Scholarlink

Climate and weather patterns have a significant influence on many aspects of our lives, including infrastructure planning, emergency preparedness, agriculture, and electricity use. It is not just a scientific endeavor but also a practical imperative to comprehend and accurately predict temperature swings in the context of Helsinki, which is a city known for its dramatic seasonal adjustments. Major weather forecasting centers use numerical weather prediction models that operate on many supercomputers to solve difficult nonlinear mathematical equations simultaneously. These models offer medium-range weather forecasts, with a grid length of 10–20 km, every 6 h to 18 h. Nonetheless, agriculture and emergency preparations frequently rely on higher-resolution regional forecasting models to provide them with more precise short- to medium-range projections. This study aims to conduct time series analysis and weather forecasting for Helsinki using facts from 2015 to 2019, utilizing models such as ARIMA, SARIMAX, LSTM, and GARCH. The experimental results based on MAE, MSE, RMSE, R2 score and MAPE, show that the SARIMAX has better performance, which makes it the preferred choice for handling the temporal and geographical intricacies of Helsinki's weather patterns.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Predictive Analytics in Weather Forecasting Using Machine Learning and Deep Learning

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Deepika Pantola,
  • Madhuri Gupta,
  • Mahim Agarwal,
  • Rupal Bohra,
  • Kritika Rawat

摘要

Climate and weather patterns have a significant influence on many aspects of our lives, including infrastructure planning, emergency preparedness, agriculture, and electricity use. It is not just a scientific endeavor but also a practical imperative to comprehend and accurately predict temperature swings in the context of Helsinki, which is a city known for its dramatic seasonal adjustments. Major weather forecasting centers use numerical weather prediction models that operate on many supercomputers to solve difficult nonlinear mathematical equations simultaneously. These models offer medium-range weather forecasts, with a grid length of 10–20 km, every 6 h to 18 h. Nonetheless, agriculture and emergency preparations frequently rely on higher-resolution regional forecasting models to provide them with more precise short- to medium-range projections. This study aims to conduct time series analysis and weather forecasting for Helsinki using facts from 2015 to 2019, utilizing models such as ARIMA, SARIMAX, LSTM, and GARCH. The experimental results based on MAE, MSE, RMSE, R2 score and MAPE, show that the SARIMAX has better performance, which makes it the preferred choice for handling the temporal and geographical intricacies of Helsinki's weather patterns.