Volume No. 7 Issue No.: 2A Page No.: 1100-1116 Oct. - Dec. 2012




Nath C. Dilip and Mwchahary Dimacha D.*

Department of Statistics, Gauhati University, Guwahati (INDIA)


Received on : September 10, 2012




An analysis of 40 years temperature data was performed by applying different statistical methods to study the variation in temperature of Kokrajhar district of Assam in India, over the period 1971-2010. Variations were examined through different sub-periodical windows and the trends of monthly, seasonal and annual temperatures were investigated. Cramerís test was applied to examine the deviation of sub-periodical means from the 40-year long term normal studentís t-test was applied to see if the two sub-periodical means significantly differed from each other Mann-Kendall test was used to detect the trends of temperatures, slopes were estimated by Senís technique and equations of trendlines were obtained by Theil-Sen trendline procedure. The means of both the beginning and ending 10-year non-overlapping, 20-year and 30-year overlapping sub-periods of substantial number of temperature series were significantly deviated from the long term normal. Minimum and mean components of temperature varied in a similar way while maximum temperature varied in an opposite way. For minimum and mean temperatures, beginning sub-period was colder while the ending sub-period was warmer than the long term normal but for maximum temperature, the beginning sub-period was warmer and the ending sub-period was colder than the long term normal. For most of the monthly and seasonal temperatures, the mean of the last sub-period significantly differed from the mean of the first sub-period. Highest number of significant sub-periodical variation was observed in minimum temperature. For monthly minimum and mean temperatures, all significant test-statistic were positive, while such were negative for maximum temperatures. As many as nine monthly, all seasonal and the annual mean temperatures had significant increasing trends at 95% confidence level. The annual and all seasonal minimum temperatures had also significant increasing trends. However, the annual and three seasonal maximum temperatures had significant decreasing trends. Level of minimum temperature was rising while that of maximum temperature was falling. Magnitudes of extremes of hot and cold both are likely to be reduced in coming years. If the trends detected by the current study persist, annual minimum temperature of the district would be increased by about 7.9°C while annual maximum temperature would be decreased by about 3.5°C by the end of the present century.


Keywords : Annual, Maximum, Mean, Minimum, Monsoon, Seasonal, Slope, Sub-period, Trend, Variation, Winter