[A-OS13] Continental Oceanic Mutual Interaction - Planetary Scale Material Circulationn

[AOS13-10] Surface air maximum temperature anomaly prediction over India at medium-range time scale using machine learning techniques

*Venkata Ratnam Jayanthi¹, Swadhin Behera¹, Masami Nonaka¹, Patrick Martineau¹, Kalpesh R Patil¹ (1.Application Laboratory, JAMSTEC)

Keywords:Heatwaves

India experiences high surface air temperatures in the months from March to June which sometimes leads to heatwave-like conditions. Predicting the surface air maximum temperature anomalies at least 10 days ahead (at medium-range time scale) would help the decision-makers and the society as a whole. In this study, we used various machine learning techniques to predict the surface air maximum temperature anomalies over India in the months from March to June. The input attributes to the machine learning models are derived using lag correlation between observed surface air maximum temperature anomalies and sea surface temperature as well as with soil moisture anomalies. The results indicate the predictions of the AdaBoost regressor and the Bagging regressor with Multi-layer Perceptron as the base estimator to have higher correlation along with higher hit rates and lower false alarm rates compared to several other machine learning techniques. The results show the machine learning models to be promising tools to predict the surface air maximum temperature anomalies over India.