Drought is a recurrent natural phenomenon resulting from prolonged water deficiency. Drought causes destructive impacts on agriculture, ecology, and society. To mitigate the impacts of droughts, it is important to distinguish between different types of drought and to understand how drought propagates from one type to another. Decreasing precipitation combined with higher evaporation rates reduce the root zone soil moisture content leading to agricultural droughts. The deficiency in soil moisture conditions severely affects plant growth and agricultural production. Hence this study attempts to investigate the propagation between meteorological drought and agricultural drought in the Palakkad district of Kerala, India and also to develop a Support Vector Machine (SVM) model for the short-term prediction of agricultural drought. The meteorological drought and agricultural drought are characterized by the Standardised Precipitation Evapotranspiration Index (SPEI), and Standardised Soil moisture Index (SSI), respectively. The Pearson correlation coefficient was used to analyse the propagation relationships between the two types of droughts. The results show that there is a time lag of 1 month for the propagation of meteorological drought to agricultural drought in the study area. The developed SVM model predicted one-month lead SSI with great accuracy (R2 = 0.8). Accurate prediction of agricultural drought helps to take proactive measures to minimize the adverse impacts of the drought event. The findings of this study provide early warning information of agricultural drought. As Palakkad is known as the Rice Bowl of Kerala, it is of great significance to conduct this study for accomplishing, drought resistance, and food security.