Sign in Subscribe
Machine learning

Climate Change Research in Rajasthan, India: Machine Learning Techniques Enhance Rainfall Prediction

Strategic Insights

— 1 min read

Climate change poses significant global challenges, requiring precise assessment and prediction to formulate effective mitigation strategies. Researchers at Rajasthan Technical University have leveraged machine learning techniques to evaluate and forecast the climatic variable rainfall in the Thar Desert, located in western Rajasthan, India. The study employed conventional Seasonal AutoRegressive Integrated Moving Average with eXogenous factors (SARIMAX), the machine learning-based Long Short-Term Memory (LSTM) model, and a hybrid model integrating Variational Mode Decomposition (VMD) with LSTM. The results indicate that the VMD-LSTM model outperforms others, achieving an average Nash-Sutcliffe Efficiency (NSE) of 0.58, Root Mean Square Error (RMSE) of 40 mm, and Mean Absolute Error (MAE) of 19.97 mm.

Key Takeaways:

  • The study used machine learning techniques to evaluate and forecast rainfall in the Thar Desert, covering a period from 1901 to 2022 (122 years).
  • The VMD-LSTM model outperformed other models, achieving an average NSE of 0.58, RMSE of 40 mm, and MAE of 19.97 mm.
  • The model demonstrated improved reliability over conventional models in handling non-linear, non-stationary climate data.
  • The five-year forecast using the best-performing model closely aligns with original data trends.
  • The research contributes a data-driven, region-specific framework for rainfall prediction, particularly valuable for climate adaptation planning in desert environments.

Statistics:

  • The study covered a period of 122 years (1901-2022).
  • The VMD-LSTM model achieved an average NSE of 0.58.
  • The model demonstrated an RMSE of 40 mm and an MAE of 19.97 mm.
  • The study focused on the Thar Desert, which covers an area of approximately 2,14,000 square kilometers.
  • The research was peer-reviewed and published in Theoretical and Applied Climatology.

Sources:

  • "Rainfall Prediction In the Context of Climate Change In Thar Desert India Using Machine Learning Algorithms." Theoretical and Applied Climatology, 2025;156(6).
  • Rajasthani Technical University, Department of Civil Engineering, Kota 324010, Rajasthan, India.
  • Hajari Singh, Rajasthan Technical University, Dept. of Civil Engineering, Kota 324010, Rajasthan, India.

Read more