Subject: Calibration Period Posted: 5/18/2015 Viewed: 17125 times
Am a little bit confused about setting up the time frame, current accounts and calibration. To be specific, I want to set up my current account as 2013 and scenarios time frame from 2014 to 2035 but the river gauging data I have is only in the period between 1976 to 1990. How can I set up my project so that I use this river data for calibration? Must the calibration period be within the project time frame?
Ms. Stephanie Galaitsi
Subject: Re: Calibration Period Posted: 5/18/2015 Viewed: 17087 times
It sounds like you'll have to use the earlier period for calibration, since that's when you have the data for. If you're using catchments, I would really focus on getting those right, and also on any legal constraints in place in the system that are driving demand.
You can either extend your model into the current time period or start a new one with the same template (when you create a new area, WEAP will give you the option to create it from a copy of your calibration-period model. Then you can change the years for the new one). Just remember to update information about changing populations, crops, land use patterns. But assuming you've gotten your calibration right, you'll have something to use an indicators of the model's use in predicting system dynamics.
Mr. Jack Sieber
Subject: Re: Calibration Period Posted: 5/18/2015 Viewed: 17083 times
Another way to do it is to initially set the time horizon to 1976-1990, then calibrate all your parameters to get a good fit for your gauging data. Once you have a good fit, you can change the time horizon to 2013-2035, using the calibrated parameters to run the model into the future.
Mr. Robert Ojwang
Subject: Re: Calibration Period Posted: 5/19/2015 Viewed: 17074 times
Hi Stephanie and Jack,
Thanks so much for the very clear advice. I will try to work it out and possibly let you know of the outcome.
One more question related to what I had asked earlier in the forum. Am modelling Roof rainwater harvesting (RRWH) potential as a future scenario (2013-2035) and am still thinking of how best to do it.
I already have the roof areas for different materials (tile, iron, concrete) for 2013 and would like to model it such that losses for each material type are taken care of seperately. Also the roof area I have only represents 'potential' out of which some percentages can be assumed to meet the demand (not all the existing roofs will install RRWH system). Finally, due to population growth, new roofs will be installed and we can assume that all new roofs install RRWH system. Later on I will consider climate change scenario.
What can be the best approach to model this?
Dr. Devaraj de Condappa
Subject: Re: Calibration Period Posted: 5/24/2015 Viewed: 17057 times
I'll attempt to reply to your question on RRWH:
* For 2013 : you can create sub-branches under your catchment object, each branch could be a type of material. And you can enter a specific value of Kc and Effective Precipitation for each branch (ie material).
* For future : I would see 2 options. 1. You "grow" your existing catchment object to account for more RRWH systems. Or 2. You create new catchment object which will represent the future additional RRWH.
If this could help and you'd need more clarification, please don't hesitate to ask.
Mr. Robert Ojwang
Subject: Re: Calibration Period Posted: 5/24/2015 Viewed: 17050 times
Thanks so much for the concise response. Just one question, since the roof losses could just be the material wetting loss which eventually evaporates, which values can be appropriate to model this using Kc and effective prec.? My study area is humid tropical.
An in your other answer am not clear on how to model a diversion? I tried using a reservoir to receive the collected water and transmit it to the City but it doesnt seem to work.
Dr. Devaraj de Condappa
Subject: Re: Calibration Period Posted: 5/25/2015 Viewed: 17003 times
Regarding tuning the Kc and the effective pcp, the value you are entering for effective pcp is the % of precipitation available for evapotranspiration, the remainder being runoff. I would presume that in your case the loss will be small, if not negligible. So your value for effective pcp could be very low (ie very high runoff), maybe around 5% or lower. If I'm correct, only the % you are entering is concerned with Kc value, so it should not matter so much which value you take for Kc, the most important parameter is the % for effective pcp.
I would advise you to try several couple of (effective pcp, Kc) values, to see which one is best.
Regarding the Diversion, I'm referring to the orange object that you can add in your schematic (the one below River). You can add a Reservoir on your Diversion to model tanks receiving the collected rainwater and then connect the Reservoir itself or the downstream part of the Diversion to the City.