Modeling Temperature and Precipitation in Hyderabad and Medak Using Copula

Abstract

testing. Modeling is done keeping in view the deep association between Temperature and Precipitation, through Copula analysis. The Mean Absolute Percentage Error (MAPE) for the best model in Hyderabad and Medak are found to be 0.04 and 0.09 respectively. R-Software and IBM SPSS statistics version 25 were used to analyze the data and interpret the results. Findings: The best Copula need not be necessarily the same for different data sets. We could find the best Copula for Hyderabad as Rotated Gumbel 270 Copula and for Medak, Frank Copula based on AIC and BIC criteria. The simulated data of Temperature showed a very close agreement with testing data, which can be seen in [Figure 5 (a) & (b)]. Novelty: This type of analysis and model fitting is not found in the literature for Hyderabad and Medak and these districts are in Telangana, the Plateau region. Our fitted models can give a good prediction of Temperature in this region.