I am building a simple model and trying use the material delay element to delay flow from Jan through March to arrive on April 1 (day 91). I have placed the below if statement in my material delay element. However when i look at the results, I see the flow rate start at 57d to some unknown value, and then drop down to 100 m3/d at 91d as i expected. Why is there a flow rate before 91d? I assume something is wrong in my formula. Please help.
Flow rate is constant = 100m3/d
It looks like you are causing an unintentional build-up of "Amount in Transit" material in the material delay. If you plot the "Amount_in_Transit" output of the Material Delay, you will see that the amount is building up and when finally released, causes your unexpected outflow.
When you say you want to delay flows until April 1, can you tell me how you want to distribute the amount that is waiting in transit during the months of Jan-Mar? Do you want this entire volume to be thrown out all at 1 time step or do you want it to distribute over a long period of time, like all of April or over the next few months? Somehow, it needs to be defined how you want to deal with the amount that you are holding back when delaying begins. If you tell me more about how you want to control the flows, we can give recommendations.
-Jason
I can't reproduce what you have until knowing the start date of your simulation. The simulation is running as elapsed time but since you are referring to the "DateTime" run property, it actually is using a calendar time basis to calculate your delay. You might want to consider changing to a calendar simulation. Once I know what your start date is, I can probably answer your question about day 57.
But before we get into this, I need to know a little more about what process you are simulating. If you are trying to simulate a process that accumulates an inflow until April 1, I think you might want to use a Pool or Reservoir instead of a Delay element. It is unusual to use a Delay element like this. Are you simulating the process of water being stored and then released on April 1?
Hi Jason, the start time is Jan 1 and run for 365 days. i have changed the simulation time from elapsed time to calendar time.
I have considered using a pool/reservoir element but i thought i would try to see if i can use the delay element. For example if I could calculate snowmelt each day but then delay its flow into a basin/pool element until April. I am still learning GoldSIM so this is not a real problem.
Okay I see. I recommend that you use this example model for estimating runoff from snow melt: Snow Accumulation and Melt Estimator – GoldSim Help Center. Please let me know if this might work for you. In your case, you can adjust the "snowmelt distribution" table so that all the snow melts by April 1. If this doesn't work for you, please let me know. We have more sophisticated snow model examples here:
Ashley,
I looked at your model and after studying your delay time calculation, I found that you are specifying a dynamic delay time approaching zero, which is unusual. Put your delay time equation into a separate equation, then you can plot it as I have done:
The Material Delay element operates like a conveyor belt, where you put material on the front and after a delay time, this same pile of material comes off the other end. To account for a delayed flow that is dispersed in both space and time common in hydrological scenarios, you can use the Dispersion input of the Material Delay.
I think you should use the "Dispersion" input of the Material Delay element. Keep the delay time constant for now. Please have a look at Dispersion: Modeling Material Delays with Dispersion (goldsim.com) for more information. Below is a screen capture of a new Material Delay I added, that you could try on your end. Notice the result over time.
Statistical dispersion can be a valuable concept for modeling hydrological processes, such as forecasting daily snowmelt rates. Dispersion accounts for the variability over time and space, which is a critical aspect of hydrological phenomena. The use of dispersion in material delay elements allows for the simulation of variability in transit times. Honestly, I have not seen any journal articles that back this claim up so please let me know if you find something. It seems like dispersion could be used in a very rough, conceptual way.
Ultimately, you will want to consider using one of the hydrological examples we have or a different tool. Hydrology – GoldSim Help Center
I hope this helps,
Jason
I forgot to mention why you have the result shown below.
Imagine you start loading a conveyor with sand. While it is running along the conveyor, you get a shovel and start running faster than the conveyor, pushing the sand forward making a growing pile above the sand in front of you on the conveyor. This isn't the best analogy but essentially you are changing how delayed in time as time progresses. Instead, please consider using the "Dispersion" input as I think this is exactly what you need.
Please sign in to leave a comment.
Comments
8 comments