By combining both the analytical approaches, it can be insinuated that Regions around Kakrapar and Amli in the months of January, March, April, May and June can have very significant drop in the Maximum temperature values and on the other hand, Godsamba and Ukai can have very significant rise in the Maximum Temperature values in the months of July, August and September. This conclusion is based on the fact that the trends detected for the above mentioned regions and Months are either increasing or decreasing for both the regions and months values. The concerning industries can be helped with this analysis to plan their resources efficiently. Both the methods have detected the trends and Sen ’s Slope values come under the confidence interval bracket which strengthens our hypothesis.
Both the non-parametric approaches donot take into account the type of data distribution in the series. One of the problems in detecting and interpreting trends in hydrologic data is the Confounding effect of serial dependence. When it comes to the weather data, there are a lot of factors that are needed to be taken into consideration so as to provide a prominent analysis. For example, the number of trees in the area can severely affect the value of maximum temperature.
For predicting the values of Max. temperature, we used both linear regression and regression trees. And as evident from the R-squared values, the value in the range of 0.3-0.5 is considered good for weather data and hence our data fits well when it comes to linear regression.
On the other hand, For the Regression trees, RMSE estimation values were under 3.0 for both the Global as well as Local model and Hence, we can use the Global model for the prediction of Max. Temperature.