Rainfall simulation is key problem for the evaluation of the energy saving potential of passive cooling by natural rainfall evaporation in humid and hot areas. This paper used a set of monthly, daily and hourly rainfall data in Guangzhou Area to identify the statistic feathers and distribution of the monthly, daily and hourly rainfall. A one-order AR model was found to be suitable for monthly rainfall estimation by AIC criterion judgment. Using a Gamma distribution model and monthly rainfall, daily rainfall was estimated by stochastically assignment. Using a Beta distribution model and daily rainfall, hourly rainfall was estimated by stochastically assignment. A comparison of statistics features of simulated results to that of measured results validates the estimation model.
Rainfall predictive models for building simulation II –rainfall estimation
Year:
2007
Bibliographic info:
Building Simulation, 2007, Beijing, China