilibrium has been reached, you will have to run the simulation at least into a major part of the second five-year climate cycle to have enough data for comparison. Comparing the beginning of the second cycle with the beginning of the first cycle will fail because equilibrium was not yet reached at the beginning of the first cycle.
If equilibrium has been reached during the first cycle, the states at the end of the second cycle will be identical with those at the end of the first cycle. But the agreement will develop gradually and will be difficult to evaluate with some automated script.
If you run through three climate cycles, two of which are unaffected by the initial state, comparing these two will be easy and can be done in an automated way. But it means you have to go through three complete cycles (one affected, two unaffected).
If you are comparing the results by eye or with some semi-sophisticated script, running through two cycles will be sufficient, but you'll have to take into account that parts of them will not be identical and you'll have to find the point from which on they agree sufficiently.
Regards,
Thomas
dynamic equilibrium will be achieved after three years or so, but you will have to run the simulation for 3*5 years to prove it.
Dynamic equilibrium has been achieved when the simulated component has "forgotten" the initial state. That is, when the current state is only determined by the recent boundary conditions, and all influence from the initial state has died away.
If the component is exposed to an infinitely repeated one-year climate, its hygrothermal state repeats year after year once dynamic equilibrium has been reached. This allows you to determine that dynamic equilibrium has indeed been reached.
If the component is exposed to a repeating 5-year climate sequence, its "memory" will be the same and it will have forgotten the initial state after about the same time as before. However, because the five years are now different among them, the hygrothermal state can only repeat every five years (even if dynamic equilibrium has been reached much earlier).
So if you want to prove that dynamic equilibrium has been reached, you will have to run the simulation at least into a major part of the second five-year climate cycle to have enough data for comparison. Comparing the beginning of the second cycle with the beginning of the first cycle will fail because equilibrium was not yet reached at the beginning of the first cycle.
If equilibrium has been reached during the first cycle, the states at the end of the second cycle will be identical with those at the end of the first cycle. But the agreement will develop gradually and will be difficult to evaluate with some automated script.
If you run through three climate cycles, two of which are unaffected by the initial state, comparing these two will be easy and can be done in an automated way. But it means you have to go through three complete cycles (one affected, two unaffected).
If you are comparing the results by eye or with some semi-sophisticated script, running through two cycles will be sufficient, but you'll have to take into account that parts of them will not be identical and you'll have to find the point from which on they agree sufficiently.
Regards,
Thomas
Re: Multiple-year weather file
Hi Thomas,
I see, many thanks for the detailed answer.
Regards,
Kostas
I see, many thanks for the detailed answer.
Regards,
Kostas