Subject: Calibration and Validation of WEAP Posted: 1/2/2017 Viewed: 21200 times
Dear Members.
Member am planning to use WEAP for a certain project but when I read through am failing to get the calibration and validation procedures. Is there any one who can help me on that??.Or any who who had ever calibrated it?.
Ms. Stephanie Galaitsi
Subject: Re: Calibration and Validation of WEAP Posted: 1/3/2017 Viewed: 21170 times
Dear Bogere,
Can you say a little bit more about the structure of the model? Have you entered flow into the river, or are you using catchments to model flow? Are you calibrating to observed gauge data for the river, or some other observed data? How many years of observed data do you have? How many stations?
Bogere Robert
Subject: Re: Calibration and Validation of WEAP Posted: 1/7/2017 Viewed: 21094 times
Thanks Stephanie for the quick response.
As I told you this my first time to use this model but my intention is use to it to allocate water resources in my study area. So my question is Do i Have to calibrate the model first and if so what are the procedures for calibrating it.
Ms. Stephanie Galaitsi
Subject: Re: Calibration and Validation of WEAP Posted: 1/9/2017 Viewed: 21082 times
Hi Bogere,
Before you can calibrate the model, you will need to enter all the demands and supplies. In the sense, calibrating the historic model is really the last step in building the historic model. You can build the scenarios before or after the calibration process (since they are looking into the future), as long as they inherit the calibrated data.
Eng. Jared Okungu
Subject: Re: Calibration and Validation of WEAP Posted: 1/13/2017 Viewed: 21062 times
This is a good discussion. Kindly assist with the callibration procedure
Ms. Stephanie Galaitsi
Subject: Re: Calibration and Validation of WEAP Posted: 1/13/2017 Viewed: 21057 times
The calibration procedure depends on the structure of your model. Are you using the water year method, or catchments? Please ask specific questions and I will be happy to assist you.
Mr. Ephraim Sekyi-Annan
Subject: Re: Calibration and Validation of WEAP Posted: 4/24/2017 Viewed: 20410 times
Dear Stephanie,
I have a question about the validation of the WEAP model. I modeled the water balance of a non-river reservoir. Due to the unavailability of inflow data, the Rainfall-runoff (Simplified Coefficient) Method was used to generate surface runoff into the reservoir from the catchment upstream. I calibrated the model with data on the water demand of the multiple water users and compared the simulated reservoir storage with the observed time-series reservoir storage. The Water Year Method was then used to forecast future water availability/unmet demands based on the current account year. After which I run the model under different scenarios. Please, is there a procedure for validation of the model after the procedure I have described above, or once the model has been successfully calibrated it is enough to run scenarios. Thank you.
With regards,
Ephraim
Ms. Stephanie Galaitsi
Subject: Re: Calibration and Validation of WEAP Posted: 4/25/2017 Viewed: 20394 times
Hi Ephraim,
Your calibration with the observed time-series storage volume from the past is valid.
However, you can't mix catchments and the water year method. If you are using a catchment (simplified coefficient method) to model the flow coming into the reservoir, the Water Year Method will not work for the catchment.
I'm not sure if you have a run-off infiltration link going directly into the reservoir, or if the catchment feeds a river, but either way, the Water Year Method won't impact the data.
But actually, if you're asking about validation, that is a historic process, so your future predictions wouldn't matter either way. A validation period is a historic period with historic data that is not used as part of the calibration period. So you would pick a span of time for calibration, calibrate the model, and then see how that works for another [historic] time period, which is the validation period (validation of the calibration). So in that sense, you can use the same process of plugging in your historic data into your catchment that feeds the reservoir, and then comparing the reservoir storage modeled in WEAP with the historic storage values.
Mr. Ephraim Sekyi-Annan
Subject: Re: Calibration and Validation of WEAP Posted: 4/25/2017 Viewed: 20388 times
Dear Stephanie,
Thank you for the response. Yes, you are right. Initially, I realized the Water Year Method did not impact on the inflow data which were generated by the catchment method (through a runoff infiltration link). And yes, the surface runoff from the catchment flowed directly into the reservoir. So what I did to overcome the challenge with the Water Year Method was that I exported the data to excel and then imported the excel file back to WEAP and subsequently deactivated the runoff infiltration link. This time round, the Water Year Method impacted on the inflow data. I then calibrated the model by adjusting the inflow data (i.e. dividing the dataset with a single factor) until the simulated time-series reservoir storage data matched with the observed data for the Current Account Year.
Concerning the validation, there were no historical inflow data for the reservoir, hence a validation of the model with such data was not possible in my case. But I hope my simulated results are still valid without the validation, as the scenarios represent future events?
Could you, please, recommend your recent publications or others' on this topic of validation? Thank you.
With regards,
Ephraim
Ms. Stephanie Galaitsi
Subject: Re: Calibration and Validation of WEAP Posted: 4/25/2017 Viewed: 20382 times
Hi Ephraim,
You say you have observed data for the Current Account Year for the reservoir storage - but do you have that data for other (earlier?) years? It's the historic data that will allow you to do either calibration or validation. Without historic data, you can do neither. It sounds like rather than "calibrating" your model, you have used a trial-and-error process to enter the correct *data* (inflow into the river headwater) for a single year.
Calibration involves looking at the results produced and making sure that those are correct against the same metrics historically (streamflow, reservoir storage volume, snowpack, unmet demand, to name a few), and the results should be compared over several years. If you don't have the historic data, you can't really say that you calibrated your model.
Are your simulated results still valid? Well you're using the Water Year Method to examine the future, which itself is conjecture. Depending on how complicated your model is, maybe having the correct relationship between inflow and reservoir operation is enough for the results you're trying to see. But with only one year of data, I think it would be hard to make that claim definitively.
In terms of sources, the WEAP publications (www.WEAP21.org/publications) has a list of recent WEAP publications, but these tend to report the results of calibration (such as the Nash-Sutcliffe), not the process. I would also caution you that you should start with calibration - only a calibrated model can be validated (not the other way around).
Mr. Ephraim Sekyi-Annan
Subject: Re: Calibration and Validation of WEAP Posted: 4/25/2017 Viewed: 20375 times
Dear Stephanie,
Thank you very much for the explanation. It was very useful.