Statistical prediction of seasonal cyclonic activity over North Indian Ocean
Keywords:Statistical prediction, Cyclonic activity, North Indian Ocean, Correlation, Screening regression, Conditional mean, Multiple regression, Validation, Jackknife method
The Northeast monsoon season of October to December (OND) is the primary season of cyclonic activity over the North Indian Ocean (NIO). The mean number of days of cyclonic activity over NIO during this season is about 20 days. In the present study, statistical prediction for seasonal cyclonic activity over the North Indian Ocean during the cyclone season of October to December is attempted using well known climate indices and regional circulation features during the recent 30 years of 1971-2000.
Potential predictors are identified using correlation analysis and optimum numbers of predictors are chosen using screening regression technique. A qualitative prediction for number of Cyclonic Disturbance (CD) days is attempted by analysing the conditional means of the number of CD days during OND over NIO for different intervals of each predictor based on the 30 year data of 1971-2000. Predictions and their validations for the subsequent test period of 2001 to 2009, based on this scheme, are discussed. An attempt for quantitative prediction is also made by developing a multiple regression model for prediction of number of CD days over the NIO during OND using the same predictors. The regression model accounts for 70% of the inter annual variance. The root mean square error of estimate is 5 days and the bias error is 0.36 days. The regression model is cross validated by Jackknife method for each individual year using the data of 29 years from the sample excluding the year under consideration. The model is also tested for independent dataset for the years 2001 to 2009. Salient features of the model performance are discussed.
How to Cite
Copyright (c) 2021 MAUSAM
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
All articles published by MAUSAM are licensed under the Creative Commons Attribution 4.0 International License. This permits anyone.
Anyone is free:
- To Share - to copy, distribute and transmit the work
- To Remix - to adapt the work.
Under the following conditions:
- Share - copy and redistribute the material in any medium or format
- Adapt - remix, transform, and build upon the material for any purpose, even