Grouping Weekly Weather Based on Weather Elements in Pagaralam by Using K-Means Clustering Analysis

Abstract

Weather is the result of the interaction and combination of various atmospheric elements that occurs in a relatively small place or region over a short period of time. This study aims to analyze weekly weather characteristics in Pagaralam using the K-Means Clustering method. The data used was weekly weather in 2022 and 2023 with 15 variables, namely Maximum temperature, Minimum temperature, Temperature, Dew, Humidity, Precipitation, Precipitation cover, Wind gust, Wind speed, Wind direction, Sea pressure, Cloud cover, Solar radiation, UV Index, and Moon phase. The analysis process began with data standardization, followed by clustering using K-Means Clustering with several K values to observe variations in cluster structure. The optimal number of clusters was determined using the elbow and silhouette methods. The best optimal K value ​​for each of the 2022 and 2023 data was K=3. A small number of weeks in both years had high temperatures and solar radiation and accompanied by lower dew, humidity, precipitation, and cloud cover than other weeks. A small number of weeks in 2022 had low minimum temperature and were also accompanied by lower cloud cover, dew, and humidity, but they had higher wind gusts and wind speeds than other weeks. Meanwhile, a small number of weeks in 2023 had lower temperatures and accompanied by higher cloud cover, wind gusts, and wind speeds than other weeks.