Health & Environmental Research Online (HERO)


Print Feedback Export to File
1608405 
Journal Article 
Estimation of Pan Evaporation using neural networks and Climate-based models 
Kim, S; Park, KiBum; Seo, YMin 
2012 
Yes 
Disaster Advances
ISSN: 0974-262X 
5 
3 
34-43 
English 
WOS:000311501400005 
The accuracy of the neural networks models for estimating the daily pan evaporation (PE) is investigated in this paper. The purpose of this paper is to develop and aly the neural networks models to estimate the daily PE, Republic of Korea. Two kinds the neural networks models such as multilayer perceptron neural networks model (MLP-NNM) and co-active neuro-fuzzy inference system model (CANFISM) for two weather stations, Daegu and Ulsan, were used to estimate the daily PE. The climate variables such as extraterrestrial radiation (R-a), sunshine duration (SD), mean temperature (T-mean), mean relative temperature (RHmean) and mean wind speed (U-mean) were used to estimate the daily PE using the various input combinations. Penman method was used to compare the performance results of the neural networks models. Therefore, based on the comparisons, it was found that the neural networks models can be employed successfully for estimating the daily PE from the climatic data available, Republic of Korea. 
Pan Evaporation; MLP-NNM; CANFISM; Penman Method