Temperatures Comparison Statistical Analysis


The data used in these analyses were collected from Abu Dhabi Statistical yearbook for the year 2009. The data represent average Maximum and Minimum Temperatures in the western area of Abu Dhabi and the Islands of Abu Dhabi in the year of 2008 as show in the table below. The aim of the paper is to analyze the temperature fluctuations within the selected area, and compare the average data of the temperature statistics. Methodology part is mainly focused on the way of data collection and analysis, while results are focused on data interpretation.

Average Max. and Minimum Temp. by Month and Region - 2008


Methodology of data collection and data analysis is based on finding the average mean of temperature variations. Hence, weather forecasts and data levels have been collected, registered and analyzed. This required precise analysis of temperature levels, while the collection process was divided into three stages:

  • Study of the forecasts
  • Study of data archives with weather recordings
  • Regular temperature measurements

These tools helped the research team to find out the values and magnitudes of seasonal temperature fluctuations for further analysis of the values. Probability distribution analysis tool was used for defining the possible ways of temperature fluctuations and magnitude of temperature differences. Considering the fact that both locations were analyzed from the perspectives of minimal and maximal values, probability distribution may be regarded as an effective tool of analyzing the possible values of temperature variations. This is closely linked with the statement that numerical weather predictions are based on precise registering of temperature (or any other parameter) values for a particular period, and the initializing an analysis for offering a range of forecast outcomes.

Results and Conclusion

Forecast analysis involves a yearly temperature valuations, and the data is collected in accordance with the statistical modeling scheme. Initial conditions are registered, while the forecasted values are analyzed in accordance with probability density function.

Linear functions of min and max temperature parameters, offered in regression analysis part are regarded as the necessary simple means parametrized.

Forecast analysis

Forecast analysis

In accordance with these graphs, standard deviation may be analyzed. However, the homogeneity of seasonal data set signifies the necessity to accept or reject a null hypothesis set. Hence, it may be accepted in three data sets, while island minimal temperature is the exception. In accordance with the data achieved, null hypothesis should be rejected.

Two-Tail Test
Lower Critical Value -2.20098516
Upper Critical Value 2.20098516
p-Value 0.00383241

Considering the confidence level achieved during the data analysis (95% in each result), it should be stated that the actual importance of weather data analysis is closely linked with the fact that data deviation parameter is regarded as the necessary aspect of temperature fluctuations. As a rule, comparison of two data sets from two different points is inevitable without cumulative residual of the second data set (island temperature). However, the deviation coefficient reveals the fact that data analysis can not be biased, while the actual importance of probability distribution analysis is explained by the fact that magnitudes of temperature deviations in weather forecasts depend on the size of data set. Hence, the importance and precision of tested sample is associated with the values of standard error, stated in the analysis. Standard error, in its turn is close to two degrees, therefore, there is a need to arrange more accurate observations, and extend the analysis timeline.