Click for new scientific resources and news about Corona[COVID-19]

Paper Information

Journal:   JOURNAL OF ENVIRONMENTAL SCIENCE AND TECHNOLOGY   FALL 2007 , Volume 9 , Number 3 (34 SPECIAL ISSUE); Page(s) 185 To 202.
 
Paper: 

ESTIMATION OF MAXIMUM DAILY AIR TEMPERATURE USING ARTIFICIAL NEURAL NETWORK AND NOAA DATA (CASE STUDEY: URMIAH LAKE BASIN

 
 
Author(s):  BANIHABIB M.E., RAHIMI KHOUB A.*, ARABI A.
 
* 
 
Abstract: 

Air temperature is measured in synoptic and climatologic stations. It is one of descriptive parameter of climate of the earth. In this research artificial neural network is used to estimate maximum daily temperature of weather in Urmiah lake basin. Artificial neural network is trained to determine relation between maximum daily temperature and input data. Input data are Land surface temperature and NDVI which are determine using NOM data. Other input data are longitude, latitude, altitude, day of year and time of NOM image. Multilayer precipitation artificial neural network with one, two and three layer are trained and tested. Three back propagation learning rules including Delta rule, Normalize cumulative delta and extended delta-bar-delta are tested. Also two transfer function including Sigmoid and Tangent hyperbolic are tested. Root mean square error (RMSE) is used for comparison of these artificial neural networks. The best artificial neural network has three hidden layer including 4, 3 and one neuron in first, second and third hidden layers respectively. Extended delta-bar-delta gives best learning. The selected artificial neural network has less RMSE with Tangent hyperbolic transfer function. Training RMSE is 0.154 C which is less than 1.57 C of multivalute regression. Therefore suggested artificial neural network can estimate maximum daily temperature of weather better than multivalute regression.

 
Keyword(s): MAXIMUM DAILY AIR TEMPERATURE, ARTIFICIAL NEURAL NETWORK, SATELLITE NOAA
 
References: 
  • ندارد
 
  Persian Abstract Yearly Visit 80
 
Latest on Blog
Enter SID Blog