|
NISCAIR ONLINE PERIODICALS REPOSITORY (NOPR) >
NISCAIR PUBLICATIONS >
Research Journals >
Indian Journal of Geo-Marine Sciences (IJMS) >
IJMS Vol.37 [2008] >
IJMS Vol.37(1) [March 2008] >
| Title: | Simulation of Antarctic sea ice area with artificial neural network |
| Authors: | Tripathi, K. C.
Das, I. M. L. |
| Keywords: | Artificial neural network
Antarctic sea ice
Fourier analysis
Activation function
Sea ice
Simulation
Forecast
Error back propagation |
| Issue Date: | Mar-2008 |
| Publisher: | CSIR |
| Abstract: | The artificial neural network (ANN) has been used for simulation of Antarctic sea ice area anomalies. Various dominant cycles present in the data have been identified using the Fourier analysis. It has been found that the data of the Antarctic sea ice area has two dominant cycles: annual and half yearly. The effect of the presence and / or absence of these dominant cycles on the simulation results have been carried out. ANN can simulate the broad trend of the sea ice area anomalies when all the cycles are present. However, the prediction skill of model for intraseasonal variability degrades as we remove the trends. Further, the forecast have been verified on the basis of various attributes of the forecast. |
| Page(s): | 77-85 |
| ISSN: | 0379-5136 |
| Source: | IJMS Vol.37(1) [March 2008]
|
|
