Wavelet enhanced time series forecasting for tomato price volatility in South Indian markets

Accurate price forecasting is crucial in the agricultural sector, where farmers face significant challenges due to price volatility. Price fluctuations directly influence the livelihoods of producers and the affordability for consumers, making crop management difficult for farmers. Agricultural time series data are often highly complex and nonlinear, making price prediction, a challenging task. While various forecasting approaches, including stochastic models, machine-learning techniques and hybrid models, have been explored, their effectiveness is often limited due to the inherent complexity of agricultural datasets. Recently, wavelet-enhanced models have emerged as a robust approach, effectively capturing both short-term fluctuations and long-term trends. Wavelet decomposition plays a vital role in denoising data and extracting inherent patterns, thereby improving predictive accuracy. This study investigates the application of wavelet-based models for forecasting the monthly wholesale tomato prices in key South Indian markets such as Bangalore, Chennai and Trivandrum. The findings address the forecasting challenges posed by the volatility of tomato prices, providing valuable insights for stakeholders, including farmers, traders and policymakers, to facilitate informed decision-making. Further, the study highlights the necessity of a robust price policy to stabilize market fluctuations, safeguard farmers’ livelihoods and ensure fair returns. Hence, incorporating advanced forecasting techniques, such as wavelet-based models can significantly improve market stability and promote sustainable agricultural development.