<b>PENDUGAAN FLUKS PANAS DAN EVAPOTRANSPIRASI DENGAN JARINGAN SYARAF TIRUAN</b><br><i>HEAT FLUX AND EVAPOTRANSPIRATION ESTIMATION USING ARTIFICIAL NEURAL NETWORK</i>

Abstract

<em>Artificial neural network (ANN) approach was used to model energy dissipation process into sensible heat and latent heat (evapotranspiration) fluxes. The ANN model has 5 inputs which are leaf temperature T<sub>l</sub>, air temperature T<sub>a</sub>, net radiation R<sub>n</sub>, wind speed u<sub>c</sub> and actual vapor pressure e<sub>a</sub>. Adjustment of ANN was conducted using back propagation technique, employing measurement data of input and output parameters of the ANN. The estimation results using the adjusted ANN shows its capability in resembling the heat dissipation process by giving outputs of sensible and latent heat fluxes closed to its respective measurement values as the measured input values are given. The ANN structure presented in this paper suits for modeling similar process over vegetated surfaces, but the adjusted parameters are unique. Therefore observation data set for each different vegetation and adjustment of ANN are required.</em>