Graph-based Local Climate Classification in Iran

TL;DR

This paper presents a novel graph-based climate classification method for Iran that outperforms existing techniques in accuracy and efficiency, and introduces seasonal analysis capabilities.

Contribution

The paper introduces the first graph-based climate classification system, overcoming limitations of previous methods and enabling more detailed and efficient regional climate analysis.

Findings

More realistic climate classification results

Less computational time compared to existing methods

Ability to analyze seasonal climate changes

Abstract

In this paper, we introduce a novel graph-based method to classify the regions with similar climate in a local area. We refer our proposed method as Graph Partition Based Method (GPBM). Our proposed method attempts to overcome the shortcomings of the current state-of-the-art methods in the literature. It has no limit on the number of variables that can be used and also preserves the nature of climate data. To illustrate the capability of our proposed algorithm, we benchmark its performance with other state-of-the-art climate classification techniques. The climate data is collected from 24 synoptic stations in Fars province in southern Iran. The data includes seven climate variables stored as time series from 1951 to 2017. Our results exhibit that our proposed method performs a more realistic climate classification with less computational time. It can save more information during the climate classification process and is therefore efficient in further data analysis. Furthermore, using our method, we can introduce seasonal graphs to better investigate seasonal climate changes. To the best of our knowledge, our proposed method is the first graph-based climate classification system.