2025 Orange County Water Demand Projection Model
100 150 200 250 300 350 400 450 500 550
Observed
Modeled
0 50
Figure 3-2: Modeled vs Historical Water Use per Account for Single-Family for One Agency
The R 2 values for all retail agencies are all above 80%, which indicates that the explanatory variables used in the model as well as their multipliers (model coefficients) explain 80% or more of the historical variability in the single-family use per account measurements.
3.4 Multifamily Regression Development
This section reviews the development of the statistical regression model for the multifamily residential sector.
3.4.1
Explanatory Variables and Fitted Coefficients
Most explanatory variables for the multifamily sector are the same as for the single-family sector. Median income and a 2-month lagged departure from precipitation were dropped. These modifications to the model design resulted in stronger measures of fit and more reasonable coefficient estimates. Final coefficient estimates presented in Table 3-4 are within the expected range for all explanatory variables
3-9
Appendix G - 46
Powered by FlippingBook