Published Paper Details:

Earth is a living, breathing planet that changes on a regular basis. Weather patterns and events play a significant role in this development. While these patterns and occurrences are required for our planet to remain habitable, they may also wreak significant damage and cost billions of dollars in repair and rescue attempts. Weather phenomena are natural events caused by one or more of the following: the water cycle, pressure systems, and the Coriolis effect. Precipitation, wind, and heat are frequently involved or related to them. The main aim of this study is tracking and analysing the variations in the weather phenomena by utilizing tree-based diffusion-based processor reallocation strategy. We devised ways for efficiently partitioning and repartitioning the nests among the processors in this paper. We look at an application of tracking numerous ordered cloud clusters in tropical weather systems as a case study. To discover such clouds, we first present a parallel data analysis approach. We created a tree-based hierarchical diffusion method that reallocates processors for the nests at a lower cost of redistribution. We do this via a novel tree restructuring method. We demonstrate that our approach has a lower redistribution cost and 53 percent fewer hop-bytes than a processor reallocation solution that ignores existing processor allocation.Fortran90 programming language have been used for implementation of algorithm.