A climate predictability index and its applications

Abstract

In this letter, a climate predictability index is proposed. It quantifies the predictability of three major components comprising the climate—temperature, pressure and precipitation. The quantification is done using a fractal dimensional analysis of the corresponding time series. The climate predictability index is calculated for 25 stations spread throughout India using the Global Historical Climatology Network dataset. Change in the index with the seasons suggests a strong influence of more than one climatic dynamics. In such cases, calculations done using mean yearly data are suspect. The index is shown to be useful in studying the interplay between various climatic components (viz. temperature, pressure and precipitation). Changes in predictability indices for temperature and pressure are seen to affect the index for precipitation. The index can be used as a discriminant for determining which stations are selected for use in developing regional climatic models.