Abstract

Accurate climatic prediction is a need of the day and it is possible if the factors affecting climate are thoroughly investigated. To know the annual temperature trend the data of past 50 years (1961- 2010) of Kohat station, Pakistan was studied. Short-term changes (10 years) and long-term changes (50 years) are analyzed by using regression models. Tenth order polynomial is the best-fitted model for short-term changes in temperature. The cyclical variation of minimum temperature showing a long-term decline and the cycle time span varies from 10 to 15 years. The maximum temperature recorded at the station exhibit cyclic variations with cycles ranging over a decade. Overall trend of the maximum temperature is showing positive trend. The annual mean temperature pattern indicates an overall increasing trend. The trend of fitted models indicate 2.08 o C rise in the maximum temperature, 0.36 o C fall in minimum temperature and 0.86o C rise in the annual temperature during the study period.