A novel thermal index improves prediction of vegetation zones: Associating temperature sum with thermal seasonality

Thermal variation plays a crucial role for governing the type and distribution of vegetation, especially in the humid region. This paper aims to provide a modified thermal index, effective warmth index (EWI) which associate temperature sum with thermal seasonality, compared with commonly used other indices for classifying and predicting of vegetation zones through a case study of Taiwan. With these different thermal indices calculated and mapped at a 100-m spatial resolution, the corresponding climate-vegetation classification schemes are applied to predict the vegetation zones. The accuracy of spatial prediction is evaluated with Kappa coefficient, referring to 651 sampling plots of vegetation. The prediction of potential natural vegetation zones using EWI is the best one (Kappa = 0.759), compared with other indices. This result suggests that thermal seasonality is effective for improving the prediction of warmth index in explaining the altitudinal zonation and distribution of vegetation in Taiwan, and potentially in humid East Asia.