Maximum and minimum temperature prediction over western Himalaya using artificial neural network

Authors

  • PIYUSH JOSHI Snow and Avalanche Study Establishment, Defense Research and Development Organization, Chandigarh - 160 036, India
  • A. GANJU Snow and Avalanche Study Establishment, Research and Development Centre, DRDO, Chandigarh (U.T.) – 160036, India

DOI:

https://doi.org/10.54302/mausam.v63i2.1423

Keywords:

Avalanche, ANN, WD, Back propagation, Mountain meteorology

Abstract

Due to eastward moving synoptic weather system called Western Disturbance (WD), Western Himalaya receives enormous amount of precipitation in the form of snow during winter months (November to April). This precipitation keeps on accumulating and poses an avalanche threat. Temperature plays an important role for the initiation of avalanches. Therefore, prediction of maximum and minimum temperature may be quite helpful for avalanche forecasting. In the present study Artificial Neural Network (ANN), a non-linear method is used for the prediction of maximum and minimum temperature using surface meteorological data observed at various observatories in Western Himalaya region. ANN provides a computational efficient way of determining an empirical possible non-linear relationship between a number of input and one or more outputs. In present study back propagation learning algorithm is used to train the network. In the training process the relationship between input and output is extracted i.e., final weights are computed. Past data of about 25 years is used for training the network and trained network is used for temperature prediction for five winter seasons (2005-06 to 2009-10). Root mean square errors (RMSE) corresponding to maximum and minimum temperature are computed. For independent data set RMSE vary from 2.18 to 2.48 and 1.99 to 2.78 for maximum and minimum temperatures respectively.

Downloads

Published

01-04-2012

How to Cite

[1]
P. . JOSHI and A. . GANJU, “Maximum and minimum temperature prediction over western Himalaya using artificial neural network”, MAUSAM, vol. 63, no. 2, pp. 283–290, Apr. 2012.

Issue

Section

Shorter Contribution