All Topics |
Topic: “Random variation in climate variables”
Author | Message |
---|
Subject: Random variation in climate variables
Posted: 7/13/2012 Viewed: 28667 times
Hello,
I am simulating streamflow in a sub-basin of the Volta basin, and I am trying to simulated random variation in climate, using the Random function in WEAP. I am trying to prescribe Monthly bounds for this random variation. For example, for temperature, each month has a different set of low and high bounds in which to randomly vary. Does anyone know how to use expression builder, or something else, to develop this monthly random variation? Thank you, Patrick |
Subject: Re: Random variation in climate variables
Posted: 7/13/2012 Viewed: 28639 times
The form of the Random function that you want is Random(LowerBound, UpperBound), where the lower and upper bounds will vary by month.
If the lower temperature bound in each month is: MonthlyValues( Jan, 0, Feb, 10, Mar, 20, Apr, 40, May, 50, Jun, 60, Jul, 65, Aug, 65, Sep, 55, Oct, 45, Nov, 35, Dec, 15 ) and the upper temperature bound in each month is: MonthlyValues( Jan, 40, Feb, 50, Mar, 60, Apr, 70, May, 75, Jun, 80, Jul, 85, Aug, 85, Sep, 75, Oct, 70, Nov, 60, Dec, 50 ) Then the expression to pick a random value in each month between the lower and upper bounds is: Random( MonthlyValues( Jan, 0, Feb, 10, Mar, 20, Apr, 40, May, 50, Jun, 60, Jul, 65, Aug, 65, Sep, 55, Oct, 45, Nov, 35, Dec, 15 ), MonthlyValues( Jan, 40, Feb, 50, Mar, 60, Apr, 70, May, 75, Jun, 80, Jul, 85, Aug, 85, Sep, 75, Oct, 70, Nov, 60, Dec, 50 )) Note, however, that the Random function will give a uniform distribution of values between lower and upper bounds. This will mean that the lower and upper extremes will occur with the same frequency as the average. Depending on your choice of bounds (e.g,. historical min and max temperatures), this may give very unreasonable (extreme) values. Jack |
Subject: Re: Random variation in climate variables
Posted: 7/13/2012 Viewed: 28628 times
Thanks Jack,
I appreciate the quick reply, and this is very helpful. The bounds refer to the 75th quantile of a dataset, so it is necessarily sampling the upper range of the data. This is good to know about the uniform distribution as this may explain odd results when I run the simulation. At the very least however, knowing the distribution is important to understanding the assumptions I'm making. -Pat |
Topic: “Random variation in climate variables”