Please use this identifier to cite or link to this item:
https://dl.ucsc.cmb.ac.lk/jspui/handle/123456789/4755
Title: | A Study of Monsoon Cloud Dynamics Using Linear Symmetry |
Authors: | Ramanayake, S. |
Issue Date: | 2019 |
Abstract: | Abstract The e ectiveness of the linear symmetry to reveal the hidden information of the natural images is presented along with a novel approach based on the cloud dynamics to predict the arrival of South-West monsoon. The linear symmetry which has been used as an e ective feature on many applications, is used to determine the cloud shape stability life time and cloud optical motions of the satellite images. Cloud images in thirty minute intervals were downloaded from the relevant site and 650E 900E longitudes and 00 250N latitudes areas were extracted for this study. The cloud shape stability begins to drop in April-May and remains in that low level in sub- sequent months of peak monsoon period (June- September) in considered years from 2012 to 2018. The cloud shape stability which is determined by the pixel orientation based on the neighbourhood of cloud images is empha- sised than the correlation investigation of the original images. When the seven day moving average cloud shape life time is plotted for the months April, May & June for all seven years, a variation can be clearly observed in potential onset period. Moreover, after decrease of the life time, no mo- mentous life time increase or uctuation is observed until the end of June. Student t-test is performed for the cloud life time variation and monsoon arrival dates are considered as the day block where null hypothesis (means are equal) fail to accept. Except 2014, all other years produced a promising output and onset days are May 16 - May 19, Jun 4 - Jun 7, Jun 2- Jun 5, May 13 - May 16, May 15 - May 18 and May 24 - May 27 in year 2012, 2013, 2015, 2016, 2017 and 2018 respectively. Neural network model is also implemented using cloud life time and pre- cipitation data as predictors. Monsoon arrival times found by using neural network model and cloud optical motion variation are agree with the above time period. |
URI: | https://dl.ucsc.cmb.ac.lk/jspui/handle/123456789/4755 |
Appears in Collections: | 2019 |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
MPhil_S Ramanayake2019.pdf | 3.01 MB | Adobe PDF | View/Open |
Items in UCSC Digital Library are protected by copyright, with all rights reserved, unless otherwise indicated.